,-----------------------------------------------------------------------------
|
| Arachne DOS Coders1 Message Board                           |       |   |
|                                                             |       |   |
| [ For the Operational discussions of the Arachne C Coders ] |       |   |
| ------------------------------------------------------------+-------+---+
|
| ============================================================================
|
| Sat, Jan 01, 2005 - 4:15:33 am 
|
| phantom>
| Happy New Year, folks! (Although things aren't that happy a bit
| north of here actually, rather traumatic in fact).
|
| Thanks for the revised entity code, Glenn. When I get back into
| things of course, I'll incorporate this and Ray's CSS fix in my
| copies of the code.
|
| Gregy, while you're tinkering, some time ago Glenn (I think) also
| had a suggestion for enlarging this text entry window. Perhaps
| you could make it, say, 10 lines long?
|
|
|
| ============================================================================
|
| Sat, Jan 01, 2005 - 9:51:15 am 
|
| ray>
| Yes, it is a bit strange -- we all want to wish a happy new year,
| but given the situation in SE Asia, I thought that the minutes silence
| that some countries observed before the stroke of 12:00 was 
| quite apropritate.
|
| BTW, Gregy, I agree with Joe's suggestion above
|
| ============================================================================
|
| Sat, Jan 01, 2005 - 10:32:41 am 
|
| glennmcc>
| That's an easy one to fix for one-time-use.
|
| F4
| scroll to <textarea........... rows="6" .......>
| change to rows="10"
| F2
|
| ============================================================================
|
| Sat, Jan 01, 2005 - 12:49:31 pm 
|
| ray>
| BTW Glenn, for those changes " return '\'' " seems as good as the form
Š| using " return atoi("39") ". The latter turns a number into a string
| and then back to a number again.
|
|
|
|
| ============================================================================
|
| Sat, Jan 01, 2005 - 12:53:26 pm 
|
| ray>
| ÉÍ»
| º º
| Èͼ
|
| ============================================================================
|
| Sat, Jan 01, 2005 - 8:59:36 pm 
|
| ray>
| Glenn, 
|
| Possible bug: Appending " #end " to the end of my call to this board
| works fine for the first call to it using (here) Shift+F10 or
| clicking on the link in history. However, if I use " R " to reload,
| she beeps twice, doesn't go to the bottom, and in the history list
| I find the tail of the URL as " ...#end#end " -- it is written
| twice. Problem is identical in 1.77.
|
| Thoughts?
|
|
|
| ============================================================================
|
| Sat, Jan 01, 2005 - 10:56:55 pm 
|
| glennmcc>
| Don't hit "R" ...... use the "Send/Refresh" button.
|
| That's why it's there....... to "Refresh" this page. <gggggggggg>
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 2:35:01 am 
|
| gregy>
| Hi....guys....<g>
| ..
| I'm tinkering with glennmcc's PHP-SRC and maybe should have this board
| pasted and uploaded by tomorrow (ooops, today's) night.....<g>
| ...
| well...but...maybe only uploaded to my TEST location to try it out
| for a few posts....but...once I smoke test it, I should be able to
Š| just upload it here and .....voila..!!! ( for francophobes...walla..!!)
| we should be in business with Glenn's nifty improvements....including
| the "preserve spaces" improvement...
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 10:53:54 am 
|
| ray>
| Thanks Gregy.
|
| Glenn, same problem here with the circling arrows button, it has
| exactly the same effect as the "R".
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 10:58:33 am 
|
| ray>
| Problem exists back to 1.73, and thats as far back as I can go, so
| this is nothing any of us have done.
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 2:48:52 pm 
|
| ray>
| I've got a candidate fix; will post to AD. Hey, you guys let me know
| if it's OK, I'm over my depth with these things.
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 3:35:08 pm 
|
| glennmcc>
| I can't seem duplicate what you're getting.
|
| Hitting the Send\Refresh button numerous times still shows only 1 #end
| at the end of the URL when I hit G to place my cursor on the URL line.
| (or left clicking the mouse in the URL line)
|
| The only way I get #end#end#end is by hitting R or linking to the
| 'reload page button' (rotating arrows) instead of the Send\Refresh
| button at the bottom of this page.
|
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 4:33:34 pm 
|
| ray>
Š| Oppps, I assumed 'send/refresh' == rotating arrows, but that's
| obviously wrong now that I ST Anyway, try my fix; seems to
| work OK with 'R' and reload icon too. There is alot of funny stuff
| happening in main.c around that neck of the code. I'm still a year
| away from realy understanding it, but the logic seems quite strange.
| It appears that any URL with an anchor (eg #end) is always loaded
| twice! Then there's that hack I introduced in B~5 to force a reload
| when you're 'F5ing' around. These things shouldn't be necessary.
|
|
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 4:34:11 pm 
|
| ray>
| S
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 4:35:42 pm 
|
| ray>
| Hmmm, It looks like one can't send an ampersand to the board. The
| previous post should have been "S (ampersand) T" (for Stop and Think)
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 4:36:12 pm 
|
| ray>
| (or is that something that I did?)
|
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 6:05:46 pm 
|
| glennmcc>
| Wait till greggy get the 'upgrades' done.
|
| & amp is one of 'em :))
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 6:07:01 pm 
|
| glennmcc>
| Damn bouncy g :(
|
| make that gregy... not greggy
|
| ============================================================================
Š|
| Sun, Jan 02, 2005 - 6:18:23 pm 
|
| gregy>
| hey...as long as it's not "late for lunch"....I'll buy any one of 'em
| ....(especially if it's chocolate for lunch.....<g g g g>.....)
|
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 7:00:31 pm 
| URL
|
| glennmcc>
| Did someone say chocolate ????
|
| Hows about some Triple chocolate cheese cake ??? :))))
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 8:16:02 pm 
|
| gregy>
| hmmmm.....<slobber></slobber> but...I've kinda gotten away from my
| normal choc binge this holiday season.....given that pumpkin pies have
| been on sale a couple times a week at my local supermarket....<g>
| ....
| As a short rest from choc....punkinpie is a FINE finger food...also..
| ...<g g g>
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 9:41:03 pm 
|
| ray>
| Gregy,
|
| Yup, as the Buddha used to say: "Call me *anything* you want
| *except* late for dinner."
|
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 9:47:44 pm 
|
| ray>
| Yeah, I think my fix is good stuff. I figure when MP was writing
| that code in main.c he just overlooked the fact that there might
| be anchors already on some URL's. Or maybe not. Hey, speaking of
| hot keys, it sure would be nice to have a hotkey for 'send/refresh'
| ... or do we already have one? It's a small pain to have to 
| scroll to the bottom of the page before being able to send.
Š|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 9:48:39 pm 
|
| ray>
| Sorry for not shutting up, but I'm trying to see If I can bust
| my fix.
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 9:49:03 pm 
|
| ray>
| Me again
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 9:49:36 pm 
|
| ray>
| Nuts, of course we have a hotkey <bonk side of head> ... F2
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 9:50:26 pm 
|
| ray>
| OK, OK, I'll shut up.
|
|
| ============================================================================
|
| Sun, Jan 02, 2005 - 10:24:50 pm 
|
| gregy>
| Ray....hmmmm...I do NOT understand your "hotkey==F2" for send/refresh
| remark....on my systems...F2==Save.....<g g g>
| ...
| per this board, Glenn has considerably changed his board source, and
| I'm getting back to him, as I go thru the source, to understand how
| it works, and to understand what his new names are for the datafiles
| on the server. So, it is taking a little bit longer than I expected
| to paste/modify his source to use over here on this board....
| ...but...
| not to worry...I'll keep at it until it's done.....actually, it isn't
| that difficult....
| Glenn's redesign is a heap more consistent with standard database
| usage, so, therefore, more understandable to me....but....such things
| as datafile names are still escaping me...
|
|
|
Š| ============================================================================
|
| Mon, Jan 03, 2005 - 10:32:46 am 
|
| ray>
| Gregy,
|
| It's 'save or send' and when I'm on the board, it sends -- quite
| clever ( I wonder how it knows which to do?) 
|
| ============================================================================
|
| Mon, Jan 03, 2005 - 10:35:11 am 
|
| ray>
| Yup, just sent that by pressing F2. 
| Re, the board, looking forward to it, but she's plenty useful right now,
| so no rush. 
|
| ============================================================================
|
| Mon, Jan 03, 2005 - 10:38:21 am 
|
| ray>
| Oh, sorry about baiting L.D, that's ungentlemanly of me and also 
| damned dangerous. I was thinking that if I tease her about her
| harangues she might get the idea that it really isn't helpful.
|
|
| ============================================================================
|
| Mon, Jan 03, 2005 - 9:14:08 pm 
|
| glennmcc>
| F2 only activates the Send/Refresh key if your cursor is in the textarea.
|
|
| ============================================================================
|
| Mon, Jan 03, 2005 - 9:16:56 pm 
|
| glennmcc>
| Or in one of the other text entry fields (nick, email or url)
|
| If the cursor has not yet been placed into one of those fields
| The F2 brings-up the save-as page in Arachne so that you can save or
| send the HTML of this message board.
|
| ============================================================================
|
| Mon, Jan 03, 2005 - 10:12:50 pm 
|
| ray>
| Cool, I like smart, multi-function keys
|
Š|
| ============================================================================
|
| Tue, Jan 04, 2005 - 8:01:44 pm 
|
| ray>
| 
|
| ============================================================================
|
| Tue, Jan 04, 2005 - 8:40:17 pm 
|
| glennmcc>
| SHIT !!!
|
| Since I'm using my 'fresh' compile which converts 32,507 characters into 
|
| Now I gotta look at the SRC of this page by hitting F6 to see what you
| actually entered. <gggggg>
|
| ============================================================================
|
| Tue, Jan 04, 2005 - 8:41:26 pm 
|
| glennmcc>
| OK, kewl.... it really _was_ a 127 <ggggg>
|
|
| ============================================================================
|
| Tue, Jan 04, 2005 - 9:00:02 pm 
|
| ray>
| Just found another bug existing in 1.77 and in b~5:
|
| If you attach a nonexistant file to an email, and then try to send,
| Arachne dutifuly notifies you that there's a problem, however if
| you then hit the big 'go back' icon "<" (an I'm assuming that the other
| ways of going back do the same, but so far I haven't checked) ...
| Arachne goes right ahead and sends the email, complete with 
| "Cannot read input file 'xxx'!"
|
| IMO She should not do that, but should return you to the error page
| again, or to be realy helpful, she should offer you a button on the
| error page for going right back to the message.
|
|
| ============================================================================
|
| Tue, Jan 04, 2005 - 9:07:13 pm 
|
| glennmcc>
| Oh fer gosh sakes Ray.
|
| Hit `M', link to `Outbox' and then to the message in question... problem sol
Š|
| Why don't you just fix your own mistakes instead of modifying Arachne to
| fix them for you ?
|
| I swear... You're getting more 'Window-fied' all the time.
|
| <ggggggggggggg>
|
|
| ============================================================================
|
| Tue, Jan 04, 2005 - 9:09:59 pm 
|
| glennmcc>
| Now, if you'll excuse me.
|
| I'm off to start working on including all 65,536 Unicode characters in our n
| <gggggggggggg>
|
|
| ============================================================================
|
| Tue, Jan 04, 2005 - 10:27:00 pm 
|
| ray>
| Now Now! You know that's one insult I don't abide. I'm quite
| capable of dealing with the above bug, it would never 'catch' me
| but just the same it shouldn't happen. Our software shouldn't
| have little traps in it even if you and me know where all of them
| are and can deal with them. Robust code and good software always
| behave in a consistent and predictable and safe manner. Remember the
| ".\cache" bug? sure, we knew how to deal with it, but that ain't
| good enough, it had to GO! Same here: a message with an invalid
| attachment should not be sent. Beware the arogance of the expert --
| the idea that because you can avoid all the little bugs in the code,
| everyone else should be able to do the same. We loose 'sales'
| becuase people get fed up with Arachne's little quirks, and I for one
| don't want to cope with them, I want to eliminate them.
|
|
| ============================================================================
|
| Tue, Jan 04, 2005 - 10:34:03 pm 
|
| ray>
| <story>
| When the guy who started me with programing sat down to test out the
| very first program that I thought was good enough to show him, you
| can imagine my horror when the first thing he did at the keyboard was
| start randomly tapping away at the keys until it crashed. To make 
| the story short, he went on to explain to me that no program is
| ready to be used until it is robust enough to deal with bad
| inputs without doing something undesired. I learned that lesson.
| He was a senior probrammer with IBM.
| </story>
Š|
|
|
| ============================================================================
|
| Wed, Jan 05, 2005 - 5:57:34 am 
|
| glennmcc>
| OK, OK.... let's fix it. <gggg>
|
| But, how the heck can we disable the 'back button(s)' at such a time ?
|
| Hold it...... I know what we need to do.
|
| All we gotta do is to _not_ put smtp: into the history.lst
|
| Hitting 'back' will then go back to then message.
|
|
|
|
| ============================================================================
|
| Wed, Jan 05, 2005 - 5:50:04 pm 
|
| ray>
|
|
| OK, your project, kill that bug Glenn. You'll find your own solution,
| but my first thought was that it should loop right back to the
| error page ... which BTW I think could be more helpful, for example
| it shouldn't be too difficult to identify the specific problem
| encountered, should it? Especialy if it's a missing attachment. It
| strikes me that a specific message for that problem shouldn't take
| more than a line or two of code to implement ... "if (FileExists( ... "
|
| And, the above reminds, me: It has always seemed to me that there's
| a whole lot of bad stuff added to the history list, SMTP for sure,
| but other stuff too. A good think to bicker about some time ;-)
|
|
|
|
| ============================================================================
|
| Wed, Jan 05, 2005 - 6:53:37 pm 
|
| glennmcc>
| Shit... that was easy.
|
| Niether smtp: nor pop3: will go inot history.lst anymore. :))
|
| ------- in urlstat.c -------
| void add2history(char *URL)
| {
Š| //!!glennmcc: Jan 05, 2005 -- do not add smtp: or pop3: into history.lst
| // if (!URL) return;
| if (!URL || strstr(URL,"smtp:") || strstr(URL,"pop3:")) return;
| //!!glennmcc: end
| ____________________________
|
| Now to work on your next suggestion of having a better error message.
|
|
| ============================================================================
|
| Wed, Jan 05, 2005 - 6:56:33 pm 
|
| glennmcc>
| OK, just took a look at the error message.
|
| What needs to be changed to make it better ?
|
| If you were trying to send e-mail with attached files, one or more files cou
| attached to outgoing e-mail. You may wish to look to outbox and delete the
| incomplete message. 
|
|
|
|
| ============================================================================
|
| Wed, Jan 05, 2005 - 7:05:38 pm 
|
| ray>
| Now that's what I call service!
|
|
|
| ============================================================================
|
| Wed, Jan 05, 2005 - 7:11:42 pm 
|
| ray>
| Ok, so the message basicaly identifies the problem you are having
| as either a bad attachment, or not enough memory. Why not have
| specific messages for both conditions? I doubt that the inner workings
| of the code leave any doubt as to which of the above is the problem
| so why not hit the nail right on the head with a spcific message?
| I remember when I started using Arachne, these, multiple choice
| error messages really bothered me; we should be as focused as possible
| in identifying errors, no?
|
|
| ============================================================================
|
| Wed, Jan 05, 2005 - 7:21:26 pm 
|
| glennmcc>
| No. <g>
Š|
|
| ============================================================================
|
| Thu, Jan 06, 2005 - 2:48:49 pm 
|
| ray>
| Well, gota disagree on that. IMO multiple choice error messages
| are only slightly better than no message at all.
|
| ============================================================================
|
| Thu, Jan 06, 2005 - 4:21:49 pm 
|
| glennmcc>
| Well then, the easy fix would be to make it 'no message at all'.
|
| We'll just do like Windows does and NOT do what the user asked and NOT
| give any message of any kind that we did not do it. <gggggggggggg>
|
| ============================================================================
|
| Thu, Jan 06, 2005 - 4:24:11 pm 
|
| glennmcc>
| Hey, that gives me a great idea.
|
| BRB
|
|
| ============================================================================
|
| Thu, Jan 06, 2005 - 4:34:46 pm 
| URL
|
| glennmcc>
| There we go.
|
| Is this better. <ggggggg>
|
| ============================================================================
|
| Thu, Jan 06, 2005 - 5:58:58 pm 
|
| ray>
| That's hunky-dory for the attachment error, now all that would be needed
| is an aditional message that is specific to the memory problem and
| I'm as happy as a clam, since there would no longer be any need to
| guess which error was the problem. BTW, I think it's nice that there
| are big fat html pages to display these messages, rather than just
| status line messages, becuase they are an oportunity to not only tell
| the user what went wrong, but to offer sage advice about how to
| prevent it in the future -- there is the oportunity to educate the user. I'd
| Arachne's error
| pages are very good in this regard
Š|
| ============================================================================
|
| Thu, Jan 06, 2005 - 6:55:18 pm 
|
| glennmcc>
| It was joke..... and you did'nt get it. :(
|
| <title> Arachne BSOD </title>
|
| <body bgcolor=blue ........
|
|
| ============================================================================
|
| Thu, Jan 06, 2005 - 8:14:44 pm 
|
| ray>
| Hmmmm, so I didn't.
|
|
|
| ============================================================================
|
| Fri, Jan 07, 2005 - 6:07:06 pm 
|
| ray>
| Glenn,
|
| I think I've got the start of a better fix for the non existant 
| attachment. Can you make this work better, like by having the
| code break back to the compose screen?
|
| ÉÍ[þ]ÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍ CGIQUERY.C ÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍ2Í[
| º} 
| º// !!glennmcc: end ±
| ºelse if (!strcmpi(cmd, "SUBJ") && mailmsg >= 0) ±
| º{ ±
| º sprintf(str, "Subject: %s\n", value); ±
| º write(mailmsg, str, strlen(str)); ±
| º} ±
| ºelse if (!strcmpi(cmd, "ATTACH") && mailmsg >= 0 && value[0]) ±
| º{ ±
| º if (!FileExists(value)) ±
| º { þ
| º Beep(); ±
| º printf("Sorry idiot, but the file \" %s \" doesn't exist!\n" ±
| º "Press any key to continue ...", value); ±
| º //Outsf("Sorry idiot, but the file %s doesn't exist", value); ±
| º getch(); ±
| º // Would it be possible to just return to the mail compose ±
| º // screen from here to repair the faulty attachment? ±
| º } ±
| º sprintf(str, "X-Attachment: %s\n", value); ±
| º write(mailmsg, str, strlen(str)); 
Š| ÈÍ ÍÍÍ 315:1 ÍÍÍÍÍ þ±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±
| F1 Help F2 Save F3 Open Alt-F9 Compile F9 Make F10 Menu 
|
|
|
| ============================================================================
|
| Fri, Jan 07, 2005 - 6:48:41 pm 
|
| glennmcc>
| That has got to be one of the biggest messes I have seen in a looooonnnnnngg
|
|
| ============================================================================
|
| Sat, Jan 08, 2005 - 10:16:14 am 
|
| ray>
| Just checking to see if Gregy had the spacing fixed. Should have
| trimmed of the 'box' characters, of course. Anyway, if you can tell
| me how to return to the compose screen from there I'll be in your
| debt.
|
|
| ============================================================================
|
| Sat, Jan 08, 2005 - 6:39:34 pm 
|
| ray>
| Glenn, 
|
| No ideas, or do you just not want to help?
|
|
| ============================================================================
|
| Sat, Jan 08, 2005 - 8:29:12 pm 
|
| glennmcc>
| Got a couple of ideas.
|
| I'm working on it. :)
|
|
| ============================================================================
|
| Sat, Jan 08, 2005 - 9:21:18 pm 
|
| ray>
| Goody!
|
| This is quite over my head. I have one idea that I don't want
| to even try since is would involve my tinkering with other files 
| besides core.exe: nonfatal.ah is called by both insight.exe errors and
| by MemErr() because both of them restart arachne with errorlevel == 128.
Š| If say MemErr restarted with say errorlevel == 129, then it could
| call a different .ah file with a focused message, but that would
| involve adding new switches.
|
|
|
| ============================================================================
|
| Sat, Jan 08, 2005 - 9:24:28 pm 
|
| ray>
| What I'd really like to know is why MemErr() needs to go through
| this whole resart hullabaloo. Why can'it it just display a message
| and be done with it? Maybe, if needed it could restart after displaying
| it's message, so the restart wouldn't have to involve switches
| and such. I dunno.
|
|
| ============================================================================
|
| Sun, Jan 09, 2005 - 4:26:38 pm 
|
| glennmcc>
| Well, I've been working on it for the last 2hrs and no luck. :(
|
| Forgive my frustration but.....
|
| Screwit !
|
| Since niether pop3 nor smtp are going to written into history.lst anymore
|
| IMO, It's good enough the way it is.
|
|
| ============================================================================
|
| Sun, Jan 09, 2005 - 5:04:38 pm 
|
| ray>
| Yeah it's nothing important, one of those polishing things that I 
| like to think about. There are more important things to do, for sure,
| although for myself, the big payoff in cracking this think would be
| just in the undertanding gained of how Arachne moves herself around.
| BTW I ended up not likeing that restriction on pop3 and smtp, I found
| that some buttons had to be hit twice to work with the restriction on --
| though again, that shouldn't happen, there is some hard link between
| going some where and that somewhere being in the history list that
| shouldn't be needed. 
|
| ============================================================================
|
| Sun, Jan 09, 2005 - 6:18:28 pm 
|
| glennmcc>
| Ah HA !
Š|
| By George...... he's got it.
|
| We now go to either of 2 different error pages. :))
|
| --- in init.c ---
| else if(argv[1][2]=='e')
| {
| sprintf(GLOBAL.location,"file:%s%snonfatal.ah",sharepath,GUIPATH);
| }
|
| //!!glennmcc: begin Jan 09, 2005
| // on email attachment errors, goto err_mail.ah instead of nonfatal.ah
| else if(argv[1][2]=='m')
| {
| sprintf(GLOBAL.location,"file:%s%serr_mail.ah",sharepath,GUIPATH);
| }
| //!!glennmcc: end
| _________________
|
| --- in misc.c ---
| if(tcpip)
| //!!glennmcc: begin Jan 09, 2005
| // on email attachment errors, goto err_mail.ah instead of nonfatal.ah
| strcat(cmd,"rm");
| else
| strcat(cmd,"cm");
| //strcat(cmd,"re");
| //else
| //strcat(cmd,"ce");
| //!!glennmcc: end
| }
| _________________
|
|
| --- system\gui\err_mail.ah ---
| <HTML>
| <TITLE>Arachne Non-fatal Email Error</TITLE>
| <BODY TEXT=BLACK>
| <IMG SRC="file://system\gui\error.gif" ALT="ERROR" ALIGN=RIGHT>
| <H1>Non-fatal Error</H1>
| <BR CLEAR=RIGHT>
| <HR SIZE=6>
| We appologise for inconvenience,
|
| Arachne was not able to complete
| requested operation due to invalid input data.
|
|
| Suggestions:
| <UL>
| <LI>If you were trying to send e-mail with attached files, one or
| more files couldn't be attached to outgoing e-mail.
|
| You may wish to look to
Š| <A HREF="reload:file://outbox.dgi">OUTBOX</A> to either edit or delete the i
| </UL>
| <HR>
| See <A HREF="file://help.htm">Arachne documentation</A> for more info.
| </BODY>
| </HTML>
| ______________________________
|
|
| ============================================================================
|
| Sun, Jan 09, 2005 - 8:03:00 pm 
|
| glennmcc>
| --- clipped from my newly modified changes.txt ---
|
| 1.81;GPL,386+ stable for DOS
| ----
| Release date: ??? ??, 2005
|
| Dec 29, 2004 - Jan 05, 2005
| Several changes for supporting 'Unicode punctuation'.
|
| Jan 05, 2005
| To prevent the 'back button' from inadvertently sending an incomplete email.
| (when an error occurred during attachment process)
|
| And to prevent the 'back button' from inadvertently downloading email again.
|
| "smtp:" and "pop3:" will no longer be written into history.lst
|
| Jan 09, 2005
| 'nonfatal.ah' will now be used strictly for memory errors. 'err_mail.ah' wil
| now be used for errors occurring during making email attachments.
|
| ??? ??, 2005
|
| ??? ??, 2005
|
| ??? ??, 2005
|
|
| ============================================================================
|
| Sun, Jan 09, 2005 - 8:34:03 pm 
|
| ray>
| God it's scarry when we think alike ;-) ... I was just about
| to write "by George, I think I've got it"
|
| I've got a solution too, that I think is quite nice. I'll examine
| your code above tomorrow, but it boils down to the same outcome --
| two different .ah's. This nice thing about mine, is it actualy
| shortens the code, too.
|
Š| Manyana
|
|
|
| ============================================================================
|
| Sun, Jan 09, 2005 - 8:37:03 pm 
|
| ray>
| Well, I couldn't help but look ;-)
| Yeah, I was going to use a new argument, too, and maybe that's best.
|
| Nitey Nite Glenn
|
| ============================================================================
|
| Sun, Jan 09, 2005 - 8:59:44 pm 
|
| gregy>
| no...gregy hasn't got this board updated for proper formatting, yet....
| ...I got sidetracked for a couple days....
| ...
| ...and...I'm trying to modify Glenn's code to "un-autoarchive"....<g>
|
| ...but...I may just temporarily go with his whole shooten-match...just
| to give me time to work thru his code...
| ...
|
|
|
| ============================================================================
|
| Mon, Jan 10, 2005 - 5:15:26 pm 
|
| glennmcc>
| You want it so it does not 'auto-archive' ?
|
| No problem.
|
| Remove this entire section.
|
| -------------
| $lastday = date("M d g:i a");
| if (
| $lastday == "Jan 31 11:59 pm" ||
| $lastday == "Feb 28 11:59 pm" ||
| $lastday == "Feb 29 11:59 pm" ||
| $lastday == "Mar 31 11:59 pm" ||
| $lastday == "Apr 30 11:59 pm" ||
| $lastday == "May 31 11:59 pm" ||
| $lastday == "Jun 30 11:59 pm" ||
| $lastday == "Jul 31 11:59 pm" ||
| $lastday == "Aug 31 11:59 pm" ||
| $lastday == "Sep 30 11:59 pm" ||
| $lastday == "Oct 31 11:59 pm" ||
Š| $lastday == "Nov 30 11:59 pm" ||
| $lastday == "Dec 31 11:59 pm")
| {
| $almost = date("g:i:s a ");
|
| echo"
| <center>
|
| <b>Current local time is $almost</b>
|
|
| <b>It's almost midnight....... the board will be up any second now.</b>
| </center>
| "
| ;
| Die();
| }
| else
|
| /////////begin caculate which numbered file to use/////////
| //change startyear and startmonth to suit your situation.
|
| $startyear = "2003";
| $startmonth = "1";
|
| $archivenumber = date("m")+((date(Y)-$startyear)*12)-($startmonth-1);
|
| ////////////end caculate file number////////////
|
| if ($archivenumber <10) 
| { 
| $countfile = "archives/count00$archivenumber.txt"; 
| $postfile = "archives/posts00$archivenumber.txt"; 
| } 
| else 
| { 
| $countfile = "archives/count0$archivenumber.txt"; 
| $postfile = "archives/posts0$archivenumber.txt"; 
| } 
|
| -------------
|
| And use just these 2 lines instead.
|
| $countfile = "count001.txt"; 
| $postfile = "posts001.txt"; 
|
| Now upload 2 zero-byte files.... count001.txt and posts001.txt
|
| There ... you're done. :)
|
|
| ============================================================================
|
| Mon, Jan 10, 2005 - 5:26:18 pm 
Š|
| ray>
| Glenn,
|
| Want to see my solution? (Won't bother you if you don't)
|
|
| ============================================================================
|
| Mon, Jan 10, 2005 - 6:14:04 pm 
|
| ray>
| Hey Gregy, I'm just now noticing that even though I have '#end'
| at the end of my call to the board, it doesn't go all the way down.
| I expect this is just part of debuging the newer code for the board,
| but thought I'd mention it.
|
|
|
| ============================================================================
|
| Mon, Jan 10, 2005 - 6:31:16 pm 
|
| ray>
| Don't know if you did anything, but that's much better.
|
| ============================================================================
|
| Mon, Jan 10, 2005 - 8:11:31 pm 
|
| glennmcc>
| That's becuase "end" is not at the end. <g>
|
| It's at the 2nd message from the end.
|
| re: Do I want to see your solution ?
|
| Well, of-course I wanna see it. :))
|
| ============================================================================
|
| Mon, Jan 10, 2005 - 8:53:22 pm 
|
| ray>
| Yeah, I now know that '#end' can be put anywhere, it's just that
| it was too high before, and it's nice now.
|
| OK my solution comming up.
|
| But first, to wax philosophical, I think this is the first time
| we've both tried to solve the same problem at the same time so as
| to make it possible for us to compare our aproaches directly.
| And, I find that we are both exactly in character:
|
| Your method is robust, cautious, and additive in nature.
Š| My method is simple, experimental, and subtractive.
| I think my method is better *if* it works with no hangups, but
| yours is more likely to be bombproof. But typical of me, I'm quite
| happy to try something that might not work, because if it fails,
| there will be a lesson in it.
|
|
|
| ============================================================================
|
| Mon, Jan 10, 2005 - 9:17:04 pm 
|
| ray>
|
| One line in arachne.bat:
|
| echo Don't worry, this is just Arachne lock file... > lock 
|
| :continue 
| if (%TRACE%) == (on) echo >> trace.log arachne %1 %2 
|
| core.exe %1 %2 
| if errorlevel 129 arachne.bat system\gui\memerr.ah 
| if errorlevel 128 goto fastexecute 
| if errorlevel 64 goto execute 
| if errorlevel 32 goto end 
| if errorlevel 16 goto graphics 
| if errorlevel 8 goto resolution 
|
| ============================================================================
|
| In errors.c (base~5):
|
| // Widely used. 
| // RAY: To test this, comment line 791 in htmldraw.c 
| void MemErr(void) 
| { 
| // RAY: Try skipping the song and dance and just exit ;-)
| // char command[80]; 
| // command[0] = '\0'; 
| SavePick(); 
| IEsaveF(s 
| IEsaveBin(s); 
| MemoryDestroy(); 
| // CloseBat(command, RESTART_NONFATAL_ERROR); 
| // exit(WillExecute(command)); 
| exit(129); 
| } 
|
| ============================================================================
|
| My thinking is that there is no point in invoking CloseBat() and
| WillExecute() since there *is* no command to execute -- all that the
| resulting '$roura$.bat' does/did is to restart Arachne with the
| switches that point the code to 'nonfatal.ah' but the same thing can
Š| be done as above with less trouble, more speed, and this aproach
| simplifies the code and is much more expandable -- if we ever need
| more .AH files for more helpful messages, we need only 'exit(130),
| exit(131)' etc. without the need to add code. 
|
| I came to realize that this aproach won't work with the attachment
| problem since even if there is an error, insight must still be run to
| generate the outbox screen, so CloseBat() is needed to make it possible
| for Arachne to capture the errorlevel returned by insight.exe. Thus I
| have left the call to nonfatal.ah exactly the same, although now its
| contents should obviously discuss the attachment problem only.
|
| Now, have I overlooked something?
|
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 4:33:39 pm 
|
| Joe>
| Ray,
|
| I'm not sure how important this is, but MP deliberately chose error
| exit codes to correspond to a specific bit position (128 = bit 7,
| 64 = bit 6, 32 = bit 5, and so on). However, your proposal for error
| exit code 129 deviates from this convention, which raises some (soft)
| alarm bells for me. It is possible that I'm reading too much into what
| MP did here, but I thought I should raise this, particularly as Glenn
| may have some idea/comment on this.
|
| Incidentally, in my current "alpha", I have redefined error exit code
| 0 as "inadvertant/unexpected exit", with the exiting ("fatal" or some
| such) error exit code 0 bumped up to 1 ...
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 4:42:22 pm 
|
| Joe>
| Glenn,
|
| Cal/Destiny's posting at freelists has reminded me that
| "http://www.drdosprojects.de" is where the "native FAT32
| project for OpenDOS" lives ...
|
| If you have backup facilities for your hard disk (just in
| case), you may be interested in trying the above. AFAIK,
| this also uses Extended INT 13 BIOS services (eg. LBA) to
| break the 8G barrier of the legacy DOSes (you can't really
| have 20 partitions of 2G each otherwise, since only the
| first 8G will be accessible via the legacy INT 13 services).
|
|
Š|
| ============================================================================
|
| Tue, Jan 11, 2005 - 4:49:48 pm 
|
| ray>
| Joe,
|
| Yeah, I know he did that -- it's hard to miss, and as with other
| things of this nature, it's hard to believe there isn't some
| reason for it, but it seems to work and I figure if there is some
| problem it will show itself sooner or later (hopefuly sooner).
| Maybe older DOS's only accept single bit errorlevels? Anyway, I'm
| much looking forward to seeing what you've been up to. Now, off to
| retrieve your .lib. BTW, isn't it customary that errorlevel 0 is
| used for problem free exits? Same with return() ... 0 == OK.
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 6:07:50 pm 
|
| ray>
| Joe,
|
| On line with your .lib now; nothing to report, all is well.
| Your .lib costs about 1K of memory, but that's a small price to
| pay for some squashed bugs.
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 6:56:37 pm 
|
| glennmcc>
| 1st off ......... What .LIB ??????????????? <gggg>
|
| 2nd off ......... As you might have suspected...... I'll stay with my
| safe & secure, works every time method for splitting nonfatal.ah into
| nonfatal.ah and err_mail.ah
|
| re:
| "Thus I have left the call to nonfatal.ah exactly the same, although now
| its contents should obviously discuss the attachment problem only.
|
| Now, have I overlooked something?".
|
| Yep..... you overlook the fact that nonfatal.ah is still going to be
| 'called' by all of the memerr() occurrences in the code if a memory
| error happens in any of those places.
|
| Or am I wrong in that conclusion ?
|
|
| ============================================================================
Š|
| Tue, Jan 11, 2005 - 7:10:08 pm 
|
| glennmcc>
| Nevermind on '1st off' above.
|
| Just now D/Led it from Yahoo.... will try it later this evening.
|
| BTW,
|
| Both v1.77 and v1.79 use Michal's build of hixlov.lib dated Sep 2004
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 7:29:51 pm 
|
| glennmcc>
| There we go.... I'm running a fresh compile right now with hixlov.387
|
| My mem stats are as follows.
|
| Michal's hixlov.lib ---- core.exe uses 452,992 bytes.
|
| Joe's hixlov.387 ------- core.exe uses 452,992 bytes.
|
| That's right.............. no difference on my system.
|
| Ray, did you compile with hixlov.387 or hixlov.086 ?
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 7:38:28 pm 
|
| glennmcc>
| Nope... that ain't the reason for the 1k diff you got.
|
| I just recompiled with hixlov.086 and core.exe is now using 456,064 bytes.
|
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 7:40:10 pm 
|
| glennmcc>
| Joe,
|
| re: http://www.drdosprojects.de
|
|
| Yep, I read that post and placed that link in my hotlist.
|
| I'll be doing exactly as you suggest in the very near future.
Š|
| ============================================================================
|
| Tue, Jan 11, 2005 - 7:43:38 pm 
|
| glennmcc>
| Ray,
|
| re: "...first time we've both tried to solve the same problem at the same ti
|
| This is getting scarry. <gggggggggggg>
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 9:36:07 pm 
|
| ray>
| Glenn,
|
| I'll double check my mem usage with Joe's new lib, your probably 
| right. As for nonfatal.ah, your wrong (I hope), because MemErr
| returns errorlevel 129 which restarts arachne with 'memerr.ah' via
| that new line in arachne.bat. Only attachment errors should
| invoke nonfatal.ah. 
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 9:46:02 pm 
|
| glennmcc>
| OK... _now_ I understand.
|
| Our solutions are 'backward from eachother'
|
| I changed it so that attachment errors go to err_mail.ah and memerr()
| stays on nonfatal.ah
|
| You changed it so that memerr() goes to memerr.ah and attachment errors
| stay on nonfatal.ah
|
| I still like mine better for 2 reasons.
|
| 1) it requires no new errorlevels
| 2) it makes no changes to arachne.bat
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 9:54:54 pm 
|
| ray>
| Ooops, got it wrong, Joe's lib gives 224 bytes more free, and
| loaded size is 423,562. My free mem dances up and down for 
| reasons that I'm not quite familiar with, less on line, less after
| I've run some other programs -- is it just the changes in the
Š| environment size?
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 10:01:11 pm 
|
| glennmcc>
| Perhaps that's it.
|
| Mine also 'dances' a bit depending upon wha I'm doing at the time.
|
| 190,656 right now.
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 10:01:51 pm 
|
| ray>
| As to the other, both of us like exactly what we could be expected to
| like, as I was saying before. However, mine is provisional, since
| there could be some defect, so you might find me retreating to
| your code with a bloody nose eventualy.
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 10:03:41 pm 
|
| glennmcc>
| Hung-up, exited, restarted, redialed and came back here.
|
| Now I'm at 190,896
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 10:06:12 pm 
|
| glennmcc>
| Exited just now without hanging up, restarted and came back here.. 190,688
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 10:06:24 pm 
|
| ray>
| Now, the next bug I'm on to is the crash that happens if wwwman is
| asked to list the files in any directory that has more than 1024
| files in it. I figure that the first thing to do is to stop the
| crash -- even if there is a 1024 limit, Arachne should politely
| give us a message (errorlevel 130 ;-)))))) that we need better
| disk organisation. 
|
Š|
| ============================================================================
|
| Tue, Jan 11, 2005 - 10:07:59 pm 
|
| ray>
| Should be easy enough to add up environment sizes and compare with
| free mem -- these variations are roughly what you'd expect from
| environment changes no?
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 10:09:31 pm 
|
| glennmcc>
| Oh... don't get me wrong on this.
|
| I don't _hate_ your method.
|
| It's just that I like mine better than yours. <gggggg>
|
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 10:11:03 pm 
|
| glennmcc>
| Oh shit, there he goes adding yet another errorlevel !!! <gggggg>
|
| ============================================================================
|
| Tue, Jan 11, 2005 - 10:39:38 pm 
| URL
|
| glennmcc>
| Is this better ?
|
| ============================================================================
|
| Wed, Jan 12, 2005 - 1:28:00 am 
|
| Joe>
| Hi all,
|
| Firstly, about single-bit error values - No, there has never been
| a DOS version that had such a limitation, this is a convention of
| MP/Arachne, whose significance is unknown to me (but as usual, I
| err on the side of caution;-).
|
| As to the memory usage with the new Lopif/Hixlov library, as I recall
| (it's now been several months since I've used the old library, as it
| doesn't work properly for me at 800x600), there was no difference as
| compared to the old library.
|
Š|
| ============================================================================
|
| Wed, Jan 12, 2005 - 1:31:32 am 
|
| Joe>
| PS. AFAIK, DOS only makes one byte (ie. 8 bits) of the error
| code value available (although I think there's a high-order
| byte used internally by DOS for its own purposes).
|
| ============================================================================
|
| Wed, Jan 12, 2005 - 1:54:22 am 
|
| glennmcc>
| I got it. :))
|
| wwwman.exe now supports dirs of up to 1536 items and will list only
| the first 1536 items in any dir exceeding 1536
|
| She auto-disables the Icons for any dir over 512 items.
|
| http://www.cisnet.com/glennmcc/my-stuff/512.zbm
| http://www.cisnet.com/glennmcc/my-stuff/513.zbm
| http://www.cisnet.com/glennmcc/my-stuff/2048.zbm
|
| ============================================================================
|
| Wed, Jan 12, 2005 - 11:29:33 am 
|
| glennmcc>
| New additions to changes.txt
|
| -------
| Jan 11, 2005
| Recompiled core.exe to use the new version of the graphics library file
| built by Joe da Silva. This new library has better support for an even
| larger variety of video cards.
|
| Jan 12, 2005
| Wwwman.exe will now handle directory listings of 1536 items.
| (old limit was 1024 items)
| Also, it will no-longer lockup or crash when there are more items
| than the limit. When a dir listing is attempted on a dir containing
| more than 1536 items, those first 1536 will be shown.
|
|
| ============================================================================
|
| Wed, Jan 12, 2005 - 6:18:48 pm 
|
| ray>
| Bloody marvelous. I was looking in core.exe, so I was probably wasting
| time. And you say she not only lists 1.5K but won't crash if there's
| more? That's killing two birds with one stone.
Š|
| I'm horny to see it! Please post code ASAP. 
|
| ============================================================================
|
| Wed, Jan 12, 2005 - 7:25:05 pm 
|
| glennmcc>
| Nope... actually it's 3 with 1 stone.
|
| increased limit to 1536 
|
| turn off icons if more than 512
|
| 'break' if the limit is exceeded and list what can be listed.
|
| Code coming up.
| (you're gonna shit when you see how freakin' simple it was) <gggg>
|
|
| ============================================================================
|
| Wed, Jan 12, 2005 - 7:42:43 pm 
|
| glennmcc>
| I hope this comes out OK on this message board.
| (had to do some formatting mods)//added some comments to indicate them
|
| --- in wwwman.cpp ---
|
| <snip>
|
| //!!glennmcc: Jan 12, 2005 -- increase limit by 50%
| #define maxdiritems 1536
| //#define maxdiritems 1024
|
| <snip>
|
| else
| isroot = 1;
| //!!glennmcc: Jan 12, 2005 -- turn icons off if more than 512 items
| if (t==513)
| {
| printf("</table><b>More than 512 items... Icons auto-disabled for this direc
| //[br] at the end of the previous line also converted for this posting.
| pics=0;
| }
| //glennmcc: end
| if(t >= maxdiritems)
| break;
| if(!d)
| ff.ff_attrib = FA_DIREC;
| done = findnext(& ff);//space inserted after & for this posting
| }
| //if(usedesc)assigndesctodir(dir,t,fdesc);
Š| if(!done)
| {
| //!!glennmcc: Jan 12, 2005 -- break if too many items
| printf("<TD><B>Ugh! Unbelievably large directory."
| " Only listing the first %d items.</B>",maxdiritems);
| break;
| //printf("<TD><I>Ugh!Unbelievable large directory, list will be splitted.</I
| //items++;
| //!!glennmcc: end
| }
| sortdirectory(dir, t);
|
| <snip>
|
| -----
|
| That's it.
|
| Change 3 lines of code.
|
| Add 3 lines of code.
|
|
| ============================================================================
|
| Wed, Jan 12, 2005 - 8:53:13 pm 
|
| ray>
| Ahh, got it. This is a major improvement. Can't try it out till
| tomorrow unfortunately. Man, are we making hay lately.
|
|
|
|
| ============================================================================
|
| Thu, Jan 13, 2005 - 5:45:01 pm 
|
| ray>
| Glenn,
|
| That's a beauty. It looks like the code was just buggy there.
| BTW Didn't you mean to use " 
| " not " [br] " 
| Also, I managed to tweek it up to a max of 1664 files, much beyond
| that and I get a very polite 'farcoreleft' error message, but she
| doesn't crash, which is very nice.
|
|
| ============================================================================
|
| Thu, Jan 13, 2005 - 6:04:34 pm 
|
| ray>
| Heh, I can't enter an HTML tag in the text ... that makes sence
Š| actualy, what I meant to say above was " less than sign 'br' greater
| than sign "
|
|
|
| ============================================================================
|
| Thu, Jan 13, 2005 - 6:29:27 pm 
|
| glennmcc>
| That's what this meant..... <ggggggg>
|
| //[br] at the end of the previous line also converted for this posting.
|
|
| ============================================================================
|
| Thu, Jan 13, 2005 - 8:10:27 pm 
|
| ray>
| AHHHH. Should have figured that out ... if I can't send a tag
| it's kinda obvious you can't send one neither!
|
| ============================================================================
|
| Thu, Jan 13, 2005 - 8:20:42 pm 
|
| glennmcc>
| Not to worry... as soon as Gregy makes the changes to this board, then
| we won't need to use [br] anymore. :)
|
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 11:57:06 am 
|
| ray>
| Glenn,
|
| Can you point me to any documentation on how to use Arachne scripts
| and what commands are alowed there? 
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 3:04:28 pm 
| URL
|
| glennmcc>
| Nope...... no idea _where_ you might find such info as that. <gggggg>
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 3:05:20 pm 
|
| glennmcc>
Š| Damn...... didn't work. :((
|
| file:doc\acf_asf.htm
|
|
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 4:02:54 pm 
|
| ray>
| Gracias.
|
| I want to take a closer look at changes that could affect speed, 
| so I'm planning to make a script that does a whole buncha things
| that I can time.
|
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 5:26:47 pm 
|
| ray>
| That's a nice file, all that is lacking is a list of available commands,
| like arachne:dial, arachne:?
|
| BTW, here's my take on the memory error file:
|
|
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 5:27:34 pm 
|
| ray>
| <HTML>
| <TITLE>Arachne Memory Error</TITLE>
| <BODY TEXT=BLACK>
| <IMG SRC="file://system\gui\error.gif" ALT="ERROR" ALIGN=RIGHT>
| <H1>Arachne didn't have enough memory for the last operation!</H1>
| <BR CLEAR=RIGHT>
| <HR SIZE=6>
| Suggestions:
|
|
|
| <LI>Try to reconfigure your system to offer more conventional DOS memory.
|
| Click here: <A HREF="reload:file://memc.dgi">mem</A> then 
| look for: "Largest executable program size". 
| If the number shown is less than about 50,000 bytes, you need to
| free up more conventional memory.
|
|
|
Š| Alternatively, you can type the "mem" command from the DOS prompt. 
| In this case, the "Largest executable program size" must be at 
| least 500,000 bytes. (Since Arachne herself is not yet loaded).
|
|
|
| Normally, the best way to free up more memory is to either load 
| fewer device drivers in CONFIG.SYS, and/or to use "loadhigh" or "lh" in 
| front of the lines that load your drivers. Remember that the more memory
| Arachne has, the better she will run. See your DOS documentation for
| more details.
|
|
|
| <LI>With your system configuration as it is, you should probably use Arachne
| extensively - only to view simple HTML pages.
| </UL>
| <HR>
| See <A HREF="file://help.htm">Arachne documentation</A> for more info.
| </BODY>
| </HTML>
|
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 5:29:37 pm 
|
| ray>
| $hit, as predictable, the BR's didn't show.
| Anyway, if you like it enough to keep it, edit it to suit and
| return, SVP.
|
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 6:19:22 pm 
| URL
|
| glennmcc>
| Done editing.....
|
| What do you think ?
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 6:30:26 pm 
| URL
|
| glennmcc>
| How about this one.
|
| Fancied-up. <gggggg>
|
|
| ============================================================================
Š|
| Sun, Jan 16, 2005 - 7:37:05 pm 
|
| glennmcc>
| re".... list of available commands,"
|
| Anything can be placed in there...... anything at-all.
|
| If it works by typing it into the URL bar and pressing enter..... it will wo
| If it works by clicking a link to it.... it willwork in an .ASF
|
| _All_ valid URLs and protocols... no exceptions.
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 7:39:15 pm 
|
| glennmcc>
| BTW,
|
| Hitting 'Q' creates an .ACF out of all links on the current page.
|
| The ASF is then run and all links on the current page get downloaded one-aft
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 7:41:11 pm 
|
| glennmcc>
| But remember........ running any ASF (including the one that gets run
| automatically by hitting 'Q'), destroys your current history.lst
|
|
|
|
| ============================================================================
|
| Sun, Jan 16, 2005 - 9:45:26 pm 
|
| ray>
| Like the page -- new style and all. Do you agree with my numbers?
| Let's make the figures shown reflect realistic minimums -- I was
| just estimating.
|
| re .ASF's, yeah, but there are only so many words that can follow
| the 'arachne:???' right? Is there a list some where?
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 8:45:23 am 
|
| glennmcc>
| Oh... I see what you mean now.
|
Š| Nope, we're gonna need to make a list ....... someday. <ggggggggg>
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 8:51:11 am 
|
| glennmcc>
| Yep... 500,000 is a realistic figure.
|
| Shall we pole the users and see which style they like ?
|
| (tables vs no tables)
|
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 5:33:47 pm 
|
| ray>
| OK, 500 grand it is. As to style, sure, why not let the
| howling mob decide ;-)
|
| BTW, my little goto yesterday sux. There is a better one line
| solution: just use 'ReloadPage()' before the 'return 1'. I found this
| out when I deleted a message from the OUTBOX, and ended up back at
| the INBOX <;-) fergodsake. Obviously ReloadPage works in both 
| BOXes.
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 6:33:33 pm 
|
| glennmcc>
| You must not have done email yet.
|
| That's what I said in my message to the mailing list earlier today. <gggg>
|
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 6:36:12 pm 
|
| glennmcc>
| Oh I did a little more tweaking today.
|
| I now have 204,128[+] green (up from 190,600)
|
| Did nothing more than change from not overlayed to overlayed on 3 files.
|
| BRB with the list of files.
|
|
Š| ============================================================================
|
| Mon, Jan 17, 2005 - 6:42:12 pm 
|
| glennmcc>
| htkernel.c
| htmlstat.c
| urlstat.c
|
| Overlaying all 3 of them reduces core.exe size in mem by 14kb
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 6:44:59 pm 
|
| glennmcc>
| Make that 13,528 bytes.... not 14kb
|
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 7:01:50 pm 
|
| glennmcc>
| <WTF>
|
| Something screwy is going on :((
|
| http://www.cisnet.com/glennmcc/my-stuff/wtf.gif
| http://www.cisnet.com/glennmcc/my-stuff/wtf-2.gif
| http://www.cisnet.com/glennmcc/my-stuff/wtf-3.gif
|
|
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 7:04:22 pm 
|
| ray>
| Never got that, infact I've had no email at all today, I wonder why?
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 7:12:22 pm 
|
| ray>
| And I can't get on to AD either, says address is bouncing emails.
|
| As far as overlaying modules, I'm looking into that too, there are
| lots of functions that should be static, for sure, since they are
| called constantly, everything in GUITICK for example. But it does
Š| look like some static functions need not be.
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 7:57:56 pm 
|
| ray>
| Glenn, be a sport and send me an email so I can see if I'm not getting
| stuff.
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 7:58:41 pm 
|
| glennmcc>
| Have a look at wtf-3.gif
|
| You got no email today 'cause the 995.ca server seems to be down.
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 7:59:58 pm 
|
| glennmcc>
| That's what I tried to do earlier.
|
| wtf-3.gif is a screen-cap of the email I tried to send you that bounced
| right back at me.
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 8:01:02 pm 
|
| glennmcc>
| I'll try again right now
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 8:02:56 pm 
|
| glennmcc>
| No go.... right back at me again.
|
| Delivery failed
| Unable to deliver to destination domain
| Cannot resolve 995.ca
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 8:03:53 pm 
|
| ray>
Š| As far as overlays, as I mentioned, I want to come up with a
| consistent speed test precicely to test the effect of moving functions
| from static to overlayed modules. For the present, what I do is
| put a Beep() in functions I want to test -- lots of beeps == should be
| static, few beeps == should be overlayed. Do you have some test you
| use?
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 8:04:34 pm 
|
| glennmcc>
| They are completely down.
|
| Can't even access http://www.995.ca/
|
|
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 8:07:52 pm 
|
| ray>
| Ahh, so my ISP is fouled up. I got the POP3 error page when I first
| tried to log on this aft but later it seemed ok, but it looks like
| I lost the days emails. I just sent and recieved an email to myself
| but since you couldn't get through, I guess what that tells me is
| that sending an email to yourself is not a reliable way to say that
| all is well with ones ISP. Thanks.
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 8:36:48 pm 
|
| glennmcc>
| Just now switched GUITICK.C to overlayed.
|
| Now I'm up to 206,480 [+] green
|
| That's a saveings of yet another 2,352 bytes.
|
|
| As for a test of wheather or not overlay.
|
| Right now I'm concentrating on memory stats.
|
| Will 'refine the process' as I go. <ggggggg>
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 8:49:04 pm 
|
| glennmcc>
Š| As for 'many beeps vs few beeps'
|
| IMO, That's not a good test.
|
| All of the various HTML rendering modules would give many, many beeps.
|
| But they are only used while rendering the HTML to screen.
| As soon as the page is fully rendered, they are no longer needed.
|
| Dumping them from memory and only reloading them the next time we render
| a new page (or re-render the current one), saves one hell of a lot of
| memory while we are not rendering.
|
| The same is true for GUITICK.C
|
| Dump that sucker out of memory till we need it again. <g>
|
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 9:10:48 pm 
|
| ray>
| I agree with what you say above vis. HTML -- only used intermitently,
| and that's exactly what the beeps demonstrate. What matters isn't
| the 'dencity' of the beeps. Durring an HTML reload, the speaker
| howls, but durring that time, the module would be overlayed anyway.
| What matters is how often the howls occur. But I disagree about
| GUITICK -- those functions are constantly on duty. But it's fun to
| play with this stuff, and what matters is real world changes in
| performance. In B5 I have rationalized master/slave function 
| families as much as possible and that simplifies things alot since
| it stands to reason that helper functions should be in the same 
| overlay as their master. Have fun with this!
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 9:13:16 pm 
|
| ray>
| It looks like my ISP is realy buggered, interesting that *this* works,
| and I can send mail and recieve mail from myself, but everyone else
| is outaluck trying to mail me.
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 9:47:38 pm 
|
| Joe>
| Hi Ray,
|
| Just sent you and Glenn an e-mail about the AD list
Š| (seems broken - hopefully a short-term problem),
| however the message to you bounced per :
|
| Returned mail
|
| --- The message cannot be delivered to the following address. ---
|
| randrews@995.ca Remote host not found. 
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 9:49:12 pm 
|
| Joe>
| Also Ray, just reading the above, 500 grand ain't enough!
| Anything below 600 grand, I consider unsuitable ...
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 10:16:53 pm 
|
| glennmcc>
| My current mem stats.
|
| 624kb free before starting results in 214,064 right now.
| (did some more tweaking) <g>
|
| So..........
|
| 500kb bytes free before starting Arachne == about 90,000 after loading
|
| Yep, Joe is right.
|
| We need close to 600kb before starting Arachne.
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 11:00:47 pm 
|
| glennmcc>
| More tweaking....... 217,568 [+] green
|
|
| ============================================================================
|
| Mon, Jan 17, 2005 - 11:18:44 pm 
|
| glennmcc>
| Just loaded a bunch of stuff low to take my memory down to 466kb before
| starting Arachne.
|
| Now my Dos Mem is at 84,448[?] black
|
|
Š| ============================================================================
|
| Mon, Jan 17, 2005 - 11:23:40 pm 
|
| glennmcc>
| Now at 535kb before starting ....... 108,368[+] green
|
| This is fine.
|
| IMO,
| Around 550kb should be our min suggested level.
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 4:18:03 pm 
|
| ray>
| www.995.ca still in limbo but still in buisiness. Just have to
| wait it out.
|
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 4:38:54 pm 
|
| Joe>
| Wow! "217,568 [+] green" - that's amazing!
|
| Anyway, just visited the AD list ... it seems to be back
| to normal. :-)
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 5:12:35 pm 
|
| glennmcc>
| Yep.....
|
| Went from 190,656 to 217,568 without making a single change to the code itse
|
| All I did was switch several files from not-overlayed to overlayed.
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 5:37:15 pm 
|
| glennmcc>
| // { RAY: 05-01-16: After deleting message(s), reload the index to prevent
| // 'holes' in it since these holes will create an error message when the ">>
| // button is used to go the next message if that message no longer exists
| // due to its having just been deleted.
| // goto index;
| //!!glennmcc: Jan 17, 2005 -- reload instead of going to the inbox
Š| //!!glennmcc: Jan 18, 2005 -- found a problem....
| //hitting the delete key while viewing a remote page or image to delete
| //that file from the cache then reloads that page or image.
| //therfore we must only reload when it's a local dir listing
| //return reloadpage(); //new line inserted on Jan 17, 2005
| //return 1; //original line
| if(strstr(GLOBAL.location,"file:")) return reloadpage(); //reload if local
| else return 1; //do not reload if not local
| }
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 5:54:20 pm 
|
| ray>
| My ISP seems to be back.
|
| Glenn, this is a first: you are venturing into teritory with 
| gay abandon, where I'm feeling cautious and hesitant ;-)
| There's no harm in fooling around with this stuff of course, but
| I would expect that below a certain size, performance must begin to
| drop off, as Arachne is forced to make do with less and less
| memory.
|
| And thanks for the code improvement above. TTTT, I didn't know
| that DEL had that function.
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 6:17:30 pm 
|
| ray>
| Well, jumping in with both feet like I do, I made every damn thing
| except globals.c overlayed and I've now got 244,192 bytes free.
| And my first impression is that the slowdown is marginal ... seems
| too good to be true.
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 6:18:45 pm 
|
| ray>
| Gotta make a speed testing .asf and get some numbers on this.
|
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 6:20:51 pm 
|
| glennmcc>
| That's exactly what I thought when I got her up to over 210,000
|
| It's like.. "Shit man, I don't see any problems".
Š|
| So I tried a few more files overlayed..... still no problems.
|
| BBL
|
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 6:27:06 pm 
|
| ray>
| testing
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 6:28:14 pm 
|
| ray>
| Too good to be true: I crashed trying to use the scrollbar.
| GUITICK back to static gives 238,960 free, but problem still there.
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 6:29:21 pm 
|
| ray>
| Or maybe not ...
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 6:30:39 pm 
|
| ray>
| But this is your project, I'll retreat back to stock and let you
| figure this all out. Man, MP would be impressed.
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 6:36:57 pm 
|
| ray>
| Watch out for "runtime overlay error" using the F5 key!
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 6:58:01 pm 
|
| glennmcc>
| Yep, I'll keep my eues open for that and other probles as-well.
|
| I now have most of the stuff ovetlayed.
Š|
| Currently at 225,184 [+] green
|
| I'm going to run this one through thae paces for the rest of this week
| and I'll let you know what happens.
|
| Gonna turn everything on for most of my testing.
|
| CSS on
| images on
| animated GIFs at 1meg/4meg
|
| I'll run it though all screen resolutions and screen layouts (all F5 options
|
| In other words...... I'm gonna beat the hell out of it. <ggggggggg>
|
|
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 7:03:42 pm 
|
| glennmcc>
| Conventional_memory:
| _Name________________Size_in_Decimal_______Size_in_Hex
| ______DOS_________91,280_(________89K_)__________16490
| _QEMM386$____________768_(_________0K_)____________300
| __COMMAND__________1,936_(_________1K_)____________790
| _____CORE________419,200_(_______409K_)__________66580
| __command__________7,200_(_________7K_)___________1C20
| _____FREE________216,528_(_______211K_)__________34DD0
|
| _Memory_Type__Total_Bytes_(_Kbytes__)__Available_For_Programs
| _Extended______________67,043,328_(_65,472K_)_______________0_(______0K_)
| _EMS__________________267,517,952_(_261,248K_)_____157,761,536_(_154,064K_)
|
| _Largest_executable_program:__216,512_(_211K_)
| _Total_Free_DOS_memory:_______294,464_(_288K_)
|
| While running memcore.dgi, she's using 8,688 additional byes compared to
| sitting here idle after rendering a page.
|
| So core.exe is using 410,512 right now.
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 9:02:48 pm 
|
| ray>
| The magic 400,000 is within sight. I'll let you beat up on this
| till your satisfied with it, and then I should have some sort of
| speed test ready, then we can make whatever tradeoffs are needed.
| Given that 200K of free memory is enough to run just about anything
| we'll need under Arachne, further memory improvements are of
Š| questionable use if there is much of a speed penalty, OTOH, like
| you, I haven't noticed much of a difference so far.
|
|
| ============================================================================
|
| Tue, Jan 18, 2005 - 11:35:26 pm 
|
| glennmcc>
| Yep, same here.
|
| So-far, not one hicup, not one cough, not one sneeze, not one problem of any
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 7:28:48 am 
|
| ray>
| We'll see how far you can go. You might try to duplicate my
| problem above -- leave only globals.c and guitick.c static. However,
| this might not work if the problem is related to computer speed or
| something like that. Which makes me think, what if all this stuff
| is OK on your fast machine, but will crash a 486? Are there any 
| static modules that might have this problem?
|
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 3:51:07 pm 
|
| glennmcc>
| None that I can think of right off-hand.
|
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 3:55:22 pm 
| URL
|
| glennmcc>
| Now as for some further testing of this new compile.
|
| Just uploaded a Cntrl+P full page capture of my nieces' modeling portfolio.
| (converted the original .BMP to GIF..... carefull, it's HUGE..... 806 x 5337
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 5:19:23 pm 
|
| ray>
| Pretty girl.
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 6:30:10 pm 
Š|
| ray>
| I did some hunting and have an updated list of 'arachne:???'
| Which should update 'aexthtml.txt':
|
|
| Protocol "arachne:"
| -------------------
| (Ray: Those marked "U" can be used as URL's)
|
| arachne:back ...goto previous document (same as "Previous document" icon)
| U arachne:dialer ...execute "Connection" ARACHNE.CFG variable
| U arachne:dialpage ...goto URL defined in "DialPage" ARACHNE.CFG variable
| arachne:exit[?errorlevel] ...exit to DOS
| arachne:ftp-send ...ftp
| arachne:ftp-upload ...ftp
| U arachne:hangup ...execute "Hangup" ARACHNE.CFG variable
| arachne:internal-config[?newURL] ...update ARACHNE.CFG
| arachne:internal-editor ...for text editor
| arachne:internal-htmledit ...edit current page
| arachne:internal-mail ...for sending mail messages
| arachne:save ...save document
|
| (In addition to the above, these have been found by Ray in the sources and 
| in the Arachne HTML pages (as of 05-01-19).
|
| U arachne:addressbook
| arachne:again
| arachne:authenticate
| arachne:copy
| U arachne:ftp-login
| arachne:fullscreen
| arachne:history
| arachne:internal
| arachne:internal-config?arachne:back
| arachne:internal-config?arachne:ftp-upload
| arachne:internal-config?arachne:restart
| arachne:internal-config?file:%s%sopt_misc.ah#scr
| arachne:internal-config?reload:
| arachne:internal-vga
| arachne:kill-cache
| U arachne:pppsetup
| U arachne:restart
| arachne:view
|
|
|
|
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 6:35:06 pm 
|
| ray>
| How, how does one start arachne with an .ASF?
Š|
| I made 'test.asf' containing one line "arachne:restart"
| and then did "DialPage file://test.asf" and all hell broke loose:
| I got MALLOCERR's and strange restarts, and even after going into
| arachne.cfg and erasing all trace of test.asf, she insisted on
| reloading it, but showed it as an open text file, but crashed anyway.
| I couldn't kill the damned thing and had to overwrite it with
| garbage to make it go away. Bloody scary.
|
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 6:47:22 pm 
|
| glennmcc>
| Like this........
|
| arachne.bat file://pop_all.asf
|
| --- pop_all.asf ---
| pop3://username:password@mail.SomeWhere.com:110/
| pop3://username:password@pop3.SomeWhereElse.com:110/
| reload:file://inbox.dgi
| ________________
|
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 6:50:44 pm 
|
| glennmcc>
| And of-coarse all hell broke loose when you have arachne:restart in an ASF
|
| That you ran at startup.
|
| startup .... run ASF which restarts
| restart .... run ASF which restarts
| restart .... run ASF which restarts
| restart .... run ASF which restarts
| restart .... run ASF which restarts
| restart .... run ASF which restarts
| restart .... run ASF which restarts
| restart .... run ASF which restarts
| restart .... run ASF which restarts
| restart .... run ASF which restarts
| restart .... run ASF which restarts
| And so-on .................forever.
|
|
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 7:11:44 pm 
| URL
|
Š| glennmcc>
| re: aexthtml.txt
|
| Done :))
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 9:14:59 pm 
|
| ray>
| Sure, I understand why my little .asf was asking for trouble, hell
| I don't expect Arachne to be able to such sillyness, but what
| caught me by surprise was the fact that I couldn't get rid of the
| damned thing -- but again, Arachne can hardly be expected to 
| cope with stuff like that.
|
|
| ============================================================================
|
| Wed, Jan 19, 2005 - 9:16:54 pm 
|
| ray>
| Oh forgot to mention: I'm experimenting with :restart, 'cause it
| might be of use in creating a speed tester, but we'll see.
|
|
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 12:34:17 am 
|
| Joe>
| Hi all,
|
| With your overlay experiments ... don't forget to try AltTab ...
| I'm sure _that_ stuff mustn't be overlaid.
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 2:36:44 pm 
|
| glennmcc>
| Done it hundreds of times since starting my tests.
|
| No problems yet. :)
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 3:52:51 pm 
|
| ray>
| Did you try my extreme? (all but globals.c and guitick.c overlayed?)
|
| Hey Joe, what do you think of all this? There seems to be little
| if any performance penalty going to overlays from static; it seems
Š| too good to be true, but true none the less. Thoughts?
|
|
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 4:29:32 pm 
|
| glennmcc>
| Here's the list of .C files _not_ overlayed in this compile.
|
| version.c
| globals.c
| html.c
| int09.c
|
| All others are overlayed.... including guitick.c
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 4:33:39 pm 
|
| glennmcc>
| I've just now recompiled with version.c and int09.c overlayed as-well.
|
| Mem stats are now up from 225,184 to 225,520
|
| I'll use this one as my 'extreme' for a few days.
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 5:35:39 pm 
|
| glennmcc>
| Been doing most of my tests running 100% on RamDrive.
|
| Now I'll switch to 100% HDD (including the cache and temp dirs)
|
| That ought to really be a good test of performance differences
| between opverlayed vs non-overlayed.
|
| Might even turn off the diskcache as an 'extreme test'. <ggg>
|
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 5:45:44 pm 
|
| Joe>
| Yeah, int09.c is the AltTab stuff that mustn't be overlayed (don't
| forget the "AltTab Yes" option in 'arachne.cfg').
|
| It's rather interesting that you've seen no performance penalty
| with this stuff :-) ... I wonder (a) is this still the case with
| old 486-level hardware and (b) are you using a RAM disk (I expect
Š| this will negate a lot of the performance hit from overlaying)?
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 7:39:20 pm 
|
| glennmcc>
| That was it... that was the method needed.
|
| 100% on HDD with no diskcache.
|
| I have now tested 50 new compiles in the last 3hrs.
|
| There are 2 modules which can be changed to overlayed with no drop in perfor
|
| version.c
| putikonw.c
|
| Overlaying those 2 saves about 7,000 bytes.
|
| Conclusion........ leave everything the way it was.
|
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 8:12:12 pm 
|
| glennmcc>
| However, since a diskcache is a _must_ anyways when running on HDD.
|
| We can go right ahead and overlay almost everything and we do not lose
| any performance what-so-ever.
|
| With the dickcache working, core.exe is held in memory so we are not
| reading the overlayed portions from disk into memory when needed.
|
| We are reading them from one section of memory to anther section of
| memory. The same as when we are running 100% on RamDrive.
|
| Thoughts ???
|
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 8:36:57 pm 
|
| glennmcc>
| I had 2 crashes with Alt+Tab
|
| int09.c is back to static.
|
| However, I changed 1 setting over on Options/Prefs. & Perfs.
|
| X Disable caching of fonts in XMS 
|
Š| My mem stats now read....... 229,664 [+] green
|
|
|
|
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 9:07:50 pm 
| URL
|
| glennmcc>
| Another screen-cap
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 9:58:06 pm 
|
| ray>
| I like your explanation: disk cache is the key to the lack of
| performance drop.
|
| BTW, remember that nice improvement you made to wwwman? Can you
| pull the same trick with insight? I crashed going to my \trash
| to retrieve something with a MALLOCERR from the same place in the code
| that it used to happen before you fixed wwwman, but I can
| access the directory fine using the URL line or "Desktop". It would
| be nice to have the same file limit for both.
|
|
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 10:04:12 pm 
|
| glennmcc>
| Good Idea.
|
| But, I gotta get up at 6:00am for work.
| So it's time to call it a night and hit the sack.
|
| I'll do it as soon as I get home from work. :))
|
|
| ============================================================================
|
| Thu, Jan 20, 2005 - 10:04:28 pm 
|
| ray>
| OK, so what are your final remaining statics? I'll try the same
| and see if I'm stable.
|
|
| ============================================================================
|
Š| Thu, Jan 20, 2005 - 10:36:45 pm 
|
| ray>
| Enjoy sleeping in, I get up at 5:30 < g g g g g g >
|
| Nite-nite.
|
|
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 2:07:40 am 
|
| gregy>
| ahhh...you oiley-boids...!!!!.....<g g g>
| ..
| Just converted2txt and clipped Glenn's "no-auto-archive" code, above,
| and will try that out.
| ...
| FWIW...I've been having problems with trying to FTPupload using Glenn's
| .BAT.....
| ...
| PROBLEM: trying to upload to (or even access) ftp.hi-line.net/~gfeig/
| keeps giving me the "no such directory or file" error from my server.
| ...
| The only way I can access my site, from the back(ftp)side is to put
| the whole URL:
| ftp://myuname:mypassword@ftp.hi-line.net/sub1/sub2/.../subn/uname/pub_html
|
| I can get into ftp.hi-line.net just fine using myuname:mypassword@
| combination, BUT...the server will NOT let me nagigate down thru the
| subdirectories......it stops returning DIRlistings at about the first
| subdirectory. So, I can ONLY access my site by knowing my complete
| directory structur/location to begin with.
|
| To get that, I use WS_FTP Lite, which, given only myuname and mypassword
| goes right to my site down thru the subDIR structure. NOW, WS_FTP does
| NOT know this subdir structure beforehand....I have never entered that
| into it....and that info is found NOWHERE in the WS_FTP files....and
| is NOT, just NOT, in the WS_FTP at all......therefore, in order for
| WS_FTP to be listing that info (i.e. my complete subdir
| structure/location) on its status/location bar, it MUST be getting that
| info returned to it FROM MY SERVER when it correctly logs in. 
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 2:10:08 am 
|
| gregy>
| So, that must be info my ISP's server returns after a correct login.
| ...
| QUESTION: Why isn't Arachne picking this up...???
|
|
| ============================================================================
Š|
| Fri, Jan 21, 2005 - 7:31:58 am 
|
| ray>
| 1 testing
| 2 testing
| 2
| 3
|
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 3:42:56 pm 
|
| ray>
| Gregy, IMHO Arachne's FTP is majorly buggy and needs a total overhaul.
| If I new the first thing about protocols I'd look into it, but
| my ignorance is total. I'd like to get the Wiz to do it, since he
| is already the author of an FTP program.
|
|
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 4:37:15 pm 
|
| glennmcc>
| 1st off
|
| When using my .BAT, you are not using Arachne.
|
| You are using Cute-FTP
|
| _______________
|
| 2nd off <g>
|
| In my .SCR file, did you change the dir structure on line 3 ?
|
| --- snipped from my .SCR -------
| user your_username
| your_password
| cd web/aqc/archives <---- change to your FTP dir structure
| lcd arch <------- change to your dir on your HDD
| nonint
| bin
| hash
| mput *.txt
| quote site chmod 666 posts025.txt <--- change filenames to match yours
| quote site chmod 666 posts026.txt
| quote site chmod 666 posts027.txt
| quote site chmod 666 posts028.txt
| quote site chmod 666 posts029.txt
| quote site chmod 666 posts030.txt
| quote site chmod 666 posts031.txt
Š| quote site chmod 666 posts032.txt
| quote site chmod 666 posts033.txt
| quote site chmod 666 posts034.txt
| quote site chmod 666 posts035.txt
| quote site chmod 666 posts036.txt
| quote site chmod 666 count025.txt
| quote site chmod 666 count026.txt
| quote site chmod 666 count027.txt
| quote site chmod 666 count028.txt
| quote site chmod 666 count029.txt
| quote site chmod 666 count030.txt
| quote site chmod 666 count031.txt
| quote site chmod 666 count032.txt
| quote site chmod 666 count033.txt
| quote site chmod 666 count034.txt
| quote site chmod 666 count035.txt
| quote site chmod 666 count036.txt
| lcd ..
| quit
|
| -------------------------------
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 6:47:53 pm 
|
| ray>
| Nope I didn't change nuthin there. Anyway you make my point 
| for me, Arachne's internal FTP is in poor shape, thus the need to
| use cuteFTP. Maybe one day I'll be experienced enough to figure
| out how all this stuff works, but that's a long way off.
|
|
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 6:49:36 pm 
|
| ray>
| ... or, we could just scrap the internal ftp and rely on cute-ftp
| as standard ...
|
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 7:38:38 pm 
|
| glennmcc>
| Sorry Ray... that was aimed at Gregy, not you.
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 7:46:08 pm 
|
| glennmcc>
Š| Now as for Insight.
|
| No changes need to be made to insight.exe after-all.
|
| It has no problem with way past 2000 massages in a dir.
|
| The problem is that Arachne can't display that many graphics files
| when we have her setup to use tables for the mail indexs.
|
| There are 2 options for viewwing very large numbers of mail messages in a di
|
| 1) over on Options/Pref. & Perf..... turn off tables for mail indexs.
|
| 2) edit idxhead.htp like so.......
|
|
| #idxrow
| <TR><TD>$msgcnt:03$.
| <TD><B>$fromto$:</B> $msgfromto$
| <TD><B>Date:</B> $msgdate$
| <TD ALIGN=RIGHT>$msgsize$ kB
|
| <TR>
|
| <!-- COMMENT-OUT THIS SECTION ........
| <TD>
| <A HREF="reload:file:$filename$" REMOVABLE><IMG SRC=pidimail.ikn ALIGN=LEFT 
| BORDER=0></A>
| -->
|
| <TD COLSPAN=4><!-- THIS WAS AT THE END OF THE LINE ABOVE -->
|
| <A HREF="reload:file:$filename$" REMOVABLE>$msgsubj$</A>
| <P>
|
| ___________
|
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 8:06:59 pm 
|
| glennmcc>
| Option 1)
| http://www.cisnet.com/glennmcc/my-stuff/no-table.gif
|
| Option 2)
| http://www.cisnet.com/glennmcc/my-stuff/no-image.gif
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 10:14:12 pm 
|
| ray>
| I'll try 'em both, 'course I don't realy know what tables even is,
Š| so caint say if I'd miss 'em if they wuz gone. As for the
| edit of the other file, on first glance, that looks like a keeper.
|
| Oh don't forget to let me know which files you've left static, I want
| to see if I'm stable with the same settings; with all my function
| moves, I might not be. I'm now guessing that only your 'int09'
| might need to be static, I merged that file with 'gui' so that's
| probably my minimum static. Anyway, to stop for a bit of shameless
| self congratulation, we've now added about 100K of free memory since
| we started work -- not too bad!
|
|
| ============================================================================
|
| Fri, Jan 21, 2005 - 10:55:27 pm 
|
| ray>
| No luck with the edited file, same crash, and I have the same problem
| with 1.79, so something is amiss. But switching to the alternate mail
| index head works fine. If the problem is the graphic, would it 
| be possible to do something similar to that clever thing you did
| in wwwman, where by it knows automagicaly not to load the graphic
| if over a certain number of files (I'm guessing 512)?
|
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 12:00:53 am 
|
| glennmcc>
| I was afraid of that.
|
| It's 'touchy' on whether or not you can handle that huage number of
| table cells even with the graphic removed.
|
| So... your only option is to switch-off the tables.
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 12:05:12 am 
|
| glennmcc>
| Static files
|
| version.c <--- it turns out that it eats-up 32 bytes when overlayed
| globals.c
| html.c
| int09.c
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 9:42:12 am 
|
Š| ray>
| OK, I'll try 
|
| globals.c (I don't have version.c -- but that's about to change ;-)
| gui.c (int09.c built in to this)
| guitick.c (I figure this should be static anyway)
|
| Free mem == 237,440
|
| BTW, I read html.c as being previously overlayed. Can you confirm?
| If so, why now make it static?
|
|
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 2:57:46 pm 
|
| glennmcc>
| Yes,
|
| Originally html.c was overlayed
|
| But back when you split it into 3 seperate files.
|
| My method was to simply stop overlaying it.
|
| Since you have it split into 3 parts.... overlay all 3.
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 3:39:46 pm 
|
| ray>
| Bbbbut that would just increase your memory useage, no? If you've got
| the old pig html.c static and we have near the same memory usage
| that's extraordinary, ain't it? I be confused. 
|
| BTW, I think I've made a good change in guievent(). Search for
| "-32512" and note the disgusting way that an overflowing keystroke
| is tested for. I've made the argument to the function unsigned int
| so it's possible to test for "0x8100" as God intended. I don't 
| think there's any time a negative number is ever used as an argument
| to the function, so it should be OK, also it will alow us to
| harness more extended keys like Alt+F??.
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 4:40:59 pm 
|
| glennmcc>
| Yes... that is what is so strange about it.
|
| If you'll remeber.... you were splitting html.c into 3 parts and I
Š| simply switched it from overlayed to static.
|
| You save about 20kb of mem..... I saved about 20kb of mem.
|
| Very strange...... but very true. :))
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 5:01:53 pm 
|
| glennmcc>
| As for -32512
|
| That _is_ the correct test for Alt+0 
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 7:37:47 pm 
| URL
|
| glennmcc>
| Hey dude,
|
| Go back over into prefs & pers. and turn tables back on for your mail indexe
|
| Now try listing that huge-ass directory with this. ;-)
| http://www.cisnet.com/glennmcc/my-stuff/insit387.zip
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 8:25:55 pm 
|
| ray>
| Ok will try it but first:
|
| That's radical about you saving memory by going static, it throws
| a wrench into my understanding of how things work. I'll back up
| to earlier code and play with it, hell if you can leave the file
| intact, AND make it static, AND save memory then God is on your side ;-)
|
| Heads up: Possible bug in 179: I crash with 'F1', and my code was
| doing the same. Traced it to main.c. Code comming if you have the
| problem too.
|
|
|
|
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 8:31:03 pm 
|
| ray>
Š| Oh, forgot: -32512 is one lousy test for an extended character.
| What self respecting programmer will allow a variable to overflow
| and then catch the slop? I made it unsigned and now test for 0x8100
| like a decent fella. Besides 'int' limits our legal characters
| a whole lot, so unless there's something I don't forsee (quite possible)
| I'm going to 'unsigned int'.
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 8:39:11 pm 
|
| ray>
| She Kicks Ass :-))))
|
|
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 8:40:26 pm 
|
| glennmcc>
| Nope... no prob with F1 here.
|
| _____
|
| Nothing is overflowng.... no 'slop' is being caught.
|
| -32512 _is_ the "bioskey output" from the keyboard.
|
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 8:43:39 pm 
| URL
|
| glennmcc>
| Grab a copy of bioskey
|
| Run bioskey.exe at the DOS prompt
|
| Now start hitting keys to see what the "bioskey output" is.
|
| Cntrl+Pause to quit out of boiskey.exe
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 9:13:47 pm 
|
| ray>
| try F1 *directly* after F8.
|
| bioskey, eh? BRB
Š|
| ============================================================================
|
| Sat, Jan 22, 2005 - 9:21:40 pm 
|
| glennmcc>
| F8, F1 ..... no problems.
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 9:25:31 pm 
|
| ray>
| Hmmm, both my code and 1.79 are doing it here. If I load any page at
| all directly after F8 then F1 is fine. 
|
| I can't find bioskey.exe at Google, it is swamped by thousands
| of pages on the bioskey function. Spare me a copy?
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 9:25:46 pm 
|
| glennmcc>
| bioskey.exe is a ver complex program
| created with this this HUGE Borland C/C++ SRC file.
|
|
| --- bioskey.c ---
| #include <stdio.h>
| #include <bios.h>
|
| void main (void)
| {
| int key;
| do
| {
| while (bioskey(1) == 0);
| key=bioskey(0);
| printf("%d\n",key);
| }
| while((key & 0xff) != 27);
| }
| _________________
|
|
| DAMN !!!!!!!!
|
| That's some complex code.
|
| <ggggggggggggg>
|
| ============================================================================
Š|
| Sat, Jan 22, 2005 - 9:27:52 pm 
|
| glennmcc>
| You can't find bioskey ?
|
|
| Did you try hitting that link just above this ????????????
|
| URL
| glennmcc>
| Grab a copy of bioskey
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 10:01:17 pm 
|
| ray>
| Sheesh, bloatware!
|
| And it looks like my problem above was something in 'crl.bat' which
| I had left in it's Greg M. format. I haven't found the offending
| line, but it's strange. When I fill up the file with pauses to
| see what's happening I get very poor screen redraws and crashes
| all over the place, that hideous line about deleting the history
| file shows up as 'bad syntax' so for now I've just REM'ed it out ;-0
| and all seems well. I makes some sence that an incorrectly emptied
| cache could cause all sorts of mischif, so that's probably what
| happens, sorry about the false alarm.
|
|
| ============================================================================
|
| Sat, Jan 22, 2005 - 10:12:33 pm 
|
| ray>
| A - HA! but this here bioskey just makes the same mistake, by not
| having the range to capture all extended key codes. There is no such
| thing as a negative keycode. Extended codes consist of two bytes, and
| the high byte can go as high as 255 whereas the low byte is usualy just
| 00, so in hex we must be able to get as high as 0xff00 in theory but
| in practice I think the highest keycode is 0xa600 (Alt + Enter) anyway,
| signed int only goes up to 0x8000 so any code above that overflows 
| into the negative range of a signed int. Thus the proper code for
| Alt + 0 is 0x8100 which overflows by 0x100 which gives -0x7f00 which
| is dec -32512 but who wants it?
|
|
| ============================================================================
|
| Sun, Jan 23, 2005 - 6:51:48 am 
|
| phantom>
| I agree with Ray about the key code thing ... however, just make
Š| sure that a -1 value isn't used somewhere, it often is in C code,
| and Arachne is no exception.
|
| Regarding statics, note that a couple of files are called "*stat.?"
| and "stat*.?" ... which means MP intended these as static (in two
| cases, some code modules were deliberately split so that selective
| bits could me made static, with the rest overlaid).
|
|
|
| ============================================================================
|
| Sun, Jan 23, 2005 - 9:12:17 am 
|
| ray>
| Joe,
|
| Quite right about the lurking '-1' I've checked and found none, but
| there could be a rude surprise in store.
|
| RE statics, this is the interesting thing about Glenn's experiments.
| There's no doubt as to Michael's intentions with some files and you're
| right that he went to some trouble over it. When I was localizing
| functions, there were cases where a helper function was static even
| though the calling function was overlayed and V.V. However, 
| preliminary results seem to indicate that Michael's structure was
| based on a no disk cache environment -- with a cache there seems to
| be little change in performance with most things overlayed.
|
| Then there is the mystery of Glenns memory improvement by going static
| with HTML.C which baffles me.
|
|
| ============================================================================
|
| Sun, Jan 23, 2005 - 1:58:04 pm 
|
| glennmcc>
| You ain't alone in that 'bafflement'. <gggggg>
|
| The differemnce is now less than it was originally.
|
| Before overlaying almost everything, making html.c static improved
| memory by 20kb
|
| Now that most all of the other files are overlayed, changing html.c
| back to overlayed only increases memory usage by 10kb
|
| But that 'bafflement' still remains.
|
| Why does html.c use 10kb less when static ????
|
| And why does version.c use 32 bytes more when overlayed ???
|
|
Š|
| ============================================================================
|
| Sun, Jan 23, 2005 - 2:14:37 pm 
|
| ray>
| I have a theory that's hard to explain, but try to figure out what I'm
| getting at: .exe size in memory == all global data + all static 
| modules + the LARGEST overlay, right? In the original build, HTML.C is
| three times bigger than any other overlay, so it 'defines' the size
| of the .exe, which is why when I chopped it up, the .exe size dropped.
| However, it may be that static modules are naturaly more efficient --
| since they live in memory all the time, the conections between static
| modules are 'always open' so to speak ...
|
|
| ============================================================================
|
| Sun, Jan 23, 2005 - 2:15:57 pm 
|
| glennmcc>
| Al right gentlemen.... I have just mad a 'smarter' BIOKEY.EXE
|
| -------- bioskey.c --------
| #include <stdio.h>
| #include <bios.h>
| void main (void)
| {
| unsigned long int key;
| do
| {
| while (bioskey(1) == 0);
| key=bioskey(0);
| printf("Use Cntl+Pause to quit this program\n");
| printf("Dec -- %u\n",key);
| printf("Hex -- 0x%X\n",key);
| printf("Oct -- %o\n",key);
| }
| while((key & 0xff) != 27);
| }
| ___________________________
|
|
|
| ============================================================================
|
| Sun, Jan 23, 2005 - 2:18:24 pm 
|
| ray>
| ... Thus, although there was a cost to making HTML static, there was
| also a benifit in that the code is more efficient. The more modules
| are static the greater the benefit, thus when there are lots of static
| modules, making HTML static is a big help, but when most of them
| are overlayed, the advantage is less, see?
|
Š| 'Course I'm just making this up, but I might have the basic Idea right.
|
|
| ============================================================================
|
| Sun, Jan 23, 2005 - 2:20:44 pm 
|
| ray>
| Grabbed your smarter code, will test in a day or two, right now,
| off to the island for some fun.
|
| Be good! ;-)
|
|
|
|
| ============================================================================
|
| Sun, Jan 23, 2005 - 2:39:36 pm 
|
| glennmcc>
| I've just made the needed changes in guievent.c and in gui.h
|
| GUIEVENT() is now unsigned int
|
| ------ gui.h ---
| //!!glennmcc: Jan 23, 2005 -- at Ray's suggestion, unsigned int
| unsigned int GUIEVENT(unsigned int key, unsigned int mys);
| //int GUIEVENT(int key, int mys);
| ________________
|
|
| ---- guievent.c ----
| //!!glennmcc: Jan 23, 2005 -- at Ray's suggestion, unsigned int
| unsigned int GUIEVENT(unsigned int key, unsigned int mouse)
| //int GUIEVENT(int key, int mouse)
| <snip>
|
| //!!glennmcc: no more "negative tests" ;-)
| else if(asc=='H' || key==0x2b1c || key==0x8100) //Alt+0,Ctrl+backslash
| // else if(asc=='H' || key==0x2b1c || key==-32512) //Alt+0,Ctrl+backslash
| ____________________
|
|
| Are there any other negative tests we need to change ?
|
|
| ============================================================================
|
| Mon, Jan 24, 2005 - 2:53:29 pm 
|
| ray>
| I think that's the only negative test, gota run now, but I typecast
| all the arguments to guievent throughout to make sure they are all 
| unsigned. I'll list them when I get back.
Š|
|
|
| ============================================================================
|
| Thu, Jan 27, 2005 - 8:19:47 am 
| URL
|
| glennmcc>
| Been a busy beaver....... fixed another bug. :)))
|
|
| ============================================================================
|
| Thu, Jan 27, 2005 - 3:49:34 pm 
|
| Joe>
| Good stuff - sqwish 'em, sqwish 'em good! <G>
|
| Been so busy lately, hope to release my code changes soon,
| just gotta find a few hours to finish things off ... :-/
|
| ============================================================================
|
| Thu, Jan 27, 2005 - 4:31:30 pm 
| URL
|
| glennmcc>
| Yep... gonna keep on sqwishin em. <g>
|
| Uploaded a simple little test page.
|
| ============================================================================
|
| Thu, Jan 27, 2005 - 8:21:16 pm 
|
| ray>
| Sheesh, I'm gone for two days and I can't keep up! ;-)
|
|
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 12:06:27 am 
|
| glennmcc>
| You try the test page ?
|
| Pretty messed-up the way the change in colors of the links on help.htm
| get placed over to the left instead of on top of the existing link.
|
| Now try it with .... http://www.cisnet.com/my-stuff/a181beta.zip
|
| Better ???? <gggggg>
|
Š| ============================================================================
|
| Fri, Jan 28, 2005 - 12:09:12 am 
| URL
|
| glennmcc>
| Duh.... try this link... it will work better. <gg>
|
| http://www.cisnet.com/glennmcc/my-stuff/a181beta.zip
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 9:33:14 am 
|
| ray>
| AYIEEE, the startreck logo ... he's one of THEM! ;-)
|
|
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 5:29:10 pm 
|
| glennmcc>
| Hell, I thought you already knew. <g>
|
| I've been a Trekie for over 30 years.
|
| Got all 3 years of Star Trek original series (including the never aired
| pilot episode "The Cage"), on VHS 
|
| Got all 7 years of the Star Trek Next Generation on VHS
|
| Got all 10 movies (some on VHS, some on DVD, some on both)
|
| Got all 7 years of Star Trek Deep Space Nine on DVD
|
| Hell, I can recite the scripts of most episodes 'from memory'. :))
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 5:30:03 pm 
|
| glennmcc>
| So...... did I sqwish that bug or what ? <gggggg>
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 6:21:27 pm 
|
| ray>
| TTTT I'm a moderate treckie myself, I loved the original series --
| most of 'em at any rate, but I found the movies quite flat and the
| spinoff series disapointed me, they seemed to be just milking the
| franchise as far as I could see.
Š|
| Errr, what bug? I'm falling behind ;-/ ... but I must say you've
| become something of a Superman lattely with all these fixes. I'm
| thinking I'm just going to have to wait for your next release and
| do my catching up then.
|
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 6:25:19 pm 
|
| ray>
| BTW what I am working on is the goback problem, and I just can't get
| a handle on it. Sometimes I click on a message, reply, send and then
| get sent back to the originating message with no problem and sometimes
| I get dumped at the 'delmail.dgi' screen, which seems the same as the
| mailto: screen AFAICT. I have no idea why. And I can't make tests
| for the problem 'cause when I send myself a message to reply to for
| testing purposes, they *always* return me to the originating message --
| it's only with 'real' messages that I get the problem.
|
| Help me Obiwan.
|
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 7:34:34 pm 
|
| ray>
| BTW, do you find that "F7" will only find the first occurance of 
| something you're seaching for if you are inside an "F4" screen?
| I'm just noticing that now with my core and with 1.79.
|
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 8:00:33 pm 
|
| glennmcc>
| F7 finds the 1st occurence.
|
| Now Cntl+L finds the next, and the next, and the next, etc...etc...
|
| As for 'goback' in email messages.
|
| That's one of the reasons why my new cores do not put them in the history.
|
| Now you can't go back to them cause they ain't in the list of files we
| can go back to.
|
| Problem solved. :))
|
|
|
|
Š| ============================================================================
|
| Fri, Jan 28, 2005 - 8:07:35 pm 
|
| glennmcc>
| Here's my history file after doing 48 emails in the inbox.
|
| file:\arachne\hotlist.htm
| file:c:\html\MYPAGES\hotlist0.htm
| file:c:\html\MYPAGES\hotlist1.htm
| http://www.hi-line.net/~gfeig/adev/advindex.php
| http://www.hi-line.net/~gfeig/adev/advindex.php#end
| file:l:\arachne\mail.htm
| file:\arachne\hotlist.htm
| file:c:\html\MYPAGES\hotlist0.htm
| file:c:\html\MYPAGES\hotlist1.htm
| file://inbox.dgi
| file:\arachne\hotlist.htm
| file://reply.dgi
| file:c:\html\MYPAGES\hotlist9.htm
| mailto:
| file:c:\html\MYPAGES\hotlist9.htm
|
| Now you tell me.
|
| Ain't that much better now ?
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 8:10:15 pm 
|
| glennmcc>
| As far as 'catching up' goes.
|
| If you'd like, I can send you all of my changes
| since the release of v1.79 on Christmas day.
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 8:16:48 pm 
|
| glennmcc>
| Here's the list of changed files.
|
| Directory of E:\ARACHNE\ASRC\ASRC181\CHANGES
|
| ENTITY.C______ 3330 _1-04-05 _5:43p
| MISC.C_______ 11313 _1-09-05 _7:58p
| INIT.C_______ 16546 _1-09-05 _8:06p
| MAIN.C_______ 36799 _1-16-05 _6:14p
| GUI.H_________ 3174 _1-23-05 _4:33p
| ARACHNE.H____ 22603 _1-24-05 _9:07p
| CONFIG.C_____ 17900 _1-24-05 _9:14p
| HTML.C______ 100939 _1-24-05 _9:17p
| URLSTAT.C____ 16519 _1-24-05 _9:18p
Š| GUIEVENT.C___ 45433 _1-26-05 _1:18p
| GUITICK.C_____ 7777 _1-26-05 _3:57p
| PRINT.C_______ 5523 _1-26-05 _6:01p
| ONMOUSE.C____ 21174 _1-27-05 12:58a
|
| 13 File(s)
|
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 9:29:43 pm 
|
| ray>
| Ctrl+L you say? But in most screens that are searchable, I can use
| F7 repetedly with no problems -- it just keeps looking for the next
| ocurance of the string, why the difference with 'F4' screens?
|
| As to the exclusions from the history list, I'm cautious about this
| but I think there could be some good in it. I didn't like adding
| 'pop3' since it prevented "DialPage file://pop3:" from working, however
| there might be a workaround for that. As to the problem I mention 
| above, I don't doubt that you've cured it, but it's like cutting of
| someone's leg to cure his flat feet < g g g > I find the goback-to-
| originating-message thing to be very convinient and I'd like to keep it
| but have it work reliably. At least I'd like to know why in hell it
| *isn't* reliable, even if your solution is the ultimate answer. 
| Near as I can figure, messages are only added to the history list 
| sometimes -- why only sometimes ??? This kind of thing bugs me.
|
|
| ============================================================================
|
| Fri, Jan 28, 2005 - 9:33:32 pm 
|
| ray>
| Oh, forgot: as to you changed files, I'lljust wait till your ready to
| upload, you sure are on a tear right now -- one man tornado. Oh and
| when you are ready to upload could we/I have a full download? Somewhere
| in the process of integrating 1.79 I managed to scramble something up,
| so a complete download would be in order around here.
|
|
| ============================================================================
|
| Sat, Jan 29, 2005 - 2:39:02 pm 
|
| glennmcc>
| By 'complete download', do you mean all of the files in 
|
| e:\arachne\asrc\asrc181\
|
| As opposed to only those 13 listed above which have been changed.
|
|
| ============================================================================
Š|
| Sat, Jan 29, 2005 - 3:30:24 pm 
|
| ray>
| Yup, I'd like to have the works, that way I'll be sure I've got
| everything right going back several releases since as I said I managed
| to foul something up between 1.77 and 1.79. (I have the .exe from you,
| but I couldn't build it from the sources.)
|
|
| ============================================================================
|
| Sat, Jan 29, 2005 - 4:23:54 pm 
|
| glennmcc>
| Ray,
|
| In your eyes, my new system of not writing .CNM .TBS .MES and .SNT
| into the history was a bug.
|
| Therefore.... I have just now finished sqwishing that bug too. <gggggg>
|
| void add2history(char *URL)
| {
| //!!glennmcc: Jan 16, 2005 -- also don't add any of the mail files themselve
| //!!glennmcc: Jan 29, 2005 -- made it configurable
| char *ptr=configvariable(A,"Mail2Hist",NULL);
| if(!ptr || !strcmpi(ptr,"No"))
| {
| if (strstr(URL,".CNM") || strstr(URL,".TBS")
| || strstr(URL,".MES") || strstr(URL,".SNT"))
| return;
| }
| //!!glennmcc: end
| //!!glennmcc: Jan 05, 2005 -- do not add smtp: or pop3: into history.lst
| // if (!URL) return;
| if (!URL
| || strstr(URL,"smtp:") || strstr(URL,"pop3:")
| //!!glennmcc: Jan 13, 2005 -- also don't add the some of the mail .DGIs
| || strstr(URL,"//movemail") || strstr(URL,"//delmail")
| || strstr(URL,"emptytrash.dgi")
| )
| return;
|
| ___________________
|
| And added a new option over on 'Options/Prefs. & Perfs.'
|
| X Save reference to mail files into History
|
| It defaults to 'No' if there is not a Mail2Hist line in arachne.cfg
|
|
|
| ============================================================================
Š|
| Sat, Jan 29, 2005 - 4:28:52 pm 
|
| glennmcc>
| Sure hope you like the 'feture'.
|
| Just look how much of a HUGE mess it has made of my history :((((
|
|
| file:l:\arachne\mail.htm
| file://readmail.dgi
| file:c:\arachne\mail\A4DOS.LST\*.mes
| file:l:\arachne\options.htm
| file:l:\arachne\system\gui\opt_misc.ah
| file:options.htm
| http://home.arachne.cz/
| file:l:\arachne\mail.htm
| file://readmail.dgi
| file:c:\arachne\mail\A4DOS.LST\*.mes
| file:C:\ARACHNE\MAIL\A4DOS.LST\A1F08D5A.MES
| file://c:\arachne\mail\a4dos.lst\A1F18C17.MES
| file:l:\arachne\mail.htm
| file://outbox.dgi
| file:C:\ARACHNE\MAIL\!0696778.TBS
| file://c:\arachne\mail\!0703933.TBS
| file://c:\arachne\mail\!0704030.TBS
| file:l:\arachne\mail.htm
| file://outbox.dgi
| mailto:
| file://c:\arachne\mail\a4dos.lst\A1F18CB6.MES
| file://c:\arachne\mail\a4dos.lst\A1F1B77D.MES
| file://c:\arachne\mail\a4dos.lst\A1F1B77F.MES
| file://c:\arachne\mail\a4dos.lst\A1F1B780.MES
| file://c:\arachne\mail\a4dos.lst\A1F1B782.MES
| file://c:\arachne\mail\a4dos.lst\A1F1B784.MES
| file://c:\arachne\mail\a4dos.lst\A1F1BB84.MES
| file://c:\arachne\mail\a4dos.lst\A1F1BB86.MES
| file://c:\arachne\mail\a4dos.lst\A1F1FB37.MES
| file://c:\arachne\mail\a4dos.lst\A1F2005F.MES
| file://c:\arachne\mail\a4dos.lst\A1F2C702.MES
| file://c:\arachne\mail\a4dos.lst\A1F2C703.MES
| file:l:\arachne\system\gui\opt_pers.ah
| file:l:\arachne\system\gui\opt_loc.ah
|
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 11:27:59 am 
|
| ray>
| Well, this is all a work in progress, so any comments are 
| tentative. I'm still trying to find a way to send a message and
| then go back to the originating message *reliably* and I think I may
| have found it -- it's a combination of Glennthink and Raythink.
Š| I'm excluding 'delmail' and 'smtp:' from Add2History() and then
| using just a single GoBack() in protocol.c. It seems that the
| GoBack problem is due to the fact that the number of goback's needed
| varies depending on a few things and they all even out if you
| exclude the above from the hisory list. Now, I'd like to exclude
| 'pop3:' too, but how can we do that and still use pop3 as the
| dialup page?
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 11:34:20 am 
|
| ray>
| Oh, just some vague conseptual ramblings:
|
| I'd say the idea with excluding stuff from the history list should
| be that pages that 'do' something (send mail, dial, delete mail, etc.)
| should almost all be excluded from history, but that pages that
| 'show' stuff should be included since when you are backing up
| through the history, you are probably wanting to 'see' something, not
| 'do' something, no? Myself, I find it very rude when I'm using the
| history list to return to, say, the send mail page, and then have to
| reach for ESC, when what I wanted was to see some previous message.
|
| What say ye?
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 1:56:30 pm 
|
| glennmcc>
| You're getting sacrry again.
|
| My code does _exaclty_ as you just discribed.
|
| BRB
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 1:58:48 pm 
|
| glennmcc>
| In urlstat.c
|
| void add2history(char *URL)
| {
| //!!glennmcc: Jan 16, 2005 -- also don't add any of the mail files themselve
| //!!glennmcc: Jan 29, 2005 -- made it configurable
| char *ptr=configvariable(& ARACHNEcfg,"Mail2Hist",NULL);
| if(!ptr || !strcmpi(ptr,"No"))
| {
| if (strstr(URL,".CNM") || strstr(URL,".TBS")
| || strstr(URL,".MES") || strstr(URL,".SNT"))
Š| return;
| }
| //!!glennmcc: Jan 05, 2005 -- do not add smtp: or pop3: into history.lst
| // if (!URL) return;
| if (!URL
| || strstr(URL,"smtp:") || strstr(URL,"pop3:")
| //!!glennmcc: Jan 13, 2005 -- also don't add the some of the mail .DGIs
| || strstr(URL,"//movemail") || strstr(URL,"//delmail")
| || strstr(URL,"emptytrash.dgi")
| )
| return;
| //!!glennmcc: end
| // printf("adding to history? GLOBAL.nothot=%d, arachne.scriptline=%d\n",
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 2:02:02 pm 
|
| glennmcc>
| And yes... the above code still allows us to use either pop3: or smtp:
| as a start page via placing then an ASF
|
| arachne.bat file://startup.asf
|
| -- startup.ASF ---
| pop3://glennmcc:*****@mail.cisnet.com:110/
| pop3://chris-pike:*******@myrealbox.com:110/
| file://inbox.dgi
| __________________
|
| BRB after trying it right from the command line instead of in an ASF
|
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 2:08:07 pm 
|
| glennmcc>
| Nope.... it does not work on the command line
|
| arachne.bat pop3://:110
|
| But I can say that the ASF method works perfectly due to the fact that
| this is the last line of my autoexec.bat
|
| C:\1BATCH\ARR.BAT 181 file://pop_all.asf /y /y
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 2:49:47 pm 
|
| glennmcc>
Š| If you feel you just must have pop3://:110 as a command line option
| again.
|
| It's this easy
|
| Simply add these 2 lines to arachne.bat at the point just above
| core.exe %1 %2 so that it looks like this.
|
| if "%1"=="pop3://:110" echo %1>history.lst
| if "%2"=="pop3://:110" echo %2>history.lst
| core.exe %1 %2
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 8:06:12 pm 
|
| ray>
| I dunno Glenn, we've got to be careful to avoid the 'workaround'
| mentality. The 'pop3:' problem should be pulled up from the roots,
| not worked around. A few of these little things are OK, but over
| time they can build up and give us a very tricky, quirky program.
| I note above that we have about the same philosophical goal with
| these history exclusions, but lets make it robust.
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 8:11:48 pm 
|
| ray>
| I spent the whole day on the blasted goback think with no luck at all.
| Here's what I know: If you have several emails and you look at one
| and delete it, the next one then pops onto the screen (quite rightly).
| But if you then reply to it and 'send', you DON'T go back to the
| originating message because it ISN'T IN THE HITORY LIST! This
| problem of the message not being in the list happens only if you
| TRASH the previous message -- any other technique for gettting
| a message on the screen results in it being entered in the history list
| automaticaly as it should. Or, hitting 'R' before replying (in the 
| above situation) also works fine to log the message in the history
| list.
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 8:16:33 pm 
|
| ray>
| What I can't figure out is where in the code, the return from
| the 'delmail1.dgi' call is (called by TRASH). If I could find it,
| a simple 'ReloadPage()' would solve the problem. Note that using
| a right click + DEL works perfectly. This imperfection with TRASH
| is a bug plain and simple and we should mash it, but I'm stumped.
|
|
|
Š|
| ============================================================================
|
| Sun, Jan 30, 2005 - 8:47:40 pm 
|
| glennmcc>
| The problem you describe is eliminated by my code because delmail does
| not get saved into history.
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 8:50:24 pm 
|
| glennmcc>
| That-is _if_ you have the option enabled to 'Save reference to mail into his
| over on Options/Prefs. & Perf.
|
| If you haven't tried my new v1.81 along with it's companion files....
| please do try it. <gg>
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 9:08:46 pm 
|
| ray>
| I just had a message from heaven and it looks like the key to this
| whole bussiness is that 'delmail1' should be saved to history because
| it points to the originating message <bonk on forhead> I was looking
| for the message with it's 'real' name. but its name in the cash
| IS delmail1.dgi. Then use one goback() only and filter 'smtp:'
| from the history list AND use global.reload = NO_RELOAD ...
| and bob's your uncle ;-)))
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 9:11:17 pm 
|
| ray>
| Sheesh, I need a break, 'cource I'll be trying 1.81 soon enough --
| once I get out of the loonie bin ;-)
|
|
| ============================================================================
|
| Sun, Jan 30, 2005 - 10:32:45 pm 
|
| ray>
| One more thing before sleepy-bye:
|
| I'm thinking that there's a deep organizational mistake in Arachne --
| probably too deep to be fixed by the likes of us: Take the 'pop3:'
| problem, it seems that the code is designed to store-then-access
Š| things, rather than access-then-store, otherwise, we'd be able to
| set DialPage to pop3, 'do' it, but then just not bother to store it
| in the history list. As it is, we've got to store it, or we can't
| 'do' it, see what I mean? Same thing with delmail1.dgi, if we don't
| store it, then we can't 'do' it. BTW, with pop3, what would happen
| if you added the <arachne nocache> tag? In that case perhaps
| the above would be wrong -- we could use pop3 with DialPage and still
| avoid having it stored. What do you think?
|
|
|
| ============================================================================
|
| Mon, Jan 31, 2005 - 12:13:59 am 
|
| glennmcc>
| Wrongo
|
| My new code does not store delmail1.dgi into history
|
| It still gets executed and deletes just fine.
|
| What does not get done and never has been done is that the file in our
| cache dir which is the result of that dgi is not stored in history.
|
|
| ============================================================================
|
| Mon, Jan 31, 2005 - 12:24:54 am 
|
| glennmcc>
| That's why using any of the mail related DGIs which move us to the next
| mail message in the box do not store the filename of that next message
| into history.
|
| When we link to the message itself in the index listing or press the
| buttons labled "<<" or ">>" we are linking directly to a message
| filename. Therefore it gets written into history.
|
| But hitting "Trash" links to delmail1.dgi which then moves us to the
| next filename. Since we did not link directly to the file name, it does
| not get saved into history.
|
| The same applies to every DGI we run no matter wheather it's a mail DGI
| or some other DGI all-together.
|
| The same applies to anything in mime.cfg which converts from one format
| to another.
|
| Linking to a .PDF automatically converts that PDF into TXT
| Reference to the PDF is stored in history... not the resulting TXT
|
| Linking to a JPG converts it to BMP
| Reference to the JPG we linked to is stored in history... not the resulting 
|
Š| ============================================================================
|
| Mon, Jan 31, 2005 - 5:43:21 pm 
|
| ray>
| Hmmmm, that makes sense. I spent the whole day yesterday trying to
| figure out why the message wasn't historied, when it *was* historied
| but under the name 'delmail1.dgi'. But I like your explanation.
| Anyway, regardless of the name, what I've got now works every time
| (Touch wood).
|
|
| ============================================================================
|
| Mon, Jan 31, 2005 - 5:46:20 pm 
|
| ray>
| ... but would it be possible to store the 'real' name of the file,
| not the .dgi? That would be nice. This whole thing is such a can
| of worms. She's your baby. I'll be content with what I've got now, and
| we'll see what you come up with. Good will come of this eventualy.
|
|
| ============================================================================
|
| Mon, Jan 31, 2005 - 7:22:45 pm 
|
| glennmcc>
| Well, I'm done with 'Mail2Hist' now.
|
| Mail2Hist No == Do not save any of the following.
|
| smtp: pop3: .CNM .TBS .MES .SNT //movemail //delmail emptytrash.dgi
|
| Mail2Hist Yes == Right back the way it was before and save all of them.
|
| And as you can see below...
| It now defaults to "Yes" if no Mail2Hist line is in arachne.cfg
|
|
| //!!glennmcc: Jan 16, 2005 -- also don't add any of the mail files themselve
| //!!glennmcc: Jan 29, 2005 -- made it configurable
| char *ptr=configvariable(A,"Mail2Hist",NULL);
| if(!ptr) ptr="Yes";
| if(!strcmpi(ptr,"No"))
| {
| if (strstr(URL,".CNM") || strstr(URL,".TBS")
| || strstr(URL,".MES") || strstr(URL,".SNT")
| //!!glennmcc: Jan 13, 2005 -- also don't add the some of the mail .DGIs
| || strstr(URL,"//movemail") || strstr(URL,"//delmail")
| || strstr(URL,"emptytrash.dgi")
| //!!glennmcc: Jan 05, 2005 -- do not add smtp: or pop3: into history.lst
| || strstr(URL,"smtp:") || strstr(URL,"pop3:")
| ) return;
| }
Š| //!!glennmcc: end
|
| if (!URL) return;
|
| // printf("adding to history? GLOBAL.nothot=%d, arachne.scriptline=%d\n",
|
|
| ============================================================================
|
| Mon, Jan 31, 2005 - 7:24:59 pm 
|
| glennmcc>
| Shit... forgot to put the space after the & symbol to keep things from
| getting messed-up here on this board.
|
| That 1st line of code should read.......
|
| char *ptr=configvariable(& ARACHNEcfg,"Mail2Hist",NULL);
|
|
| Hey gregy, 
|
| When you gonna fix that ???? <gggggggg>
|
|
| ============================================================================
|
| Mon, Jan 31, 2005 - 8:30:00 pm 
|
| ray>
| Yeah! Sarge, where are you? Aren't we supposed to be getting
| proper spacing?
|
| As for the exclusions, this is your project Glenn, so we'll see how
| it survives testing, I think the 'store-then-do' problem is real
| but we'll see.
|
|
| ============================================================================
|
| Mon, Jan 31, 2005 - 8:32:11 pm 
|
| ray>
| Oh, forgot, I should post my 'final' <heh, heh> incarnation of
| the goback thing. I'll send it to AD so's it formats properly.
|
|
| ============================================================================
|
| Wed, Feb 02, 2005 - 3:25:49 pm 
|
| glennmcc>
| New beta uploaded with both of our experiments inside. <gggg>
|
|
Š| ============================================================================
|
| Thu, Feb 03, 2005 - 3:31:07 pm 
|
| glennmcc>
| Feb 03, 2005
| Linking to the Arachne version number at the top of the screen will
| now take us 'Up one level' when we are viewing a remote file.
| (will still go to 'about.htm' if we are currently local)
|
| ============================================================================
|
| Thu, Feb 03, 2005 - 3:48:12 pm 
| URL
|
| glennmcc>
| New beta on its way. :))
|
| ============================================================================
|
| Thu, Feb 03, 2005 - 6:59:37 pm 
|
| ray>
| Very interesting. 
|
|
| ============================================================================
|
| Thu, Feb 10, 2005 - 7:54:04 pm 
|
| ray>
| hee hee hee ... got Alt+T workin' 
|
| ============================================================================
|
| Thu, Feb 10, 2005 - 8:22:04 pm 
|
| glennmcc>
| Show me the money.
|
| I mean the code. <ggggg>
|
| ============================================================================
|
| Thu, Feb 10, 2005 - 10:15:46 pm 
|
| ray>
| Not until you help me parse that line ;-)
| ... well, the fact is I figured it out but I'm more confused than ever:
| why in hell are the arguments to insight.exe -2 .... stored in
| $query$.tmp (or whatever it is) ?? Why are they not just passed to
| the command line? For such trouble to have been gone to there must
| be a reason.
|
|
Š|
| ============================================================================
|
| Fri, Feb 11, 2005 - 12:26:21 pm 
|
| mht>
| There is at least one reason: DOS command line length is VERY limited --
| 128 characters for everything (the command itself and its parameters).
|
| ============================================================================
|
| Fri, Feb 11, 2005 - 7:24:30 pm 
|
| ray>
| Hi Michal!
|
| That might be all the reason necessary, thanks.
|
| Glenn,
|
| I'm not quite the hotshot I thought yesterday, see AD for my problem
| with Alt+T. This should be fixable. BTW, I tried to acsess the
| .DGI code to do a 'trash' directly but it is so hidiously complecated.
| It will be years before I can attempt something like that, so you
| were basicaly right about that aproach being either impossible or
| at least close to.
|
|
| ============================================================================
|
| Sat, Feb 12, 2005 - 8:37:09 am 
|
| ray>
| Joe, Michal,
|
| We've got a live one!
|
|
|
| ============================================================================
|
| Sat, Feb 12, 2005 - 3:20:08 pm 
|
| glennmcc>
| A live one what ?
|
| ============================================================================
|
| Sat, Feb 12, 2005 - 4:00:10 pm 
|
| ray>
| Fresh meat! ... Andrej ;-)
|
|
| ============================================================================
Š|
| Sun, Feb 13, 2005 - 8:44:45 pm 
|
| ray>
| Ya know, I like this Alt+T already, if you've read a measage and want
| to trash it, instead of having to 'home' then tab to the trash button
| (which you have to watch) Alt+T, Enter does the job every time and you
| don't have to think about which link you're on before you press enter.
| Next: Alt+S to go to the 'Send mail now' button.
|
|
|
| ============================================================================
|
| Mon, Feb 14, 2005 - 3:56:58 pm 
|
| glennmcc>
| Home, Tab, Tab, Enter is now reduced to Alt, T, Enter
|
| For a savings of 1 keystroke.
|
| You are one lazy SOB... you know that ?
| <gggggggggggggg>
|
| ============================================================================
|
| Mon, Feb 14, 2005 - 5:40:55 pm 
|
| ray>
| yes, but you miss the point, Alt+T is just for testing, what I'm
| really after is the ability to do this sort of thing. I picked
| 'trash' for testing because it's easy to add a line to arachne.bat
| that automaticaly restores the file to the inbox after each test in
| preparation for the next and the next add infinitum. Also, if the
| cursor is low in the document, you never know what link TAB will take
| you too, so it can be an arbitrary number of keystrokes to get you
| back to the button of your choice. This might sound dumb, but I like
| to be able to use a keystroke that is guaranteed to have the desired
| result without me having to, say, page up, and then TAB several times
| while keeping an eye on the screen. When I'm working with my email,
| I really find reaching for the mouse to be irritating, so a friendlier
| set of keystrokes for navigating the messages will be nice for me 
| even if the idea never catches on.
|
|
| ============================================================================
|
| Mon, Feb 14, 2005 - 6:56:10 pm 
|
| ray>
| testing
|
|
| ============================================================================
|
Š| Mon, Feb 14, 2005 - 7:24:45 pm 
|
| glennmcc>
| OK, I see the point of your experiments.
|
| But you're still a lazy SOB <gggggggggggggg>
|
| ============================================================================
|
| Mon, Feb 14, 2005 - 9:35:57 pm 
|
| ray>
| 'course! ... I'm a mailman.
|
|
|
| ============================================================================
|
| Tue, Feb 15, 2005 - 6:53:17 pm 
|
| ray>
| Glenn,
|
| See that little fix I made to the mouse colors? Now what would be
| really nice would be to make it so that when a page loads, it
| checked the possition of the mouse, and verified if it was on a
| hot link, as opposed to just starting off with the assumption that
| it's not. I tried a few things with no luck, and even some nasty
| crashes. Any ideas?
|
|
| ============================================================================
|
| Tue, Feb 15, 2005 - 7:54:16 pm 
|
| glennmcc>
| You ae not 'on a hot link' simply because the cursor is over it.
|
| You are 'on it' only after moving the cursor over it.
|
| So..... jiggle it a little... it'll open. <ggggggggg>
|
| ============================================================================
|
| Tue, Feb 15, 2005 - 8:35:02 pm 
|
| glennmcc>
| There we go.... no more jiggling needed. <ggggggg>
|
| void mouseon(void)
| {
| char *mouseonlink=NULL;
| if(!global_nomouse)
| {
| mouseonlink=onmouse(0);//check to see if we are over a link
Š| if (mouseonlink)x_yncurs(1,mousex,mousey,user_interface.brightmouse);
| else
| x_yncurs(1,mousex,mousey,user_interface.darkmouse);
| }
| }
|
| void mouseoff(void)
| {
| if(!global_nomouse)
| x_yncurs(0,mousex,mousey,user_interface.darkmouse);
| }
|
|
|
|
| ============================================================================
|
| Tue, Feb 15, 2005 - 8:48:55 pm 
|
| glennmcc>
| Unexpected 'pleasent side effect'.
|
| When take the first link on page and then come back to that page
| via Ctrl+leftarrow without having moved the mouse nor having used the
| TAB key on the new page we can TAB to the next link on the orignal page.
|
| The mouse no longer gets 'stuck' on the first link. :))))
|
|
| ============================================================================
|
| Tue, Feb 15, 2005 - 9:29:54 pm 
|
| glennmcc>
| While I was at it, I fixed the same kind of problem where the mouse
| gets 'stuck' while tabbing through the links on page and one is hit that
| is under an image of less than 2 pixels in width.
| (went ahead and did it for height also just to be sure) <gggg>
|
| --- in html.c -------
|
|
| if(getvar("HEIGHT",
| {
| int i=try2getnum(tagarg,frame->scroll.ysize);
| if(i<2)i=2;//!!glennmcc: Feb 15, 2005 -- fixes 'sticky mouse'
| if((!znamrozmery || i!=img->size_y) && !egamode && !vga16mode)
| {
| img->resize_y=i;
| img->resize_x=img->size_x;
| }
| img->size_y=i;
| znamrozmery=1;
| }
|
Š| if(getvar("WIDTH",
| {
| int max=p->docRight-p->docLeft;
| int i;
| if(max<0) max=0;
| i=try2getnum(tagarg,max);
| if(i<2)i=2;//!!glennmcc: Feb 15, 2005 -- fixes 'sticky mouse'
| if((!znamrozmerx || i!=img->size_x) && !egamode && !vga16mode)
| {
| img->resize_x=i;
| if(!img->resize_y)
| img->resize_y=img->size_y;
| }
| img->size_x=i;
| znamrozmerx=1;
| }
|
| //kill adds (468x60)?
| if(user_interface.killadds && reg && img->size_x==468 && img->size_y==60)
| break;
|
|
| ============================================================================
|
| Tue, Feb 15, 2005 - 9:30:31 pm 
|
| ray>
| Beautiful. Two minds are better than one, I love synergy.
|
| I write it this way:
|
| void MouseOn(void) 
| { 
| if (!gGlobalNoMouse) 
| // { RAY/glennmcc: 05-02-15: Make sure mouse's off-link color is correct. 
| // ... and 'giggle' the mouse to see if it should be hot or cold 
| // when a new page is loaded ;-) 
| // x_yncurs(1, MOUSEX, MOUSEY, 15); 
|
| if (OnMouse(0)) // We are on a link. 
| x_yncurs(1, MOUSEX, MOUSEY, sUI.hotmouse); 
| else 
| x_yncurs(1, MOUSEX, MOUSEY, sUI.coldmouse); 
| // } 
| } 
|
|
|
|
| ============================================================================
|
| Tue, Feb 15, 2005 - 9:32:08 pm 
|
| ray>
| What, more improvements? ... a dude can't keep up.
Š|
|
| ============================================================================
|
| Tue, Feb 15, 2005 - 9:46:22 pm 
|
| glennmcc>
| EXCELLENT !
|
| Your version of the code in guitick.c makes the EXE 16 bytes smaller
| than it is with my version.
|
| Two heads ARE better than one. <g>
|
| ============================================================================
|
| Wed, Feb 16, 2005 - 7:25:47 am 
|
| ray>
| There's something instinctive about wanting to create a variable to
| pass a value, but I've been working hard to overcome it and just
| using return values from functions directly; you get tighter cleaner
| code, and single use variables are often confusing since you can't be
| sure they aren't used again somewhere.
|
|
|
| ============================================================================
|
| Wed, Feb 16, 2005 - 7:27:13 pm 
| URL
|
| glennmcc>
| v1.81 has just been released :))
| http://www.cisnet.com/glennmcc/
|
|
| ============================================================================
|
| Wed, Feb 16, 2005 - 9:57:54 pm 
|
| ray>
| Good, I've got some catching up to do.
| Grab it manyana.
|
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 5:24:51 pm 
|
| ray>
| tesing
|
|
| ============================================================================
Š|
| Thu, Feb 17, 2005 - 5:26:24 pm 
|
| ray>
| testing
|
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 5:35:08 pm 
|
| ray>
| testing
|
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 5:38:04 pm 
|
| ray>
| testing
|
|
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 5:38:22 pm 
|
| ray>
| (sorry about this ;-)
|
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 5:43:08 pm 
|
| ray>
| testing
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 5:46:12 pm 
|
| ray>
| Glenn, 
|
| This mouse thing still isn't perfect. Load a page that has
| links and checkboxes (like inbox). position the mouse next to a
| checkbox, reload the page and hit tab such that the cursor pops
| to a ckeckbox. Now tab around the screen, and notice that as soon
| as the cursor hits a link it turns 'hot' as it should, but now it
| stays 'hot' even if you tab 'round to the ckeckboxes again. Moving
| the mouse by hand at any time fixes the colors. Can yo confirm?
|
|
Š|
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 6:01:08 pm 
|
| ray>
| testing
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 6:04:41 pm 
|
| ray>
| last time!
|
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 7:23:20 pm 
|
| ray>
| Glenn,
|
| problem above confirmed in 1.81. The fact that the cold to hot
| transitions occur with not problem whereas the opposite is not
| the case should be a good hint where the problem lies.
|
|
|
| ============================================================================
|
| Thu, Feb 17, 2005 - 8:19:33 pm 
|
| ray>
| Usin' 1.81 ;-)
|
| Hey Glenn, you didn't do the reload thing with the inbox! ... or does
| that require some new config option? ... yeah now I remember ...
|
|
| ============================================================================
|
| Fri, Feb 18, 2005 - 9:23:36 pm 
|
| ray>
| Nuts, there is always one more thing:
|
| ONMOUSE.C ==================================================================
|
| if (*ptr == '#' && click && click != MOUSE_RELEASE) 
| { 
| GotoAname(p 
| goto nolink; 
| } 
Š| gLastOnMouse = linkonmouse; 
| return ptr; 
| } // End: if (atomptr->type == HREF 
| else 
| { 
| // { RAY/glennmcc: 05-02-15: Make sure mouse's off-link color is correct. 
| // NB must turn mouse off then on to change color. 
| x_yncurs(0, MOUSEX, MOUSEY, sUI.coldmouse); 
| x_yncurs(1, MOUSEX, MOUSEY, sUI.coldmouse); 
| >>>>>>>>>> DefaultMsg(); // Erase status bar. 
| // } 
| if (atomptr->type == FORM) 
| return NULL; 
| } 
| goto nolink; 
| }// end analyse atom 
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 5:07:11 pm 
|
| ray>
| Glenn,
|
| Is that HTTPSTUB stuff working now? As of 1.79 I seem to recall that
| it was still a work in progress.
|
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 5:28:40 pm 
|
| glennmcc>
| Nope......... not a 'work in progress'.
|
| It works perfectly,
| albeit vvvveeeerrrryyyy sssssllllooooowwwwllllyyyy in Win9x <gggggg>
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 5:30:18 pm 
|
| ray>
| OK roger that.
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 5:38:19 pm 
|
| ray>
| ... so you can use Arachne with a Winmodem? ... EEEK! ... but only
Š| under Windows itself, obviously.
|
| BTW, a small bug: Remember that change in popget to go to HEX 
| names for the mail? This is very unlikely to ever bite, but just as
| a point of principal, the format specifier you use will produce lower
| case "123def45.cnm" but the overflow test is for a capital letter:
|
| if(fname[8]>'F') ... shit, in 50 years, someones's gona bitch!
|
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 6:22:43 pm 
|
| glennmcc>
| In 50 years I will be 95yrs old... and I will be saying
| "Fuck you, fix it yourself !!!" to everyone who bitches about it. ;-)
|
| But it's a moot point anyways.
|
| Borland C/C++ programs themselves will only work correctly for 33 more yrs.
|
| In April 2038, date handling routines in the compiler itself will 'overflow'
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 6:44:04 pm 
|
| glennmcc>
| Oops, I was off a little bit on that date.
|
| Here are the headers of 2 emails I just saved into my outbox.
|
| Reset date to Jan 18, 2038
|
| Date: Mon, 18 Jan 2038 20:38:22 -0500 
|
| Reset date to Jan 19, 2038
|
| Date: Thu, 01 Jan 1970 00:00:00 -0500 
|
| That's it, from Jan 19, 2038 and beyond... all dates will be locked at
| Jan 01, 1970 and the time will locked at midnight.
|
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 6:46:20 pm 
|
| ray>
| Well I was sorta pulling your leg; let 'em bitch. Did I find anything
| interesting in that CSS stuff? Logical patterning is hardly a
| reliable guite in these things, but you never know.
Š|
| BTW I just pulled in you 'time remaining' code -- works nice, but
| a bit jumpy at first. Now, leme see if I can improve it, like I
| do ...
|
| Oh and no comment on that last stuff in onmouse, do you not aprove?
| Looks to me like I've got the mouse behaving right all the time.
|
|
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 6:55:35 pm 
|
| glennmcc>
| Nope, I do not approve of this........
| // NB must turn mouse off then on to change color. 
|
| That's the kind of SHIT that _I_ do. 
|
| Come on now Ray.......... You can do better than that. <gggggg>
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 6:59:14 pm 
| URL
|
| glennmcc>
| As for the CSS stuff.... test it, see if your code works to keep the
| change of color on a CSS link within a framed page aligned properly.
|
| http://www.cisnet.com/glennmcc/my-stuff/testcssf.htm
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 7:05:49 pm 
|
| ray>
| Na, I caint do no better! There might be a better way of course, but
| at least it works for now. As to the CSS stuff, I haven't changed
| anything, just making comments, it's your baby. But if you don't
| want to kick it around, send me some page or other where I can 
| test it, but hell, I have no idea what CSS even is <;-(
|
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 7:06:41 pm 
|
| ray>
| Nuts, which one of us is going to drop dead first? <gggg>
|
Š|
| ============================================================================
|
| Sat, Feb 19, 2005 - 7:20:56 pm 
|
| glennmcc>
| I ain't never gonna die.
|
| In 2038 I'll just have to switch to different compiler to do v99.81 <g>
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 7:28:14 pm 
|
| glennmcc>
| As for what CSS is.......
|
| Options/Prefs. & Perf.
|
| X Enable stylesheets (CSS/1.0) 
|
| Save the change, choose to use new settings.
|
| Now come back here again and move you mouse over the URL on my post up above
| (or any of the other links on this page)
|
| See it change from blue to red ?
|
| That is being done by these lines of CSS at the very top of this HTM file.
|
| <style>
| A:link { text-decoration: none; color: #00f}
| A:hover { text-decoration: none; color: #f00}
| </style>
|
| My test page for CSS in a right-hand frame is.......
| http://www.cisnet.com/glennmcc/my-stuff/testcssf.htm
|
|
|
| ============================================================================
|
| Sat, Feb 19, 2005 - 9:44:56 pm 
|
| ray>
| I'll look at it tomorrow, g'nite.
|
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 10:54:41 am 
|
| ray>
| Yup, get the blue to red OK, so I take it that's not the problem.
Š| Off we go to ...testccf.htm
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 10:56:53 am 
|
| ray>
| Oh ... what is supposed to happen? Another color change? If so,
| exactly where on the page?
|
|
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 1:17:57 pm 
|
| mht>
| The "Year 2038 problem" is `typedef long time_t;` in TIME.H and related
| stuff. Using `unsigned long` instead delays the problem until year 2106.
| Modern C/C++ compilers already use `unsigned long`.
|
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 3:36:20 pm 
|
| glennmcc>
| 101 years ought to be enough for anyone.
|
| But I would prefer 640 <gggggggggg>
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 3:50:06 pm 
|
| glennmcc>
| Ray,
|
| The CSS is being used in the right-hand frame.
| http://www.cisnet.com/glennmcc/my-stuff/testcssf.htm
|
| http://www.cisnet.com/my-stuff/cssbroke.gif (v1.79)
|
| http://www.cisnet.com/my-stuff/cssfixed.gif (v1.81)
|
|
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 3:52:31 pm 
|
| glennmcc>
| Duh !!
| (forgot my own name) <gggg>
|
Š| http://www.cisnet.com/glennmcc/my-stuff/cssbroke.gif (v1.79)
|
| http://www.cisnet.com/glennmcc/my-stuff/cssbroke.gif (v1.81)
|
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 3:53:34 pm 
|
| glennmcc>
| Shll we try this one last time ? <g>
|
| http://www.cisnet.com/glennmcc/my-stuff/cssfixed.gif (v1.81)
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 4:48:06 pm 
|
| ray>
| OK thanks. BTW these test pages of yours are just the thing -- I like
| the concept and the inplementation.
|
|
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 4:55:24 pm 
|
| ray>
| Yuk, that's awful. But I tried 'testcssf.htm' again and it looks
| perfect with the code just as you had it (unless I fixed something
| by accident formating it for B5 he he he ...). Are you telling me
| that page should look buggered up like your busted example above?
|
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 5:21:40 pm 
|
| glennmcc>
| No, no, no........
|
| It _should_ look 'fixed' ;-)
|
| Hence the filenames cssbroke.gif and cssfixed.gif <gggg>
|
| ============================================================================
|
| Sun, Feb 20, 2005 - 10:11:29 pm 
|
| ray>
| Ya ya, I mean the file 'testcssf.htm' looked ok, if that one isn't
| a problem then can you show me a page that is a problem, so's I can
| see a sample defect?
|
Š|
|
| ============================================================================
|
| Mon, Feb 21, 2005 - 6:43:23 pm 
|
| ray>
| Glenn,
|
| You've probably caught this already but 'pppglenn.htm' is missing
| from 1.81
|
| ============================================================================
|
| Mon, Feb 21, 2005 - 8:06:05 pm 
|
| glennmcc>
| Nope, not missing... renamed to pppenhan.htm
|
| LFN would read......... ppp_enhanced.html <g>
|
| As for testcssf.htm
|
| That page will be messaed-up when you place the mouse pointer over any
| of the links in the right hand frame when using any version of Arachne
| before v1.81
|
| The change of color from blue to red will not place the new color
| directly over top of the old color. Instead, it will be shifted left
| as is shown in cssbroke.gif
|
| The code I changed for v1.81 shifts it over to the right where it's
| supposed to be.
|
| Oh yeah, for doing the testing on testcssf.htm
|
| You _do_ have both CSS and frames enabled..... don't you ???
|
| ============================================================================
|
| Mon, Feb 21, 2005 - 8:40:13 pm 
|
| ray>
| re: pppglenn.htm:
|
| It's just that 'clearlok.asf' reffers to it by it's old name and when
| you hit the 'clear cache' button ... etc. etc.
|
| re: css:
|
| Ok so *that* page should look right (and does). Do you have a
| page that shows what is still broke (the suff your commented .....
|
| HEY! .... here I'm trying to talk myself into some work that I don't
| know anything about ... silly ass < g g g g g >
Š|
|
| ============================================================================
|
| Tue, Feb 22, 2005 - 7:21:08 am 
|
| glennmcc>
| By using testcssf.htm you can see wheather or not your CSS code i broken.
|
| If it's broken you will see what v1.79 does.... cssbroke.gif
|
| If it's fixed you will see what v1.81 does ... cssfixed.gif
|
| The same page (testcssf.htm), was used for both tests.
|
| ============================================================================
|
| Tue, Feb 22, 2005 - 2:10:41 pm 
|
| ray>
| Maybe I missunderstood something, I took it from the comments in your
| code that there was still some problem besides the one you fixed.
| What was it now ... something about up and down? No matter, it's your
| baby.
|
|
| ============================================================================
|
| Tue, Feb 22, 2005 - 6:20:14 pm 
|
| glennmcc>
| Yes, that's it exactly.
|
| The code that is 'active' fixes the left-right shifting of CSS LinkColorChan
|
| The commented-out code was my failed attempt to fix the up-down shift.
|
|
| ============================================================================
|
| Tue, Feb 22, 2005 - 9:56:53 pm 
|
| ray>
| Ok, if you find a page that displays the up-down problem, send me the
| URL so's I can have a peek. I doubt that I'll be able to do anything
| much but my curiosity is pricked. BTW, I sent myself an email with
| a JPG attached and there was no problem so I'm guessing that if Sam
| had a goback headache is is some witches' brew combination of my
| goback strategy mixed with your history list exclusions. Back out
| of that Glenn, she's nothing but trouble. Can't remember the name, but
| there's a function like" RemoveFromCache()" that might do the trick,
| so you could let things work normaly then after the fact remove the
| garbage from the cache and (with any luck) the history list too. But
| my bet is that the only way to realy make this work right is to turn the
| code upside down so it 'executes then stores' as opposed to 
Š| 'storing then executing'.
|
|
| ============================================================================
|
| Wed, Feb 23, 2005 - 3:52:56 pm 
|
| glennmcc>
| I think I have figured-out what was going on during your tests.
|
| What you experiencing was the fact the you were starting Arachne with one of
| these 2 situations.
|
| 1) arachne.bat pop3://110:
| 2) arachne.bat (with dialpage set to pop3://110:)
|
| In these situations, you must store pop3://110: into history because
| that 1st thing that happens is that she shells-out to run the dialer.
| (miniterm, Lsppp or whatever you might be using to dial)
|
| Now when the dialing/connecting/packet_driver_loading is done, you are
| returned into Arachne to execute what was saved in the history before
| shelling-out to dial.
|
| That is the only situation in which we must store something into history
| before we can do it.
|
| If we are already on line...... history does not come onto play for
| anything other than going back to what we did before.
|
| This is the chain of events when already connected
|
| 1) enter item on URL line and press enter or take a link to the item
| or enter it on the command line after arachne.bat or have it as dialpage
| 2) save reference to the item into history
| 3) the item in question is 'executed' via URL not via reference to it in his
|
|
| Now the chain of events when we are not already connected
|
| 1) enter item on URL line and press enter or take a link to the item
| or enter it on the command line after arachne.bat or have it as dialpage
| 2) save reference to the item into history
| 3) shell-out to dial/connect/load packet driver
| 4) the item in question is 'executed' via reference to it in history due
| to the fact that our most recent URL was arachne:dial instead of the
| item in question
|
| Now.... have I made that about as clear as mud ??? <gggggggg>
|
|
|
| ============================================================================
|
| Wed, Feb 23, 2005 - 5:44:09 pm 
Š|
| ray>
| Well, I think you have the bull by the horns, you may be right about
| the reason for my theory about 'store then execute'. The only thing I am
| sure of is that I don't realy know what the hell IS going on <gg>.
| But IIRC I had trouble sending mail even after I was on line if
| 'smtp:' was excluded. Anyway, she's your project.
| I have no doubt that this whole thing will eventualy be
| solved much to the improvemnt of the history list; it is a very
| worthwile project.
|
|
|
|
| ============================================================================
|
| Thu, Feb 24, 2005 - 11:28:52 am 
| URL
|
| glennmcc>
| Here you go.
|
| The same test page but with the 2 frames top/bottom instead of right/left.
|
|
| ============================================================================
|
| Thu, Feb 24, 2005 - 4:43:31 pm 
|
| ray>
| got it.
|
| ============================================================================
|
| Thu, Feb 24, 2005 - 9:11:20 pm 
|
| ray>
| Glenn, I'm noticing that the highlight color varies randomly 
| with my core and with 1.81. Is this normal, or am I missing some
| option in Arachne.cfg?
|
|
|
| ============================================================================
|
| Thu, Feb 24, 2005 - 9:27:27 pm 
|
| glennmcc>
| It's normal so to speak.
|
| That is if we can call a small bug that I have yet to fix " normal ". <gggg>
|
| ============================================================================
|
| Mon, Feb 28, 2005 - 4:41:25 pm 
Š|
| ray>
| Glenn,
|
| Looking forward to seeing your frames fix. This is the first problem
| where I've tried to get any idea how low level screen writes are done,
| so seeing the variables used, and the functions called, will be 
| most informative.
|
|
| ============================================================================
|
| Mon, Feb 28, 2005 - 5:45:52 pm 
|
| glennmcc>
| //!!glennmcc: begin Jan 26, 2005 -- added to compensate for CSS linkhover in
| frame
| if(dx<p->htmlframe[atomonmouse.frameID].scroll.xtop)
| dx-=p->htmlframe[atomonmouse.frameID].scroll.xtop;
| //!!glennmcc: end
|
| xx=atomonmouse.x-dx+p->htmlframe[atomonmouse.frameID].scroll.xtop;
|
| //!!glennmcc: begin Feb 27, 2005 -- added to compensate for CSS linkhover in
| frame
| if(p->htmlframe[atomonmouse.frameID].scroll.ytop>p->htscrn_ytop &&
| dy<p->htmlframe[atomonmouse.frameID].scroll.ytop)
| dy-=p->htmlframe[atomonmouse.frameID].scroll.ytop-p->htscrn_ytop;
| //!!glennmcc: end
|
| The 'X' and 'Y' fixes are identical with the exception that we only do the f
| when current p->htmlframe[atomonmouse.frameID].scroll.ytop
| is greater than p->htscrn_ytop and then the 'Y fix' itself
| requires also subtracting....... p->htscrn_ytop
|
|
| p->htscrn_ytop is the current 'Y cooridate of the top of the page itself.
|
| 0 when in fullscreen layout
| 25 when the small Ikons, URL/title bar are across the top
| 50 when the small Ikons, URL bar, Title bar are across the top
| 50 when the buttons are on right side, URL bar, Title bar across the top
| 100 when the large buttons, URL bar, Title bar are across the top
|
| _That_ was the 'magic number' we needed.
|
| The current number stored in p->htscrn_ytop
|
|
| There are still 2 small problems to fix.
|
| 1) keep the CSS link color change from 'overlapping' the frame boudery
|
| 2) refine the 'Y fix' so that the color change goes back to the original
| color of the link when moving to the next link.
Š| (the 'X fix' already does it... the 'Y fix' does not)
|
|
|
|
| ============================================================================
|
| Mon, Feb 28, 2005 - 6:32:48 pm 
|
| ray>
| So there's our magic variable. Thanks for that. Keep me posted if
| you fix the rest.
|
| ============================================================================
|
| Mon, Feb 28, 2005 - 10:27:13 pm 
|
| ray>
| Well, that works like a charm. But it seems to me that the two
| problems you mention above are really just one problem. Using the
| left-right test page, if you scroll to the right then try the
| highlight, notice that where the links are 'in bounds' they change 
| colors back and forth as they should. Only the 'out of bounds'
| text takes and holds the wrong color. Thus, it seems to me that
| the Y problem is really a variation of the out of bounds problem.
| Do we have another variable that keeps track of the scroll offset?
| If we did, we could use it to truncate the link text at the proper
| left margin of the frame. This already happens for unhighlihted text,
| so it shouldn't be too much trouble to teach the trick to the
| highlighting function.
|
|
| ============================================================================
|
| Tue, Mar 01, 2005 - 7:31:27 am 
|
| ray>
| Or maybe I said that backwards. The trouble seems to be not
| in the turning off of the highlight, but in the drawing of the atom
| when a highlight is apropriate. In other words, she draws atoms
| fine until such times as they need to be drawn with a highlight, then
| she 'forgets' about the stuff out of bounds.
|
|
|
| ============================================================================
|
| Thu, Mar 03, 2005 - 2:03:47 pm 
|
| ray>
| "XSWAP memory manager" won't light up either.
|
|
| ============================================================================
|
Š| Thu, Mar 03, 2005 - 7:25:25 pm 
|
| ray>
| But I see that all of the above is identical in 1.79 and 1.81, so
| it wern't my doing ;-)
|
|
| ============================================================================
|
| Thu, Mar 03, 2005 - 8:45:37 pm 
|
| glennmcc>
| Scroll further on down the page in about.htm
|
| You'll see that all of the links which do 'light up' (but are missaligned)
| in previous versions..... no longer 'light up' at-all.
|
|
|
| ============================================================================
|
| Thu, Mar 03, 2005 - 9:19:05 pm 
|
| ray>
| You mean the links which wrap to the next line? I find them
| dead in 1.79 and ...
|
| ============================================================================
|
| Thu, Mar 03, 2005 - 9:20:53 pm 
|
| ray>
| ... just as dead in 1.81, eg. "See Suntech homepage"
|
| So it ain't that. What links are you talking about?
|
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 4:59:10 pm 
|
| glennmcc>
| No, not the ones that wrap.... that's a different bug altogether that
| we'll need to tackle some other time.
|
| All of the rest of the links about.htm which do not wrap should also
| change color.
|
| All the way down at the bottom of the page is a link to file://system.htm
| (minimal config)
|
| That one changes color too.... it does the color change but missaligned
| to shifted upward in 1.79 and back.
|
| But now it does nothing.
Š|
|
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 5:07:30 pm 
|
| glennmcc>
| If you scroll down untill
| COMTOOL.COM - public modem tool, see details
|
| is right the bottom of the HTML screen and is cut in half (top half visible,
| bottom half scrolled below the HTML screen) then place the mouse over the
| next link up which is for PLANY..... the see details link for PLANY
| works the same as in v1.79 and back... the color change works but is
| still shifted upward. The see details link on comtool does the same.
|
| If you scroll up just a few pixells so that the comtool line is no
| longer cut in half but is fully visible. Now the comtool link changes
| color but is shifted upward and the plany link does not change color.
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 5:10:38 pm 
|
| glennmcc>
| Scroll a little bit more so that this is now visible....
|
| Credits:
| All registered users of Arachne. 
|
| Now all of the links currently showing on screen are 'color change dead'.
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 5:14:58 pm 
|
| glennmcc>
| Now, 'right-click' somewhere in the about.htm fram so that only about.htm
| will show all by itself instead of being in the lower frame of our
| frameset.
|
| Scroll anywhere on the page and checkout the links.
|
| Every one of that is not wrapped to 2 lines or does not begin all the
| way to the left of its particular line changes color and is aligned
| correctly.
|
| When we can get every link on about.htm working correctly..........
| our job is done. <ggggggggg>
|
|
|
| ============================================================================
|
Š| Fri, Mar 04, 2005 - 7:26:35 pm 
|
| ray>
| All those problems are negative here, that's looking at 'about Arachne'
| via 'testcssf2.htm live from you site.
|
| Do we have code confusion? I'll repost to AD right now. Or could it
| be some config option?
|
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 7:43:21 pm 
|
| glennmcc>
| Do you mean to say that all of the links on about.htm when viewed in the
| lower frame of testcss2.htm work correctly ?
|
| That even 'minimal configuration' at the very bottom of the page _does_
| change color when your mouse is over it ?
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 7:44:37 pm 
|
| glennmcc>
| BRB with a screen-cap
|
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 7:56:50 pm 
|
| glennmcc>
| 1 screen-cap is good... 3 is better. <gg>
|
| http://www.cisnet.com/glennmcc/my-stuff/css-01.gif
| http://www.cisnet.com/glennmcc/my-stuff/css-02.gif
| http://www.cisnet.com/glennmcc/my-stuff/css-03.gif
|
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 7:58:45 pm 
|
| glennmcc>
| #1 and #2 are while in the frameset
|
| #3 is about: viewed by itself and not in the framesset.
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 8:18:22 pm 
|
| glennmcc>
Š| eHighlight = 1; // Make sure link highlights with sUI.hotmouse color.
|
| What if it was already that color ????
|
| ie: the CSS in a page has it start out as red and your hotmouse color is red
|
| What happens is....... no color change.
|
| Q2:
| What if your hotmouse color is the same as the BGcolor of the page where
| the link is ?
|
| ie: link starts out as blue on a red BG.... hotmouse color is red
|
| What we get is an invisible link because it is now a red link on a red BG.
|
| The CSS sets the start color and change-to color.
|
| We must not over-ride this.
|
| Here's what it looks like.........
|
| <style>
| A:link { text-decoration: none; color: #00f}
| A:hover { text-decoration: none; color: #f00}
| </style>
|
| All links on the page will be blue and change to red when we mouse-over
| and then back to blue when we move the mouse off of them.
|
| With your eHightlight=1; ......... they start as blue and change to
| hotmouse color. (which will be different in all setups)
|
| If someone has their mouse set to not change colors.... they change to
| white which means that a blue link on a white BG changes to a white
| link on a white BG.
|
| To sum it up.........
|
| eHighlight = 1; // Make sure link highlights with sUI.hotmouse color.
|
| VERY BAD IDEA. :(
|
|
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 9:21:57 pm 
|
| ray>
| re: screen caps: I am showing those links all just fine. If you have
| confirmed my code, what's next? Should I send you my .exe to confirm
| my results?
|
| re: eHighligh: First, this wasn't exactly meant to be a permanent
Š| solution. Still what is better? As it was, the highlight could be
| any color at random, which is hardly acceptable and a random color
| could itself have all the same problems you mention
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 9:25:48 pm 
|
| ray>
| Do pages contain info on the color that is desired for highlighed
| links? If so, how do we grab it? If not, then we just have to 
| pick one and take our chances, no? Or, would it be possible to
| grab the pages' colors and then write a simple algorithm to insure
| that hot links are always made with a unique color?? Baring that, I'd
| say that eHighlight is acceptable since, as with the mouse itself, one
| always runs the risk of invisibility due to background, nicht?
|
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 9:30:17 pm 
|
| ray>
| Ooops, re-reading you post, I see that there is a specified color
| for links, so then, why is it not being used and how do we fix it?
| Perhaps the hot mouse color should only be a default if nothing else
| is specified. I dunno, I no speakieeee HTML but I'll take a fixed
| color over a random color any day.
|
|
|
|
|
| ============================================================================
|
| Fri, Mar 04, 2005 - 10:14:52 pm 
|
| glennmcc>
| I gotta get this SOB figured-out or it's gonnna drive me freakin' nuts !<g>
|
|
| Please zip-up and send me your current copy of onmouse.c
|
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 10:07:40 am 
|
| ray>
| Ok. BTW, I'm glad you said it -- let's not drive ourselves 
| loonie eh? The fact is, we're both working too hard and somethings
| got to crack eventualy. I've told myself to keep it down to
| 5 hours per day < GGGGGG >
|
|
Š|
| ============================================================================
|
| Sat, Mar 05, 2005 - 10:09:02 am 
|
| ray>
| Oh, about this onmouse stuff, just off hand, it looks to me like
| yo might be suffering from an improper initialization, or something
| like that.
|
|
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 3:48:53 pm 
|
| glennmcc>
| Mr. Andrews.... you are a GENIUS !!!!!!!!!
|
| I just 'melded' your code into my copy of onmouse.c
|
| It works perfectly. :)
|
| You squished that bug but proper.
|
| Thank you very much for all of this hard work.
|
| Gotta admit tho...... I killed 'eHighlight' in my copy.
|
| Don't like that idea.. ain't gonna use it. <gggg>
|
|
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 4:13:27 pm 
|
| ray>
| Thanks,
|
| re: eHighlight, as you mentioned yesterday, if there is a proper
| tag for hover color, how come we can't use that? If we did
| then eHighlight would become either pointless, or at best a
| sort of default. I'm out of my deapth, but even if we don't have
| the ability to parse hover color right now, it seems like it would
| be very simple to do.
|
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 4:19:37 pm 
|
| ray>
| Oh, one more thing about CSS, as you probably noticed, only two
| screen edges are 'guarded' right now (left, with "x1 = 0, and
Š| the bottom of the HTML window (to protect the scroll bar). I'd like
| to see what would happen if a highlighted link was forced by scrolling
| to be off the window to the right. Can you make as suitable test 
| page and give it a go? I think it will crash.
|
|
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 4:23:00 pm 
|
| glennmcc>
| Yes, there is a 'proper tag'.
|
| Yes, we _can_ use it.
|
| In-fact, it _is_ being used........ but there is a bug in it.
|
| Just as there _was_ a bug in linkhover within frames.
|
| Wanna help me squash another bug ? <ggg>
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 4:24:29 pm 
|
| glennmcc>
| Off the right ??? 
|
| BRB
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 4:48:38 pm 
| URL
|
| glennmcc>
| See what happens here.
|
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 4:54:16 pm 
| URL
|
| glennmcc>
| Great news.
|
| With my fresh compile containing your code.
|
| No crashes. :))
|
| Those links that go past the right by just a little bit get highlighted
| but the highlighting stops just short of the scrollbar.
|
Š| Those links that are way too long do not get highlighted at-all.
|
|
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 6:55:37 pm 
|
| ray>
| Ok, that's cool then. I don't like to add more range checking code
| than necessary.
|
| As to the other bug, tell me what you know and I'll charge into it.
|
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 8:40:36 pm 
|
| glennmcc>
| I think I got it.
|
| Been testing it and so far it's working perfectly in all of my tests.
|
| It needs more testing to confirm that it keeps working correctly. <g>
|
| ----- in parsecss.c --------
|
| if(objmap[TAG_SPECIAL_A_HOVER])
| {
| sheet->usehover=1;
| sheet->hoversetbits|=(setbits&(UNDERLINE|ITALIC));
| sheet->hoverresetbits|=(resetbits&(UNDERLINE|ITALIC));
| if(text){sheet->hoverR=textR;sheet->hoverG=textG;sheet->hoverB=textB;}
| //!!glennmcc: Mar 05, 2005
| //default to a contast from both link color & background color
| else
| {
| sheet->hoverR=(sheet->backR + sheet->linkR) / 2;
| sheet->hoverG=(sheet->backG + sheet->linkG) / 2;
| sheet->hoverB=(sheet->backB + sheet->linkB) / 2;
| }
| //!!glennmcc: end
| }
|
| _____________________________
|
|
|
| ============================================================================
|
| Sat, Mar 05, 2005 - 8:47:58 pm 
|
| ray>
| That was fast! I won't touch it until you confirm you've got it.
Š| (A real fix beats my hack any day ;-)
|
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 10:39:54 am 
|
| ray>
| What does this stuff in http.c do:
|
| "User-agent: xChaos_Arachne/4.%s"
|
| I've seen the "User-agent" string inside emails, but I've never
| seen it followed by "XChaos ..." as it is above.
| And specifcaly, what is the "/4." for?
|
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 1:33:07 pm 
|
| glennmcc>
| The '/4' was inserted in our UserAgent string several years ago to 'get arou
| those web sites which redirect us to a "Upgrade your Browser" page
| instead of giving us the page we wanted when theye detect that we are
| not using at-least a version 4.0 or higher browser.
|
| BRB with the full 'User Agent string'.
|
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 1:36:56 pm 
|
| glennmcc>
| This is what the current release sends to almost every web site we visit
| to identify what browser we are using.
|
| User_Agent -- xChaos_Arachne/4.1.81;GPL,386+ (DOS x86;WATTCP/1.05;
| 800x600,HiColor; www.arachne.cz)
|
|
| Amnd this what my current testing version identifies itself as. ;-)
|
| User_Agent -- xChaos_Arachne/4.1.83;UE07 (DOS x86;WATTCP/1.05; 800x600,HiCol
| www.arachne.cz)
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 1:50:27 pm 
|
| glennmcc>
| BTW,
| My 'fix' up above only fixed the problem of black being used by default when
| a.hover is in a CSS page but {color: colorname or number} is missing
Š| from that line.
|
| It does not fix the problem of the specified color changeing.
| (seemingly at random)
|
| Further diagnoses showed me that it is not random.
|
| Upon first visiting and of thos CSS test pages on my site, the link coor
| is correct.
|
| If we take a link to a new file that then shows up in the right or
| bottom frame and mouse-over any of the links which is now using a
| diferent color that our last frame was using.... when we go back to the
| last frame the color has now changed to match the color of that second
| framed page containing a different a.hover {color specification.
|
| A precise example.......
|
| visit testcssf.htm and all of the links in the right frame are correctly
| changed from blue to light blue.
|
| Now link to about.htm and mouse/over any the links which will then
| correctly change from green to yellow.
|
| Now go back in history so that help.htm is again displayed if the rigjht
| frame. All of links now change from blue to yellow when we mouse/over
| them.
|
| Further strangeness........ if our mouse happens to be already over one
| of those links when we go back to help.htm in the right frame....
| it is correct in color for the change at light blue.
| Move the mouse off and it correctly changes back to blue.
| Move back to the same link or to any other link and it uses yellow
| from about.htm
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 1:52:57 pm 
|
| glennmcc>
| If I come here, go to testcssf3.htm, come back here and go back to testcss3.
|
| The links now use red from this page instead of their correct spec of light 
|
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 1:55:48 pm 
|
| glennmcc>
| The incorrect color only happens when the page in question is within a frame
|
| Right-clicking on the page so that it shows by tself instread of within
| the frameset shows correct color change.
|
Š| Back in history to the frameset and they are back to using correct color. 
|
| Back here and forward to the frameset and the use red from here again.
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 2:01:47 pm 
|
| glennmcc>
| We _can_ go to a page which does not contain CSS a.hover and trhen come
| back and the color is OK.
|
| So what's happening is the a.hover color from the last rendered page is
| being used.
|
| When we right click to show just the page and not in a framset, it gets
| re-rendered and coming back to it in the frameset now shows the correct
| color.
|
| So, to fix this problem we are going to need to store the a.hover color
| for each page within a frameset in a different variable for each page.
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 2:03:43 pm 
|
| glennmcc>
| I never have liked frames and this is a further example of why. <ggggg>
|
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 3:57:30 pm 
|
| ray>
| Thanks for info on "/4."
|
| re: CSS, just off the top of my head, why not try this: take your
| fix and insert it into the same place in onmouse.c where I put
| my little hack -- that seems to be the right location, so if you add
| the right fix, she's done.
|
| the right fix
|
| ============================================================================
|
| Sun, Mar 06, 2005 - 4:00:57 pm 
|
| ray>
| I figure that should work since the hover color changes far more often
| than the one time a page is loaded, so the fix has to be in onmouse
| which is used to handling such things as colors changing off and on 
| and off and on.
|
|
Š| ============================================================================
|
| Sun, Mar 06, 2005 - 8:45:34 pm 
| URL
|
| glennmcc>
| Squashed another bug. :))
|
| http://www.cisnet.com/glennmcc/testing_&_symbol/&-amp-bug.htm
|
| All 7 links now work perfectly.
|
|
| ============================================================================
|
| Mon, Mar 07, 2005 - 2:39:02 pm 
|
| ray>
| give it over ;-)
|
|
| ============================================================================
|
| Mon, Mar 07, 2005 - 2:55:40 pm 
|
| glennmcc>
| No problem, dude.
|
| --- in entity.c ---
|
| void entity2str(char *str)
| {
| int i=0,j=0,l=strlen(str);
| char *ptr;
|
| while(i<l)
| {
| //!!glennmcc: Mar 06, 2005 -- fix '&' and '&' in URL bug
| //http://www.cisnet.com/glennmcc/testing_&_symbol/&-amp-bug.htm
| if(str[i]=='&' && !(i<l && str[i+1]==' ') &&
| (str[i+2]==';' || str[i+3]==';' || str[i+4]==';' ||
| str[i+5]==';' || str[i+6]==';')
| )
| // if(str[i]=='&' && !(i<l && str[i+1]==' '))//original line
| //!!glennmcc: end
| ___________________
|
| ============================================================================
|
| Mon, Mar 07, 2005 - 2:57:57 pm 
|
| glennmcc>
| Oops. forgot, the & bug still exists in THIS page.
|
| Hey Gregy, when ya gonna fix that ? <g>
Š|
| --- in entity.c ---
|
| void entity2str(char *str)
| {
| int i=0,j=0,l=strlen(str);
| char *ptr;
|
| while(i<l)
| {
| //!!glennmcc: Mar 06, 2005 -- fix '&' and '& amp;' in URL bug
| //http://www.cisnet.com/glennmcc/testing_& _symbol/& -amp-bug.htm
| if(str[i]=='& ' & & !(i<l & & str[i+1]==' ') & &
| (str[i+2]==';' || str[i+3]==';' || str[i+4]==';' ||
| str[i+5]==';' || str[i+6]==';')
| )
| // if(str[i]=='& ' & & !(i<l & & str[i+1]==' '))//original line
| //!!glennmcc: end
| ___________________
|
|
| ============================================================================
|
| Mon, Mar 07, 2005 - 3:08:57 pm 
|
| glennmcc>
| Hmmm,
|
| Now that I'm looking at it more closely, that change in entity.c also
| fixed the display here.
|
| It just looked odd due to not being formated correctly.
|
| Does the first posting look OK on your end ?
|
| ============================================================================
|
| Mon, Mar 07, 2005 - 3:12:51 pm 
|
| glennmcc>
| Nevermind
|
| The only thing that does not didplay correctly in that first
| post is the & amp; on the first line. (same in both 1.81 and 1.83)
|
| ============================================================================
|
| Mon, Mar 07, 2005 - 6:00:34 pm 
|
| ray>
| Oh THAT bug ... I thought you meant the hover color bug. 
| Happy to have this one too though :-)
| Please send latest on hover color bug, I can't wait to see this
| thing killed dead since, as I said before, this is my first
| try at working with low level video routines (I can feel the power :-)
Š|
|
|
|
|
| ============================================================================
|
| Mon, Mar 07, 2005 - 6:31:46 pm 
|
| glennmcc>
| Oh, that one ?
|
| I haven't done any more on it.
|
| Figured I'd let you use the power instead of just feeling it. <gggg>
|
| ============================================================================
|
| Mon, Mar 07, 2005 - 9:44:38 pm 
|
| ray>
| Ah, well then, would you be so kind as to repost your color fix?
| It's now scrolled away into the void without my having captured it.
| I must say, that was a cute algorithm you cooked up for insuring
| a unique color for the highlight ... which makes me think: so where
| was the bug? IIRC, what you wrote was more like a default mechanism.
| Do I take it then that when a hover color is specified, that it is
| acted upon?
|
|
| ============================================================================
|
| Tue, Mar 08, 2005 - 5:09:35 pm 
|
| glennmcc>
| Yes, when the hover color is specified... it does get used.
|
| When it was not specified.... black was being used.
|
| Now when it's not specified, the code calculates a color which isa
| contrast of BGcolor and link color by simply adding link color to
| BGcolor and then dividing by 2.
|
| BRB
|
|
|
|
| ============================================================================
|
| Tue, Mar 08, 2005 - 5:11:15 pm 
|
| glennmcc>
| ----- in parsecss.c --------
|
Š| if(objmap[TAG_SPECIAL_A_HOVER])
| {
| sheet->usehover=1;
| sheet->hoversetbits|=(setbits&(UNDERLINE|ITALIC));
| sheet->hoverresetbits|=(resetbits&(UNDERLINE|ITALIC));
| if(text){sheet->hoverR=textR;sheet->hoverG=textG;sheet->hoverB=textB;}
| //!!glennmcc: Mar 05, 2005
| //default to a contast from both link color & background color
| else
| {
| sheet->hoverR=(sheet->backR + sheet->linkR) / 2;
| sheet->hoverG=(sheet->backG + sheet->linkG) / 2;
| sheet->hoverB=(sheet->backB + sheet->linkB) / 2;
| }
| //!!glennmcc: end
| }
| _____________________________
|
| ============================================================================
|
| Tue, Mar 08, 2005 - 7:51:46 pm 
|
| ray>
| Got it, thanks. Believe it or not, I was dreaming about your
| little algorithm last night!
|
|
| ============================================================================
|
| Tue, Mar 08, 2005 - 7:53:14 pm 
|
| ray>
| Oh, one more thing, can you please repost that file which we were using
| to check for links that went out of bounds on the right? I'd like to
| fool around with that some more. This time I'll save it.
|
|
|
| ============================================================================
|
| Tue, Mar 08, 2005 - 9:55:48 pm 
| URL
|
| glennmcc>
| Al of them are in the 'my-stuff' dir.
|
| http://www.cisnet.com/my-stuff/
|
| ============================================================================
|
| Tue, Mar 08, 2005 - 9:56:29 pm 
| URL
|
| glennmcc>
| Forgot my own name again <g>
Š|
| http://www.cisnet.com/glennmcc/my-stuff/
|
| ============================================================================
|
| Tue, Mar 08, 2005 - 10:24:47 pm 
|
| ray>
| Ok, its the third css test I'm looking for.
|
| BTW, I'm having trouble with the color fix. It looks to me like
| that code is only active when a hover color is set anyway, thus
| obviating the need for a default. Looking at help.htm, I found that
| the hover color was rendered acurately at all times but when I 
| disabled hover color, I got no hover color at all inspite of you code,
| and adding a 'beep' there confirmed that this code isn't called
| unless a hover color is set. Wasgoinon brutha?
|
|
|
| ============================================================================
|
| Tue, Mar 08, 2005 - 10:29:55 pm 
|
| ray>
| Aaaa wait a minute, spoke too soon, sheesh, it depends whether you've
| clicked in the frame or not -- in frame, color is fine, out of frame
| color is ... why should it be different anyway?
|
| ============================================================================
|
| Thu, Mar 10, 2005 - 8:10:13 pm 
|
| ray>
| Man, you're indestructable.
|
| That suff I showed you yesterday? It works nice -- gives a default
| hover everywhere, not just with CSS, but there's a flipping problem:
| in INBOX, or any page where a right click gives a yellow background,
| the yellow disapears once the mouse is off it :( I've tried to find
| the code that handles this yellow background but I can't. Do you
| know where it is?
|
|
| ============================================================================
|
| Fri, Mar 11, 2005 - 7:32:24 pm 
|
| glennmcc>
| Near the top of onmouse.c .............
|
| if(click) //copy to clipboard
| {
| if(click==2 && atomptr->type==IMG)
| {
Š| struct picinfo *imgptr=(struct picinfo *)ie_getswap(atomptr->ptr);
| GLOBAL.postdata=0;
| arachne.target=0;
| return imgptr->URL;
| }
| if(click==1 && atomptr->type==TEXT && atomptr->linkptr==IE_NULL)
| {
| ptr=(char *)ie_getswap(atomptr->ptr);
| if(*ptr)
| {
| outs(ptr);
|
| ie_appendclip(ptr);
|
| GLOBAL.clipdel=0;
| memcpy(ae,atomptr,sizeof(struct HTMLrecord));
|
| highlightatom(ae);
| return NULL;
| }
| }
| }
|
| And then further down.
| (just below where we made the changes for aligning the CSS linkhover stuff)
|
|
| if(click) //copy to clipboard
| {
| if(click==2 && atomptr->type==IMG)
| {
| struct picinfo *imgptr=(struct picinfo *)ie_getswap(atomptr->ptr);
| GLOBAL.postdata=0;
| arachne.target=0;
| return imgptr->URL;
| }
| if(click==1 && atomptr->type==TEXT && atomptr->linkptr==IE_NULL)
| {
| ptr=(char *)ie_getswap(atomptr->ptr);
| if(*ptr)
| {
| outs(ptr);
|
| ie_appendclip(ptr);
|
| GLOBAL.clipdel=0;
| memcpy(ae,atomptr,sizeof(struct HTMLrecord));
|
| highlightatom(ae);
| return NULL;
| }
| }
| }
|
| And here's the function itself over in highligh.c
Š| ( x_setfill(0,14); // black text on yellow BG)
|
|
| void highlightatom(struct HTMLrecord *foundatom)
| {
| struct HTMLframe *frame=&(p->htmlframe[p->activeframe]);
| // struct TMPframedata *htmldata=&(p->tmpframedata[p->activeframe]);
| mouseoff();
|
| x_setfill(0,14);
| {
| int x1=frame->scroll.xtop+foundatom->x-frame->posX;
| int x2=frame->scroll.xtop+foundatom->xx-frame->posX;
| if(x1<frame->scroll.xtop)
| x1=frame->scroll.xtop;
| if(x2>frame->scroll.xtop+frame->scroll.xsize)
| x2=frame->scroll.xtop+frame->scroll.xsize;
| x_bar(x1,(int)(frame->scroll.ytop+foundatom->y-frame->posY),
| x2,(int)(frame->scroll.ytop+foundatom->yy-frame->posY));
| }
|
|
|
| ============================================================================
|
| Fri, Mar 11, 2005 - 9:12:43 pm 
|
| ray>
| Thanks. 
|
| ============================================================================
|
| Fri, Mar 11, 2005 - 9:18:00 pm 
|
| ray>
| And now, today's intractable mystery: So we've got the hover
| working OK ... oh that reminds me, I made one more change, must remember
| to post it ... now, when you scroll (like with testcss3.htm) such that
| the link moves off the screen, if you 'hover' it the text will
| skew very slightly left or right. I figured out that what is happening
| is that the library function that writes text always truncates at
| a whole character such that the remaining line is 'pulled' more or less
| to the left depending on the proportional width of the character.
| This is probably curable, since 'normal' links scroll off the page
| seamlessly. So how can we get hovered links to do the same??
|
|
| ============================================================================
|
| Fri, Mar 11, 2005 - 9:22:15 pm 
|
| ray>
| ²
| x1 = 0; ²
| //x2 = AonM.xx + FRAME.xtop; ²
Š|
| // NEW:
| // This allows *visible* part of over long link to hover. ²
| x2 = p->htscrn_xsize + p->htscrn_xtop; ²
| ²
|
| y1 = (int)(AonM.y + FRAME.ytop - p->htmlframe[atomptr->frameID].posY); ²
| y2 = (int)(AonM.yy + FRAME.ytop - p->htmlframe[atomptr->frameID].posY); ²
| ²
| // Guard against intrusion into the status line when there is no scrollbar. 
| if (y2 > p->htscrn_ysize + p->htscrn_ytop) ²
| y2 = p->htscrn_ysize + p->htscrn_ytop; ²
| ²
|
|
| ============================================================================
|
| Fri, Mar 11, 2005 - 9:23:37 pm 
|
| ray>
| What are those little '2's ?
|
|
| ============================================================================
|
| Fri, Mar 11, 2005 - 9:24:26 pm 
|
| ray>
| Ah, just unerased border characters.
|
|
| ============================================================================
|
| Fri, Mar 11, 2005 - 9:52:55 pm 
|
| glennmcc>
| Just now recompiled with your new x2= line
|
| Excellent..... it works like a charm.
|
| Thank you. :)
|
| Now to fix that 'skew bug' ;-)
|
|
| ============================================================================
|
| Fri, Mar 11, 2005 - 10:41:28 pm 
|
| glennmcc>
| Found a bug in the x2= code.
|
| CSS linkhover now does not work on http://www.cisnet.com/glennmcc/
|
| Works fine on http://www.cisnet.com/glennmcc/ar181gpl/
Š|
| Also works fine on http://www.cisnet.com/glennmcc/ if we edit-out
| <basefont size="4"> from the page.
|
|
|
| ============================================================================
|
| Fri, Mar 11, 2005 - 11:04:19 pm 
|
| glennmcc>
| Here's the fix.
|
| Use your new line only when we are in a frame. 
|
| When not in a frame, use your original line.
|
| //Ray: Mar 11, 2005
| // This allows *visible* part of over long link to hover.
| if(FRAME.xtop>p->htscrn_xtop)
| x2 = p->htscrn_xsize + p->htscrn_xtop;
| else
| x2 = atomonmouse.xx + FRAME.xtop;
| //Ray: end
|
|
|
| ============================================================================
|
| Sat, Mar 12, 2005 - 7:38:57 am 
|
| ray>
| Right, mustn't do that outside a frame.
| As to the skew bug, fixing that would be nice, since it looks so
| bloody amaturish. I figure on finding the call to drawatom() that
| handles non hover links (which scroll smooth) and steal the 
| coordinate system used.
|
|
| ============================================================================
|
| Sat, Mar 12, 2005 - 7:41:03 am 
|
| ray>
| OH, regarding this FTP problem, IIRC I once used a site of yours to
| practice uploading. If Gregy doesn't give me a site, can I borrow
| yours again? ... "junk1" "junk2" etc. etc.
|
|
| ============================================================================
|
| Sat, Mar 12, 2005 - 8:14:14 am 
|
| ray>
| Just tried that new code, seems to work, but I'm suspicious.
Š| Frames or no frames, why should changing the x2 point make any
| difference to a link that doesn't overlap anything any time?
| My guess is that if you take your homepage, and make a frame out of
| it for test purposes, that things will become clearer.
|
|
| ============================================================================
|
| Sat, Mar 12, 2005 - 3:18:28 pm 
|
| glennmcc>
| testcssf4.htm
|
| Now that links don't change color when the page containing
| <basefont size="4"> is in a frameset.
|
|
| ============================================================================
|
| Sat, Mar 12, 2005 - 4:50:09 pm 
|
| ray>
| Yech, that look horrible. The left frame is mostly black, ditto with
| 1.81
|
|
| ============================================================================
|
| Sat, Mar 12, 2005 - 4:53:05 pm 
|
| ray>
| So what gives with this basefont size=4? How about some other sizes?
|
| BTW, I've go an idea for making those cache exclusions work ;-)
|
|
| ============================================================================
|
| Sat, Mar 12, 2005 - 4:55:12 pm 
|
| ray>
| Hey, even your homepage is screwed up.
|
|
|
| ============================================================================
|
| Sat, Mar 12, 2005 - 5:03:09 pm 
|
| glennmcc>
| It's not just <basefont size="4">.... <font size="4"> is also a problem
| with your new x2= line compiled in.
|
| size="1" 2 and 3 are fine.
|
Š| >3 and the linkhover no longer changes color.
|
| ============================================================================
|
| Sun, Mar 13, 2005 - 3:18:51 pm 
|
| glennmcc>
| F2 FTP upload info
|
| ftp.tripod.com
| UN == arachne-gpl
| PW == scotty
|
| ray/ subdir created and ready for testing.
|
| ============================================================================
|
| Sun, Mar 13, 2005 - 6:21:37 pm 
|
| ray>
| Hmmm, >3 is busted ... that makes me think that something
| else is amiss, not the hover code per se. Always somethin else :(
|
| Re FTP, thanks for that, with any luck some junk is comming.
|
|
|
| ============================================================================
|
| Sun, Mar 13, 2005 - 6:32:01 pm 
|
| ray>
| nuts. "ftp://arachne-gpl:scotty@ftp.tripod.com:21/"
|
| don't work with 1.79, get "530 login error". What am I doing wrong?
|
|
| ============================================================================
|
| Sun, Mar 13, 2005 - 6:39:07 pm 
|
| glennmcc>
| Yep, I thought better of having accessible to anyone who might visit
| here and see the UN/PW I left up above.
|
| So I logged-in and changed the password.
|
| I'll email the new PW to you.
|
| BRB
|
| ============================================================================
|
| Sun, Mar 13, 2005 - 6:41:05 pm 
|
Š| glennmcc>
| You've got mail ;-)
|
|
| ============================================================================
|
| Sun, Mar 13, 2005 - 7:05:34 pm 
|
| glennmcc>
| Found the fix just a few lines further down in onmouse.c
| (simply remove '2 *' ...... ploblem solved)
|
| sz = (long)((x2 - x1 + 1) * (y2 - y1 + 1) + 4 * sizeof(int));
| if (sz > 0 && sz < MAXHOVER)
| //if (sz > 0 && 2 * sz < MAXHOVER)
|
|
|
| ============================================================================
|
| Sun, Mar 13, 2005 - 10:15:07 pm 
|
| ray>
| Yup, that got me on, but alas, no 'junk1.txt' got uploaded.
| For the 100th time, what do I enter into the 'filename' box?
| just the filename or do I need a path or what? No, don't answer that,
| I've got the answer in my own docs some where and I should go and
| find it out myself.
|
| And, is that the solution for the >3 bug? ... shit, I suppose it
| must be, that's great. Why do you suppose it was done that way? --
| seems like your fix is the natural way.
|
| Bed time.
|
|
| ============================================================================
|
| Mon, Mar 14, 2005 - 9:20:40 am 
|
| glennmcc>
| FTP upload.......
|
| F2 while viewing junk1.txt
|
| Server: ftp.tripod.com
| Username: arachne-gpl
| Password: ********
| Filename: ray/junk1.txt
| X Use binary mode
|
| Hit Upload button
|
|
|
Š| ============================================================================
|
| Mon, Mar 14, 2005 - 9:24:24 am 
| URL
|
| glennmcc>
| Just now tested it by uploading my hotlist.... worked fine.
|
| ============================================================================
|
| Mon, Mar 14, 2005 - 4:39:48 pm 
|
| ray>
| Yup, mine works too. I used "/ray/junk1.txt" yesterday. Why ... WHY?
| can't I remember this stuff? IMHO, the FTP box should have a link to
| "the complete idiots guide to Arachne FTP" so that you don't have
| to spend the rest of your life repeating this stuff.
|
| ============================================================================
|
| Mon, Mar 14, 2005 - 5:23:51 pm 
| URL
|
| glennmcc>
| Do you perhaps mean something like this ??? <ggggggg>
|
|
| ============================================================================
|
| Mon, Mar 14, 2005 - 5:58:17 pm 
|
| ray>
| Ha ha 
|
| ============================================================================
|
| Mon, Mar 14, 2005 - 6:57:32 pm 
|
| glennmcc>
| Yep, that one is of-coarse just a joke between friends.
|
| But seriously... what do think of putting something like those example
| fields right on saveas.ah ?
|
| Do you think it would help ?
|
| We could make it more like this.........
|
| Hit the F2 key while viewing somefile.txt
|
| Server: ftp.provider.com
| Username: your_username_for_this_FTP_site -- Password:
| your_password_for_this_FTP_site
| Filename: destination_path/destination_filename.txt
| X Use binary mode
Š| [enabled for most situations, otherwise some files will be
| converted from DOS format to UNIX format by the FTP server]
|
| Hit the Upload button and away it goes.
|
|
| ============================================================================
|
| Mon, Mar 14, 2005 - 9:05:26 pm 
|
| ray>
| I'd have a 'help' button that links to a file. And since I'm the
| ultimate Arachne FTP idiot, I should write it; gime a few days.
|
|
|
| ============================================================================
|
| Wed, Mar 16, 2005 - 9:50:03 pm 
|
| ray>
| Glenn, your site still looks fouled up, what gives?
|
|
| ============================================================================
|
| Thu, Mar 17, 2005 - 4:59:09 pm 
|
| glennmcc>
| Fouled-up in what way ?
|
| Using which version of core.exe ?
|
| ============================================================================
|
| Thu, Mar 17, 2005 - 5:03:05 pm 
| URL
|
| glennmcc>
| No problem with my v1.85
|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 12:41:43 pm 
|
| ray>
| The wallpaper is missing in horizontal bands intersperced with bands
| where the wallpaper is ok. The text is all fine. Sometimes the 
| wallpaper is replaced with black, sometimes with grey. It's the same
| with 1.79, 1.81 and my core. Also, when the page first shows, it's
| fine, but then something else loads, and the problem happens.
| Any ideas?
|
|
Š|
| ============================================================================
|
| Fri, Mar 18, 2005 - 12:45:37 pm 
|
| ray>
| Oh, BTW, the reason I was there was to confirm you fix of the 
| oversized hover, nice work. Now all that remains is to figure out why
| things skew when lines are off the page to the left.
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 1:16:20 pm 
|
| glennmcc>
| As you can see from my screen capture.
|
| I have no such problem with the 'wallpaper' as you call it.
|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 1:22:15 pm 
|
| glennmcc>
| Wait a second........ 
|
| I'll bet you are referring to <hr noshade>
|
| Those are supposed to be there...... they are my separators between
| the various subjects of each section.
|
| Does this look more like you would expect to see if it weren't for my
| <hr noshade> separators ?
|
| http://www.cisnet.com/glennmcc/no-sep.htm
|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 1:29:23 pm 
|
| glennmcc>
| BTW,
|
| This message board (and mine for that matter), use the same separator
| between each post.
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 3:50:41 pm 
|
| ray>
| Nope, it's not those lines, it's the ... not wallpaper, what do you call
| it, the background. I'll send a capture ... 
Š|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 4:08:36 pm 
|
| glennmcc>
| The background is simply http://www.cisnet.com/glennmcc/images/paper.gif
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 4:43:04 pm 
|
| glennmcc>
| Just got your screen cap.
|
| Something is totally FUBAR on your end.
|
| As you see in this one,
| http://www.cisnet.com/glennmcc/images/no-prblm.gif
|
| There's nothing wrong here (or with the page itself)
|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 4:48:20 pm 
| URL
|
| glennmcc>
| That does look pretty kewl tho ;-)
|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 6:15:43 pm 
|
| ray>
| Yeah, obviously something here, but what? Maybe I'll try the virgin
| 1.81 distro ...
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 6:30:54 pm 
|
| ray>
| Nuts, it's not my core -- works fine in the 1.81 virgin distro directory
| so what else could it be? Nothing in mime I hope. Some config option?
| Let's check that ...
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 6:36:39 pm 
|
| ray>
Š| Nope, not something in arachne.cfg
| Sheesh, it's just a page rendering, entirely internal to core.exe no?
| So what outside of core could be doing this?
|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 6:41:11 pm 
|
| glennmcc>
| I don't understand you logic.
|
| If it works fine in the 'virgin 1.81 distro' but is FUBAR in your directory.
|
| The problem _is_ your core.exe......... right ?
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 6:43:41 pm 
|
| glennmcc>
| Or do you mean that you placed your core.exe into an 'otherwise virgin'
| directory ?
|
| If so, there are only 2 files which come into play other than core.exe
| arachne.cfg and mime.cfg
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 6:44:19 pm 
|
| ray>
| Nope, my core imported into 1.81 virgin is OK, and your 1.81 core
| has the same problem as my core in this directory, so it ain't in core.
|
| But I just killed my cache and all is well -- I think it gets too big
| to handle.
|
|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 6:48:44 pm 
|
| glennmcc>
| Oops, forgot one...... arachne.pck
|
| I'll bet ya that's it.
|
| If the 2 copies of arachne.pck are for 2 different vid setups
| one of 'em is gonna be FUBAR
|
|
|
| ============================================================================
Š|
| Fri, Mar 18, 2005 - 6:51:07 pm 
|
| glennmcc>
| The cache _can't_ hget too big to handle.
|
| It can only contain 256 items.
|
| When #257 is saved... #1 gets deleted.
| When #258 is saved... #2 gets deleted.
| etc, etc, etc....
|
|
|
|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 6:53:35 pm 
|
| glennmcc>
| Now, if your disk itself is too full and there's not enough room to
| store the files themselves..... that's another situation all-together.
|
|
| ============================================================================
|
| Fri, Mar 18, 2005 - 9:51:35 pm 
|
| ray>
| Hmmm, I didn't know that -- 256 max eh?
| But I like your .pck theory. While I was working on the hover thing
| I was going back and forth to your site, and changing resolutions like
| an Italian parlament at the same time, so maybe the wallpaper got
| stored at a different rez. or some such thing. Gota remember to
| restore a known good .pck as a first recovery step when trouble strikes.
|
|
| ============================================================================
|
| Mon, Mar 21, 2005 - 5:24:28 pm 
|
| ray>
| Hey Glenn,
|
| Is there any one place in the code that is always passed for any
| POP3 or SMTP whether you need to dial or not?
|
|
| ============================================================================
|
| Mon, Mar 21, 2005 - 5:59:35 pm 
|
| glennmcc>
| I'm not sure what you mean.
Š|
|
| ============================================================================
|
| Mon, Mar 21, 2005 - 8:36:31 pm 
|
| ray>
| I'm working on this little mail indicator I talked about on the list.
| The trick is knowing when to update the display. There is probably
| an elegant way/place to poll for any change to the status of the
| inbox/outbox, but so far I'm just polling after any 'guievent' --
| clumsy, but it works -- except for after a mail upload or download.
| In that case the inbox/outbox are checked right after the initiating
| keystroke, but before any mail has been moved, so the display is
| inacurate (just until the next keystroke or mouse click, but still ...)
| So I'm wondering if there is some place in the code to stick a 
| call to the little function that would capture all mail events.
|
|
|
| ============================================================================
|
| Mon, Mar 21, 2005 - 8:38:15 pm 
|
| ray>
| Oh, no need to mention it, but if this interests you, I'll post the
| code, it's simple enough and it will work in your sources.
|
|
| ============================================================================
|
| Mon, Mar 21, 2005 - 11:06:18 pm 
|
| glennmcc>
| Well, of-course I wanna see it. <gggg>
|
| I'll have a look at the SRCs visa vi a different place to call it.
|
|
| ============================================================================
|
| Tue, Mar 22, 2005 - 3:57:34 pm 
|
| ray>
| Ok, let me just do as much as I myself can for a day or two, and I'll
| show you what I've got.
|
| ============================================================================
|
| Tue, Mar 22, 2005 - 6:02:32 pm 
|
| ray>
| Ah, think I've got it -- very crude but seems to cover all the bases.
|
|
Š| ============================================================================
|
| Tue, Mar 22, 2005 - 6:32:22 pm 
|
| glennmcc>
| So.... you gonna give it up ?
|
| Or is it 'not ready for prime time' just yet ? <g>
|
|
| ============================================================================
|
| Tue, Mar 22, 2005 - 8:55:46 pm 
|
| ray>
| Yeah, sent it to AD an hour or two ago. Find a better place for the
| update test, and I'll love you forever ;-)
|
|
| ============================================================================
|
| Tue, Mar 22, 2005 - 10:12:23 pm 
|
| glennmcc>
| How often do you want it to be updated ?
|
| Or should I ask... Under what circumstances do you want DrawMailFlag();
| to be called again ?
|
| ============================================================================
|
| Wed, Mar 23, 2005 - 7:15:42 am 
|
| ray>
| Whenever it 'needs' it, i.e. when a first CNM is added, last CNM
| removed, and ditto for TBS. What I have now does no harm by checking
| too often, but it would be more elegant to only check after
| mail creations/deletions/movements or after pop3 or smtp, but I tried
| that and never managed to cover all the bases.
|
|
|
|
| ============================================================================
|
| Sat, Mar 26, 2005 - 6:43:32 pm 
|
| ray>
| Glenn,
|
| I find that the newer line in mime.cfg:
|
| file/printprn.dgi |[100]TYPE _*.txt>PRN\ntype _>prn
|
| ... doesn't work, but the older line is fine:
Š|
| file/printprn.dgi |[100]TYPE _4prt.txt > prn.
|
|
| ============================================================================
|
| Tue, Mar 29, 2005 - 5:30:29 pm 
|
| glennnmcc>
| Does it work OK with command.com but not with 4dos ?
|
| ============================================================================
|
| Wed, Mar 30, 2005 - 6:07:41 pm 
|
| ray>
| I'm always running A under command.com now via my filthy little trick.
|
| BTW, Michal wrote this cute little custom cursor maker, want a look?
| You 'kin make any shape of cursor you want, within the 16X16 grid.
|
|
|
|
| ============================================================================
|
| Wed, Mar 30, 2005 - 7:32:05 pm 
|
| glennmcc>
| Yes, please....
|
| I'd love to have that. :)
|
| ============================================================================
|
| Wed, Mar 30, 2005 - 9:55:04 pm 
|
| ray>
| OK, I'll upload it to Yahoo manana. (BTW, what I really want for my
| birthday is a 'Marvin the Martian' cursor ;-))).
|
|
|
| ============================================================================
|
| Thu, Mar 31, 2005 - 6:51:02 pm 
| URL
|
| glennmcc>
| So...... you like Marvin...... do ya ???? <gggggg>
|
| http://www.gargaro.com/marvin.html
|
|
| ============================================================================
Š|
| Thu, Mar 31, 2005 - 9:08:46 pm 
|
| ray>
| Thanks. Nuts, we need the internet just for sites like that.
| Michal says that when he was a kid, they had W.B. cartoons in Poland,
| but no Marvin -- I guess the rotten commies censored him out.
|
| Bastards.
|
|
|
| ============================================================================