DOS Coders1 Message Board 
______________________________________________________________________________

,-----------------------------------------------------------------------------
|
| Arachne DOS Coders1 Message Board                          |        |  |
|                                                            |        |  |
| [ For the Operational discussions of the Arachne C Coders  |        |  |
| ]                                                          |        |  |
| -----------------------------------------------------------+--------+--+
|
|
| ============================================================================
|
| Thu, Jul 01, 2004 - 5:58:26 am 
|
| phantom>
| FYI ...
|
| I may be out of contact for the next few days - I've come
| down with a horrible 'flu ... :-(
|
|
| ============================================================================
|
| Thu, Jul 01, 2004 - 8:37:52 am 
|
| ray>
| Well, I suppose if you're too week to sit at your computer 
| you can have one or two days off <g>
|
|
| ============================================================================
|
| Fri, Jul 02, 2004 - 9:37:58 pm 
|
| ray>
| Joe,
|
| How's that nasty plague treating you?
|
|
|
| ============================================================================
|
| Fri, Jul 02, 2004 - 9:40:42 pm 
|
| ray>
| Hmmmm, this could be a break:
|
| Trying to send the above msg. I got my usual 4 characters ending
| with a '!' error. I went to errase them before sending and got
| a MALLOCERR pointing to line 310 of onmouse.c ... maybe that's the
| locus of our problem? I'll post the line next:
|
|
| ============================================================================
|
| Fri, Jul 02, 2004 - 9:44:59 pm 
|
| ray>
| ... in context, the line is:
|
| atomptr = (struct HTMLrecord *)IEgetSwap(currentHTMLatom);
| if (!atomptr)
| MALLOCERR();
|
| if (currentHTMLatom == p->lastonscr)
| nextHTMLatom = IE_NULL;
|
|
| ============================================================================
|
| Sat, Jul 03, 2004 - 12:11:14 pm 
|
| mht>
| and maybe not... We can suspect that, when we can see the usual
| garbage in IE, the some of the swap 'row' pointers are FUBAR and
| anything can happen at any place where the swap is used...
| BTW, as the config file is kept in memory also in IE swap, the
| source of smiley faces may be the same.
|
| Anyway, the above piece of code may be worth looking at.
|
| But currently, I noticed that BMP cutting and scrolling is FUBAR
| as well, and have to look into GIF routines to figure out how it
| was intended to work.
|
|
| ============================================================================
|
| Sat, Jul 03, 2004 - 1:16:56 pm 
|
| ray>
| Michal,
|
| I have move the translations for the IE modules, and already I can
| see how complex all this stuff is. When I get the garbage
| characters,
| it is always on a new line, and I can see by the comments that there
| are maybe 5 or 6 different tests for newlines in the code --
| plenty of places for bugs to hide.
|
| ============================================================================
|
| Sat, Jul 03, 2004 - 1:20:11 pm 
|
| ray>
| BTW, I've done the translations for all of Haro's files, I you think
| they could be of use to you, let me know and I'll send them along.
|
| And, I have a puzzle that I think you'll like:
|
| You have 12 identical weights exept that one of them is very
| slightly
| either too heavy or too light. Using an ordinary beam balance no
| more than three times, find the defective weight.
|
|
| ============================================================================
|
| Sat, Jul 03, 2004 - 2:16:37 pm 
|
| glennmcc>
| Betcha' I can do it by using the scale just once. ;-)
|
|
| ============================================================================
|
| Sat, Jul 03, 2004 - 2:31:23 pm 
|
| mht>
| Ok, please send them. They should be a bit easier to work with,
| even if I understand more or less quite many of the Czech comments.
|
| I solved the puzzle with Muslims, but I wasn't sure if it was the
| best possible solution (it was quite long). Now I cannot find the
| piece of paper where I wrote it.
|
| That one isn't difficult, as long as you are only interested in
| finding the defective weight and don't need to know whether it is
| too heavy or too light. Is this the case?
|
|
| ============================================================================
|
| Sat, Jul 03, 2004 - 3:02:21 pm 
|
| ray>
| Betcha CAN'T ....
|
| Michal,
|
| OK, there on the way ... now it I could just get Joe's code, I'd
| be ready to release the whole show, and appart from some odds and
| ends I'd be finished :)
|
| It seems to me that the interesting thing about the three Muslim
| couples is that the number of round trips needed is only one or two
| (I can't recall right now) greater than if anyone could go with
| anyone.
|
| As for the 12 weights, I found that one quite hard, myself. But 
| it seems that once you've found the weight, it is usually possible
| to
| tell if it is heavy or light with no extra trouble, but not always.
|
|
| ============================================================================
|
| Sat, Jul 03, 2004 - 4:51:03 pm 
|
| mht>
| I reproduced my solution with Muslim pairs. The numbers are 11 and
| 9.
| As for the 12 weights, the keyword is 'usually'. 
|
| After looking a bit into HDRAWGIF.C, I have an impression that
| HDRAWBMP.C has been created (as many, many other sections of Arachne
| code) using "cut, paste and modify (and screw up a bit)" method
| instead of "find similarities and structurize". There are many
| similar
| pieces of code in both files, and while I found much of the sloppy
| coding known to me from HDRAWBMP.C in HDRAWGIF.C as well, some
| details
| in HDRAWGIF.C look as after my early attempts to optimize
| HDRAWBMP.C.
| It looks that some changes partially break the scroll/cut
| functionality with BMP images, unfortunately. Well, I never liked
| this part of HDRAWBMP.C, it's high time to re-write it ;-)
|
|
|
| ============================================================================
|
| Sat, Jul 03, 2004 - 6:44:56 pm 
|
| ray>
| Yes, 9, and 11, its is -- but it is interesting that only one
| extra round trip is needed, no?
|
| As for the weights, as you say, it depends on whether or not the
| beam balances on the first weigh, no? I like that puzzle because
| it is asymetrical -- in theory 27 discriminations are possible, and
| only 24 answers are required (H | L * 12 weights) but in practice,
| the
| mechanics are such that one has 'too much' information at some
| times,
| and 'too little' at other times to give an answer to the heavy/light
| question every time. BTW, it is possible with 13 weights, too, but
| with that number, it gives too much of a hint what the first
| division should be.
|
| As to the code, my guess is that you're going to have to rework 
| all of Haro's stuff sooner or latter ;-(
|
| And...
|
| I've done all the translating except for a very few Czech comments
| that Christof missed (99.9% acuracy ain't bad!) and he hasn't yet
| translated 'putikonw.c' and 'x_lopif.h' (the horror!) -- once those
| are done and incorporated it's done ;-)
|
|
| ============================================================================
|
| Sun, Jul 04, 2004 - 4:00:36 am 
|
| mht>
| As to the Muslims -- yes, interesting. Such a complicated rule and
| only
| two times two people have to go back. It would be interesting to
| prove
| whether there is a solution for N > 3 pairs, and if so, find a
| formula
| for the number of trips needed for N pairs. Anyone feels fed up with
| computers? <G>
|
| As to the weights, after writing the solution on a piece of paper,
| I can see that you can always determine if the defective weight is
| heavier or lighter (but with 13 weights, not always). My thought
| solution was sloppy. And, generally, the solutions of this kind of
| puzzles are mostly asymmetrical -- at least there is no simple
| binary
| symmetry.
|
| As to the code... well, all that is mess... Insight as well, as we
| could see recently...
|
| I got your message, so please expect some comments within a few
| days ;-)
|
|
| ============================================================================
|
| Sun, Jul 04, 2004 - 4:52:59 pm 
|
| ray>
| Have any of you ever wondered what 'roura' means -- as in
| '$roura$.bat'??
| It means 'pipe'. What say we rename it '$pipe$.bat' and let the
| light shine brighter?
|
|
|
|
| ============================================================================
|
| Sun, Jul 04, 2004 - 5:29:30 pm 
| URL
|
| little_birdie>
| ;-)
|
| ============================================================================
|
| Sun, Jul 04, 2004 - 5:30:48 pm 
|
| little_birdie>
| Why not $tube$.bat or $duct$.bat ? <gggggggggggggggggggg> ;-)
|
| ============================================================================
|
| Sun, Jul 04, 2004 - 5:58:17 pm 
|
| ray>
| (nice dictionary!)
|
| ...
|
| Becuz Michal sez!
|
|
|
|
| ============================================================================
|
| Sun, Jul 04, 2004 - 9:09:53 pm 
|
| ray>
| Help me, Obewan Kenobe,
|
| There is a keyword "Toolbar" that seems to be used nowwhere except
| that
| it is refered to in 'keywords.txt' (*nice* file BTW -- the sort of
| thing I like to rant about having -- embarasing when I *do* have it
| but don't know I do) ... anyway 'Toolbar' is supposed to point to
| 'toobar.cfg' which is a file that seems to exist nowwhere. Is this
| something that's obsolete?
|
|
| ============================================================================
|
| Mon, Jul 05, 2004 - 7:09:09 am 
|
| ray>
| Now I get it ... sort of. Toolbar.tb is a config file. So why isn't
| in called toolbar.cfg? And why is there code in init.c that offers
| to
| read the non existent 'Toolbar' config variable so as to let you
| choose a differnt toolbar if you want? We should either implement
| 'Toolbar' or cut the code that refers to it and make 'toolbar.tb'
| standard, no? I'll bet there's simmilar stuff happening with
| 'mime.cfg'.
|
|
| ============================================================================
|
| Mon, Jul 05, 2004 - 5:01:32 pm 
|
| little_birdie>
| --- arachne_main\system\gui\edittb.ah ---
|
| <TITLE>Toolbar Editor - advanced.tb</TITLE>
| <BODY NORESIZE>
| <FORM ACTION="arachne:internal-config?arachne:restart">
| <INPUT TYPE=HIDDEN NAME="Toolbar" ARACHNECHECKED="Toolbar"
| VALUE="toolbar.tb">
|
| <!--
| Use
| <INPUT TYPE=RADIO NAME="Toolbar" ARACHNECHECKED="Toolbar"
| VALUE="toolbar.tb">
| this toolbar definition file <B>toolbar.tb</B>
| <INPUT TYPE=RADIO NAME="Toolbar" ARACHNECHECKED="Toolbar"
| VALUE="oldstyle.tb">
| traditional toolbar <B>oldstyle.tb</B>
|
|
| -->
|
| <TEXTAREA ROWS=35 ACTIVE ARACHNEROWS COLS=74
| ARACHNEEDITOR="toolbar.tb"></TEXTAREA>
| <TABLE>
| <TD>
| <INPUT TYPE=SUBMIT VALUE="Save toolbar & restart Arachne [F2]">
| <TD>
| <FORM ACTION="arachne:back">
| <INPUT TYPE=SUBMIT VALUE="Cancel">
| </TABLE>
| ___________________
|
|
| ============================================================================
|
| Mon, Jul 05, 2004 - 6:27:35 pm 
|
| ray>
| So we have a toolbar editor then. Nuts, I modified mine all by trial
| and error by hand. Well that answers that except for the 
| name. Why not 'toolbar.cfg' just as it says in 'keywords.txt'?
|
|
| ============================================================================
|
| Mon, Jul 05, 2004 - 6:30:16 pm 
|
| ray>
| Just looked at that edittb.ah not too much of an improvement over
| doing it by hand, but still nice.
|
| ============================================================================
|
| Mon, Jul 05, 2004 - 6:57:32 pm 
|
| little_birdie>
| This is a small typo that CAV made.
|
| "Toolbar" [Path+Filename] to your toolbar.cfg
|
| It _should read_ ............
|
| "Toolbar" [Path+Filename] to your toolbar.tb
|
|
|
|
| ============================================================================
|
| Mon, Jul 05, 2004 - 10:14:05 pm 
|
| ray>
| That's one solution, but isn't 'toolbar.cfg' the right answer?
| Arachne is full of these tiny, niggling iritations that cause so
| much more confusion that they are worth. Why don't we fix 'em??
| It's your call, of course, but IMHO, consistency is a virtue.
|
| arachne.cfg, mime.cfg and toolbar.??? ;-/
|
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 12:05:59 am 
|
| little_birdie>
| OK, shall we rename tham all then to be more 'standard' ???
|
| No more .TB, it shall now be .CFG
| No more .HTP, they will now be .HTM
| No more .HTT, they will now also be .HTM
| No more .AH, they will also now be .HTM
| No more .IKN, they will now be .ICO
| No more .ACF, they will become .CFG
| No more .ASF, they will also now be .SCR
|
|
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 12:11:22 am 
|
| little_birdie>
| Do you get my point ?
|
| IMO, you are now proposing changes simply for the sake of change.
|
| If it ain't broke..... don't fix it.
|
| It it _is_ broke...... start figuring-out how to fix it.
|
| Throwing-in this cahnge, and that change and another change does
| absolutely nothing in the way of making this a better program.
|
| All it does is to make it 'fit the mould' of what YOU feel is best
| for YOUR needs.
|
| If that's all you are interested in..... have at it.
|
| But you can for damned sure count me out of it.
|
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 2:59:20 am 
|
| mht>
| Ok, let's analyze the above proposals without emotions.
|
| TB -> CFG is reasonable, because both files have similar functions
| and both are text files. More consistency is less confusion and more
| user-friendliness. 
|
| ACF -> CFG is reasonable as well, all .ACF files are just versions
| of ARACHNE.CFG. (The opposite, ARACHNE.CFG -> ARACHNE.ACF, is
| reasonable, too, but less standard and more cryptic.)
|
| IKN -> ICO makes no sense at all. ICO is well-established format of
| Windows and OS/2 icon bitmaps, while IKN is a different proprietary
| image format by IBASE & xChaos.
|
| HTP -> HTM makes little sense. HTP templates are not ordinary HTM
| files,
| they contain special $...$ tokens replaced by Insight with
| appropriate
| values when they are processed.
|
| AH -> HTM makes little sense as well, they contain a lot of
| Arachne-specific stuff.
|
| ASF -> SCR: maybe, but better avoid confusion with Netware login
| scripts
| and Windows screen savers.
|
| HTT -> HTM: why, they are special files and exist in cache only,
| anyway.
|
| And yet another proposal:
|
| COOKIES.LST -> COOKIES.IDX, because after Joe's modifications it is
| no longer a plain text file. HISTORY.LST still is.
|
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 4:19:24 pm 
|
| ray>
| I agree with Michal completely. It's probably too small an issue
| to get into a fuss about, but I don't understand your going over
| the top
| about this sugestion, Glenn. toolbar.tb **IS** a config file of
| exactly 
| the same type as arachne.cfg and mime.cfg -- it is read the same 
| way by the same function and follows the same rules. But for the
| different extension, I would have figured that out much sooner.
| As for the others you mention, with them, the 'different' extension
| tells us something useful and there is no confusion, so changing
| them
| (except as Michal suggests) would be foolish.
|
| Why fight this? Anyway, it's such a small thing that perhaps the
| best reason not to change it is simple backwards compatability.
| Either way, it's not much of an issue if you ask me.
|
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 4:33:22 pm 
|
| little_birdie>
| </bad mood over>
|
| <concession mode>
| toolbar.cfg it is then
| </concession mode>
|
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 4:59:29 pm 
|
| ray>
| Glenn,
|
| Well, I've long since given up trying to understand you, but thanks
| ;-)
|
| And:
|
| This is a change that has consequences outside of core itself, so
| it's up to you to make sure that it's documented in the next 
| release, Capn'.
|
|
|
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 5:33:46 pm 
| URL
|
| glennmcc>
| Done.
|
| arachne_main\toolbar.tb has now been renamed as toolbar.cfg
|
| --- in init.c ---
| //---toolbar.cfg loading
| ptr=configvariable(A,"Toolbar",NULL);
| if(!ptr)
| #ifdef POSIX
| strcpy(TOOLBARcfg.filename,"toolbar.cfg");
| #else
| sprintf(TOOLBARcfg.filename,"%stoolbar.cfg",exepath);
| #endif
| else
| strcpy(TOOLBARcfg.filename,ptr);
| TOOLBARcfg.killcomment=1;
| rc=ie_openf_lim(T,CONTEXT_SYSTEM,LINES); //Toolbar
| if(rc==2)
| memerr0();
| else if(rc!=1 || TOOLBARcfg.lines==0)
| cfgerr(T);
|
| _________________
|
|
| --- in arachne_main\system\gui\edittb.ah ---
| <TITLE>Toolbar Editor - advanced.cfg</TITLE>
| <BODY NORESIZE>
| <FORM ACTION="arachne:internal-config?arachne:restart">
| <INPUT TYPE=HIDDEN NAME="Toolbar" ARACHNECHECKED="Toolbar"
| VALUE="toolbar.cfg">
| <!--Use
| <INPUT TYPE=RADIO NAME="Toolbar" ARACHNECHECKED="Toolbar"
| VALUE="toolbar.cfg">
| this toolbar definition file <B>toolbar.cfg</B>
| <INPUT TYPE=RADIO NAME="Toolbar" ARACHNECHECKED="Toolbar"
| VALUE="oldstyle.cfg">
| traditional toolbar <B>oldstyle.cfg</B>
|
|
| -->
| <TEXTAREA ROWS=35 ACTIVE ARACHNEROWS COLS=74
| ARACHNEEDITOR="toolbar.cfg"></TEXTAREA>
| <TABLE>
| <TD>
| <INPUT TYPE=SUBMIT VALUE="Save toolbar & restart Arachne [F2]">
| <TD>
| <FORM ACTION="arachne:back">
| <INPUT TYPE=SUBMIT VALUE="Cancel">
| </TABLE>
| ____________________________________________
|
| http://www.cisnet.com/glennmcc/ar177gpl/changes.txt
| will be updated in just a few minutes
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 5:36:51 pm 
|
| ray>
| OK, I'll mirror it.
|
|
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 5:42:57 pm 
|
| glennmcc>
| Upload done.
|
| July 06, 2004 
| Toolbar.tb will now be toolbar.cfg 
| (all needed changes made in init.c and edittb.ah) 
|
|
| ============================================================================
|
| Tue, Jul 06, 2004 - 8:22:43 pm 
|
| ray>
| Movin' right along :-)
|
| Michal has me playing with some cool stuff like '_fastcall' first
| efforts gave Dos memory (KB) 170[+] -- it crashed, but never mind,
| Michal will figure it out.
|
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 4:07:52 pm 
|
| mht>
| Instead of changing "toolbar.tb" to "toolbar.cfg" in every place,
| what about
|
| #define TOOLBAR_CFGFILE "toolbar.cfg"
|
| or something like that in, say, ARACHNE.H?
|
|
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 4:53:00 pm 
|
| Mr. Clean>
| Good idea, especialy if we do it for all of the string litterals
| that point to the various files Arachne neeeds.
| BTW, in B4 it would be in 'defines.h' ... check it out.
|
|
| Glenn,
|
| Since you are in such a cooperative mood <G G G>,
| would we be pushing our luck asking for "$pipe$.bat" ?
|
| (I thought so) ;-)
|
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 6:19:54 pm 
|
| mht>
| When speaking about $pipe$.bat, I'd get rid of these dollars (spend
| them? <G>). Why? Dollar signs are often used in reserved names of
| installable device drivers. [And I know there are some drivers that
| provide *nix-like pipe functionality in DOS network environment,
| although I have no idea what names they use]. Anyway, $pipe$ sounds
| as asking for trouble (the extension doesn't matter).
|
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 7:33:08 pm 
|
| glennmcc>
| Michal,
|
| re:"Instead of changing "toolbar.tb" to "toolbar.cfg" in every
| place,"
|
| It only needed to be changed on these 3 lines
| (and the 1st is just a comment)
|
| //---toolbar.cfg loading
| strcpy(TOOLBARcfg.filename,"toolbar.cfg");
| sprintf(TOOLBARcfg.filename,"%stoolbar.cfg",exepath);
|
| And since the 1st line of 'active code' is only used... #ifdef POSIX
|
| It actually required changing just one line of Dos code.
|
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 7:36:23 pm 
|
| glennmcc>
| Mr. Clean,
|
| Personally, I agree with Michal
|
| 'pipe.bat' seems like a reasonable change to implement.
|
| BRB
|
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 8:01:41 pm 
| URL
|
| glennmcc>
| Done.
|
| --a--- 443,752 7-07-04 9:48p l:\arachne\core.exe
| --a--- 65 7-07-04 9:55p l:\temp\arachne.tmp\pipe.bat
|
| --- clipped from fresh upload of .....
| http://www.cisnet.com/glennmcc/ar177gpl/changes.txt ---
|
| July 07, 2004
| $roura$.bat will now be pipe.bat
| (all needed changes made in the SRC code files and 'companion'
| files)
| _____________
|
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 8:15:11 pm 
|
| ray>
| Hey, that was too easy!!
|
|
|
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 8:41:35 pm 
|
| glennmcc>
| Oh yeah......
|
| While I was at it..... $roura2.bat is now pipe2.bat in the SRC code.
|
| ---------- e:main.c
| [741] strcat(mman,"pipe2.bat");
| [794] strcat(mman,"pipe2.bat");
| ______________
|
|
| In addition, every occurence of the word 'roura' has been changed
| to 'pipe'
|
| --- find /n "roura" *.* ---- (done before the changes)
|
| ---------- e:arachne.h
| [395]void rouraname(char *fname);
|
| ---------- e:config.c
| [314]//. Roura : name of a batch file. Allow us to use external
| executable. Like Pipe in UNIX
| [315]void rouraname(char *fname)
| [318] strcat(fname,"$roura$.bat");
| [321]//create a batch file executing cmd , uses name created by
| rouraname( ..)
| [327] rouraname(bat);
| [334] return EXIT_FAST_EXECUTE; //%ARACHNETEMP%\$roura$.bat
| [336] return EXIT_EXECUTE; //.\$roura$.bat
|
| ---------- e:main.c
| [741] strcat(mman,"$roura2.bat");
| [794] strcat(mman,"$roura2.bat");
|
| ---------- e:urlovrl.c
| [489] //in $roura$.bat (means "pipe"), and if the image is already
| not
| [557] rouraname(command);
|
|
| ___________
|
|
| --- find /n "roura" *.* ---- (done after the changes)
|
| ________________________
|
|
| Yep, that's right.... not found anywhere... in any of the files. :)
|
|
| And yes....
| All needed changes have also been made to the SRC for insight.exe
|
| A brand-new insight.zip with the SRC code and precompiled
| insight.exe
| has been uploaed to ...... http://www.cisnet.com/glennmcc/ar175gpl/
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 9:06:16 pm 
|
| ray>
| Marvelous! If this keeps up much longer we'll have a team here.
|
| NBI: (That's "Next Brilliant Idea" incase you were wondering ;-)
|
| So now we have configuration options for tool bar, history list,
| cache index, and cookie file. What *don't*
| we have a configuration option for ... yup 'mime.cfg', and that's
| the
| one we need more than any of the others. Glenn, remember how you
| did(do?) offer two seperate downloads who's only major difference
| is that one uses 'mime lite' and the other uses 'mime industrial'?
|
| All we gota do is add a keyword like "MimeFile" and the rest is
| a piece of cake -- just pick the mime of your choice. The coding
| would take two minutes.
|
|
|
| ============================================================================
|
| Wed, Jul 07, 2004 - 9:36:16 pm 
|
| glennmcc>
| Is this what you mean ?
|
| -- add this line to mime.cfg --
|
| file/.mcf |COPY $1 MIME.CFG
|
| _______________________________
|
| There you go.
|
| Link to any .MCF file and it will be copied over top of mime.cfg
|
|
|
|
| ============================================================================
|
| Thu, Jul 08, 2004 - 7:25:35 am 
|
| ray>
| That's powerful, too, but I meant a keyword in 'arachne.cfg' that
| would let you choose between different mime.cfg's non-destructively.
| That way newbies could start with "MimeFile mimelite.cfg" in
| arachne.cfg
| and progress to "MimeFile mimefull.cfg" as needed with one
| config file change. I think it would be very useful to have several
| mime config files to choose between. Two minutes work.
|
|
| ============================================================================
|
| Thu, Jul 08, 2004 - 7:11:31 pm 
|
| glennmcc>
| 2 min work... huh ?
|
| OK, stopwatch in hand.
|
| Ready, set.... GO !
|
|
|
| ============================================================================
|
| Thu, Jul 08, 2004 - 7:13:36 pm 
|
| glennmcc>
| Time's up........ you lose. <g g g g g>
|
|
| ============================================================================
|
| Thu, Jul 08, 2004 - 7:41:04 pm 
|
| ray>
| Hmmmm, this could be fun.
|
| For a proper test, I've got to make the change and test it, which
| means modifying arachne.cfg and making two different versions of
| mime.cfg and testing the loading of them both.
|
| .... OK how about 1/2 hour ;-)
|
|
|
| ============================================================================
|
| Thu, Jul 08, 2004 - 10:08:15 pm 
|
| ray>
| Glenn,
|
| Screwed things up royaly trying to implement the 'pipe.bat' thing
| (and it seemed like such a good idea at the time). Can you give
| me a list of files involved? Obviously there's more to it than
| meets the eye. core.exe, arachne.bat, and I suspect insight.exe
| are involved but I'd be grateful for any help you can give.
|
|
|
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 3:20:20 am 
|
| Joe>
| Michal,
|
| While it is true that "strange" characters like '$' are often used
| in
| device names, you should be aware that it is a DOS convention to use
| the '$' character in file names that are temporary. Hence
| '$roura$.bat'
| or '$pipe$.bat' comply with this convention. Even if there is a
| device
| driver whose name is '$pipe$', why would this conflict with a batch
| file of this name? AFAIK, there is no problem with this. Therefore,
| I
| would prefer to continue to follow the "temporary file name"
| convention
| and avoid the chance that a user already has a batch (or executable)
| file called 'pipe.*'.
|
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 3:24:26 am 
|
| Joe>
| Glenn,
|
| I like your idea of a *.MCF scheme, although we have a convention
| that
| special Arachne files have a '*.A*' extension, so perhaps *.AMC or
| *.AMF
| would be appropriate?
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 3:29:20 am 
|
| Joe>
| Ray,
|
| I've started integrating my changes into B3a, although it's harder
| than
| I expected. I keep getting muddled with mismatched brackets (or is
| it
| braces? I always get them mixed-up;-), making a mess of things and
| restarting from scratch. Perhaps this will improve with practice ...
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 7:29:06 am 
|
| ray>
| Joe,
|
| Good thinking on all points.
|
| Yes, importing across editions isn't much fun, is it?
| When I do it, I use three monitors :-(
|
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 6:06:56 pm 
|
| mht>
| Joe,
|
| You are right, there may be a conflict between pipe.bat and some
| other executable named 'pipe', but pipe.bat is either called with
| full path (it is written to temp directory) or it can be called
| as .\pipe.bat (this should be changed, see arachne.bat).
|
| But, there would be a conflict between $pipe$.bat and \dev\$pipe$.
| Please try, e.g., "copy somefile.txt somepath\con.txt" (or replace
| con.txt with any devname.txt).
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 7:25:35 pm 
|
| ray>
| Hmmmm, very high level stuff, so I'll keep my nose out of it 
| except for one very small point: From a janitors perspective
| anything that makes the name of the file 'exotic' is good since the
| word 'pipe' is liable to occur in many places besides 'pipe.bat';
| thus I'd be happiest with anything like '~pipe~.bat' or '$$pipe.bat'
| or anything else that makes a unique search string -- but it's 
| a very small point, not worth bothering about if there are more
| pressing reasons to choose the name.
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 7:26:26 pm 
|
| ray>
| Glenn, where are you when I need you?
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 7:40:43 pm 
|
| glennmcc>
| Michal and I are in agreement on this point.
|
| Joe,
|
| I have aleady removed the dollar signs and I'm afraid it's going to
| take
| a lot better arguement than that for me to put them back. <g g g g>
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 7:46:25 pm 
|
| glennmcc>
| Michal,
|
| Which section of arachne.bat are you refering to ?
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 7:54:28 pm 
|
| glennmcc>
| Ray,
|
| These are the files in which to change $roura$.bat to pipe.bat
|
| --a--- 2,261 7-07-04 10:18p k:\arachne\arachne.bat
| --a--- 10,531 7-07-04 10:16p k:\arachne\changes.txt
| --a--- 443,752 7-07-04 9:48p k:\arachne\core.exe
| --a--- 83,856 7-07-04 10:37p k:\arachne\insight.exe
| --a--- 18,068 7-09-04 9:44p k:\arachne\mime.cfg
| --a--- 11,020 7-07-04 9:51p k:\arachne\doc\dir.htm
| --a--- 567 7-07-04 9:51p k:\arachne\system\dgi\clr.bat
| total files 7 total bytes 570,055
| disk free space 813,301,760
|
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 7:56:14 pm 
|
| glennmcc>
| In addition....
|
| $roura2.bat and $roura3.bat must be changed to
| pipe2.bat and pipe3.bat
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 8:07:26 pm 
|
| ray>
| Thanks Glenn,
|
| To tell the truth I made one hell of a mess -- had to restore
| my whole arachne directory after doing an outrageous SRon the 
| whole thing, including .exe's 
|
| BTW, in arachne.bat won't we need to be careful that whereas:
|
| "del $roura?.bat"
|
| will kill both $roura$.bat and $roura2.bat,
|
| "del pipe?.bat" 
|
| will not kill 'pipe.bat' but only 'pipe2.bat'
|
| and we should use:
|
| "del pipe*.bat"
|
| instead?
|
| Also, while we're renaming things howbout renaming
|
| RouraName() to PipeName() ... seems like a no-brainer.
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 8:18:46 pm 
|
| ray>
| Huh ... just tried that delete thing and it looks like I'm wrong:
| "del pipe?.bat" will kill both. And to think that for the last
| 30 years I thought the '?' would only match an existing character.
|
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 8:52:18 pm 
|
| glennmcc>
| re: "RouraName() to PipeName() ... "
|
| Beat ya to it..... <g>
|
|
| --- from last evenings' post ----
|
| In addition, every occurence of the word 'roura' has been changed
| to 'pipe'
|
| --- find /n "roura" *.* ---- (done after the changes)
| ________________________
|
| Yep, that's right.... not found anywhere... in any of the files. :)
|
|
| ============================================================================
|
| Fri, Jul 09, 2004 - 9:26:44 pm 
|
| ray>
| Hmmmm, not bad janitoring, I must say.
|
|
| ============================================================================
|
| Sat, Jul 10, 2004 - 7:25:08 am 
|
| phantom>
| Michal,
|
| You are correct. I have CLOCK$ (DR-DOS 6.0) and when I tried to
| rename a file to this, I got the message "file already exists"
| (or words to that effect).
|
| Nevertheless, it *is* a convention to use the '$' character in
| temporary file names, which is why $ROURA$.BAT was used by MP.
|
| AFAIK, device driver names never _start_ with '$'??? If so, then
| $PIPE$.BAT is still OK, just as $ROURA$.BAT is/was.
|
| Anyway, now you know the reason for the '$' character. Also, note
| that although it is not uncommon for device names to use the '$'
| character, _more_ device names seem to just use just alphanumeric
| characters, so we are just as likely to encounter a device driver
| called PIPE as PIPE$, and much more likely than $PIPE$.
|
|
|
| ============================================================================
|
| Sat, Jul 10, 2004 - 7:37:39 am 
|
| ray>
| Joe has the best case here, guys. Besides, isn't it just
| good sence to make the name something that is certain to be
| unique and obviously a temp file? I'm going to hold off
| making this change incase the consensus changes on this.
|
| BTW, just to add more confusion: If 'roura' was replaced with
| another 5 letter term, like 'pipe!' instead of a four letter
| term like 'pipe', it would be possible to do a hot swap including in
| the .exe's with no problems since no string sizes would change, no?
| Just an idea.
|
|
|
| ============================================================================
|
| Sat, Jul 10, 2004 - 3:20:50 pm 
|
| mht>
| About device driver names:
| CLOCK$ exist in any DOS version. DR-DOS 7 Fixed Disk Security module
| appears in MEM listing as $SECURE$ device; I never used "login"
| security
| under DR DOS 6, so I don't know whether it behaves similarly or not.
| I also seem to remember I saw some other driver starting with $ 
| (not including one I once wrote myself). 
|
| Well, too bad that CONFIG.SYS AVAILDEV= directive (causing more
| *nix-like behavior of device names) died with DOS 2.x...
|
| And yes, I agree that $ characters are sometimes used in temporary
| file names. I didn't say that $pipe$ is an error, I just suggested
| some kind of 'defensive programming' which costs zero extra code.
|
| About meaning of '?' wildcard:
| Well, Ray... 30 years ago there was no MS/PC DOS (and even no CP/M
| IIRC),
| so the meaning of '?' on the system you were using at that time
| could
| have been a bit different. <G> 
| But wait... what can I say about it? I am 30 years old! <G G G>
|
|
|
| About ARACHNE.BAT:
| I meant the following section
|
| :execute
| $roura$.bat
| goto err
| :fastexecute
| if not (%ARACHNETEMP%)==() %ARACHNETEMP%\$roura$.bat
| if exist %TEMP%\ARACHNE.TMP\$roura$.bat
| %TEMP%\ARACHNE.TMP\$roura$.bat
| if exist %TEMP%\$roura$.bat %TEMP%\$roura$.bat
|
| etc.
|
| My suggestion to use .\pipe.bat was not necessarily good. This
| wouldn't
| work if Arachne is started from other directory. (Is this possible,
| BTW?)
| Where is the best place fot pipe.bat?
|
|
| ============================================================================
|
| Sat, Jul 10, 2004 - 6:13:54 pm 
|
| ray>
| Glenn,
|
| Am I missing something? That post on the list -- we already have
| 1.75, no?
|
|
|
| ============================================================================
|
| Sat, Jul 10, 2004 - 7:19:43 pm 
|
| glennmcc>
| I am using v1.77;GPL,386+
|
| ============================================================================
|
| Sat, Jul 10, 2004 - 8:44:06 pm 
|
| ray>
| OOOOooooowwwww,
|
| I could have sworn your URL advertised 1.75, anyway off to the
| download site ;-)
|
| ============================================================================
|
| Sat, Jul 10, 2004 - 9:08:22 pm 
|
| glennmcc>
| Yes, it _does_ advertise 1.75
|
| But that's the release.... as my email on May 8th said.............
|
| Well folks,
| It was exactly 6 months ago today that Michael Polák
| gave his 'OK' to release Arachne v1.73;GPL
|
| Today seems as good a time as any to declare that all
| futher developemnet of v1.75;GPL is "officailly closed".
|
| I will now begin working on v1.77;GPL
|
| Please visit my site for more info.
|
| And for what you might call "a very nice surprise" !!! :))))
|
|
| ============================================================================
|
| Sat, Jul 10, 2004 - 9:10:39 pm 
|
| glennmcc>
| Also,
|
| This is 'advertised' on my front page as-well........ <g g g>
|
|
| Coming soon..... Arachne GPL 1.77
|
|
| ============================================================================
|
| Sat, Jul 10, 2004 - 10:36:30 pm 
|
| ray>
| Right.
|
| Joe, any ETA on that code?
| (BTW I case you don't know why I'm so hot to get it, it's because
| I can't use it until I get it worked into my own code because
| of the fact that I need to run my own cores constantly to see
| if I've busted something -- so I'm feeling a little impoverished).
|
|
|
| ============================================================================
|
| Sun, Jul 11, 2004 - 11:31:45 am 
|
| ray>
| All,
|
| I notice that we have defaults for "CachePath" and "MailPath", but
| no
| default for "DownloadPath" -- looks like another little
| thing to fix, no?
|
|
|
| ============================================================================
|
| Sun, Jul 11, 2004 - 7:30:17 pm 
|
| phantom>
| Just an idea ...
|
| What about '!PIPE$.BAT' as a "unique" name? Arachne already uses '!'
| for some file names, so hopefully this really will be unique to
| Arachne.
|
|
| ============================================================================
|
| Sun, Jul 11, 2004 - 10:03:15 pm 
|
| ray>
| I'm happy with anything that can't be confused with anything else
| and '!pipe$.bat' is that, for sure.
|
| But what about my brilliant observation that if the name is kept
| the same length it can be S and R'd everywhere, including in .exe
| files with no need to rebuild? ... come to think of it, so bloody
| what, as if that matters.
|
|
| ============================================================================
|
| Mon, Jul 12, 2004 - 2:12:39 am 
|
| phantom>
| Ray,
|
| In answer to your question ...
|
| Assuming I get more adept with practice, I hope it will be only a
| week or so to integrate the code ...
|
|
|
| ============================================================================
|
| Mon, Jul 12, 2004 - 7:32:17 am 
|
| ray>
| Joe,
|
| OK, I can wait. Or, if you're pressed for time, I could do it, Joe.
| Unpleasant work but not otherwise un-doable, I just assumed you
| could
| do it easier than I could. If you've got bigger fish to fry let me
| know, and I'll do the integration.
|
| BTW I've been working on this 'all configuration opptions parsed in
| one function' project for a few days now on an experimental basis --
| it looks good so far; code size is breaking even and it is so much
| neater and better organized not having to hunt down option parsing
| all over the place, especialy nice is that all default values can
| be scrutinized in one place.
|
|
|
| ============================================================================
|
| Tue, Jul 13, 2004 - 3:36:25 pm 
|
| mht>
| Glenn,
|
| (or anyone else who tried to compile v1.75 using v1.73 sources plus
| CHNGDSRC.ZIP) -- I didn't, I feel I'm doing too many things at the
| same time anyway ;-)
|
| Vladimir, who is working on localized Russian version of Arachne
| (let
| me remind that this is Vladimir who pointed out that there is a
| problem
| with characters above 128, this way allowing me to find the "signed
| vs.
| unsigned char" problem; I hope he will start reading our message
| board
| soon) tried to compile and got:
|
| > "linker error: group DGROUP exceeds 64K"
|
| Have you experienced similar problems? Can you help? Maybe he
| needs a new .PRJ file with different options? (Vladimir isn't
| an experienced C programmer.) 
|
|
| ============================================================================
|
| Tue, Jul 13, 2004 - 5:27:22 pm 
| URL
|
| glennmcc>
| Yep, sounds to me like a corrupted .PRJ
|
| I'll upload nokey.prj to http://www.cisnet.com/glennmcc/my-stuff
| nokeyprj.zip
|
|
|
| ============================================================================
|
| Wed, Jul 14, 2004 - 1:54:03 pm 
|
| mht>
| The problem is that the near data segment is nearly full (close to
| 64 KB). Translated Russian messages are somewhat longer, and the
| maximum size is exceeded. This may happen also for other localized
| versions -- English is known to be a particularly concise language
| ;-)
| It may be the time to think about moving some data (maybe messages?)
| into far data segment(s). Too bad that Borland C cannot (at least
| it's not documented) put constant data into code segments...
|
|
| ============================================================================
|
| Wed, Jul 14, 2004 - 4:06:19 pm 
|
| ray>
| Michal,
|
| Aren't the messages constant data?
|
|
| ============================================================================
|
| Wed, Jul 14, 2004 - 10:25:55 pm 
|
| ray>
| Finished the rough first effort at doing all parsing of arachne.cfg
| in config.c ... seems to be working :-) 
|
|
|
| ============================================================================
|
| Thu, Jul 15, 2004 - 2:07:38 am 
| URL
|
| vbm>
| Michal,
| And how can I find longest messages?
|
| ============================================================================
|
| Thu, Jul 15, 2004 - 3:43:35 am 
|
| mht>
| The problem is that, if we decide to move messages into far data
| segments, we can no longer do
|
| #define SOME_KIND_OF_MSG "Some kind of message"
|
| and then
|
| puts(SOME_KIND_OF_MSG);
|
| or something similar. It seems that we need to do it this way:
|
| const char far SomeKindOfMsg[] = "Some kind of message";
|
| (we can combine the above declaration with #define, of course) and
| then
|
| puts(SomeKindOfMsg);
|
| Maybe there is a better way? What do you say, our C experts? ;-)
|
| A possibility to stuff constant strings into code segment would be,
| of course, ideal (Watcom C can be told to do this, Turbo Pascal does
| it by default). The strings used by an overlaid module would be
| removed from memory every time the module is unloaded then.
| Unfortunately, to achieve this in Borland C, we'd have to put the
| messages into an .ASM file, it seems. [Experts? ;-)]
|
|
| ============================================================================
|
| Thu, Jul 15, 2004 - 6:57:02 am 
|
| ray>
| Welome Vladimir!
|
|
|
| ============================================================================
|
| Fri, Jul 16, 2004 - 3:05:42 am 
|
| Joe>
| Welcome Vladimir.
|
| Michal, I can't check just now, but I think I shrunk the Data
| Segment
| requirements a bit in 1.73J4, perhaps the Russian text would fit in
| that version???
|
|
| ============================================================================
|
| Fri, Jul 16, 2004 - 3:08:15 am 
|
| Joe>
| Ray,
|
| I'm still working on "B3b" - when this is complete, how do we verify
| that I haven't made any mistakes?
|
|
| ============================================================================
|
| Fri, Jul 16, 2004 - 4:22:22 am 
|
| mht>
| Vladimir says that the board is difficult to use for him, because
| his dial-up connection breaks every 3-5 minutes -- to often to stay
| online. He prefers e-mail contact.
|
|
|
| ============================================================================
|
| Fri, Jul 16, 2004 - 4:36:36 pm 
|
| ray>
| Joe,
|
| The hell if I know! It's your code, so only you know when it's
| debugged, but we know it works in B0x, so there's not much reason
| that it shouldn't work in B3x; the only pain in the arse should be
| stuff like coping with function reanames, other than that, it should
| be just a line for line delete/add/change unless you bump into
| one of my trivial changes like a loop converted from 'while' to
| 'for'.
| Unless I'm overlooking something, I'd describe the work as 
| unpleasant but not dangerous. Don't hesitate to contact me if
| there's
| some problem or other.
|
|
| ============================================================================
|
| Fri, Jul 16, 2004 - 6:01:18 pm 
|
| ray>
| All,
|
| I tried the experiment Michal mentioned above -- changing all the
| message defines into far strings. The results (I'm using it now)
| are that '_DATA' segment size drops by 4,058 bytes :-), but the
| .exe size increases by 3,888 bytes :-( and Arachne reports
| Dos memory 165[+] (down about 2K from before). So do we 
| really want to do this? Maybe its what we *have* to do if
| our data segment overflows, but not unless. Perhaps Vladimir
| would be interested?
|
|
|
|
| ============================================================================
|
| Sat, Jul 17, 2004 - 10:32:10 am 
|
| phantom>
| Michal,
|
| It's very sad that Vladimir has such a bad connection! I've never
| heard of such a bad connection, he should try another modem to
| see if the problem is that or his telephone line.
|
| If he prefers e-mail, I'd suggest joining the ArachneDevelopment
| list and using its mailing list feature.
|
|
|
| ============================================================================
|
| Sat, Jul 17, 2004 - 10:35:30 am 
|
| mht>
| Every time a pointer to a far string is passed to a function,
| instead
| of a near string, the code should grows by two bytes (PUSH SEG
| _MyString
| instead of PUSH DS) and the EXE header should grow by four bytes
| (one
| relocation item). The total size of all data segments (all
| _FAR_DATA,
| DATA, _BSS, etc.) should remain unchanged, at least I expected this.
|
| But there is a small catch, I just noticed it while analysing
| compiler's
| output on a small sample program. Every far constant declared in the
| way I proposed goes into its own segment, which is
| paragraph-aligned,
| and therefore its size is rounded up to the next multiply of 16
| bytes.
| This way we lose 7.5 byte per message, on average. But there is a
| way
| to put all far constants into one segment and avoid this loss!
| It is -zEname option. We can put #pragma option -zE_FARMESS before
| all
| const char far MsgXYZ[] = "blah blah blah" and #pragma option -zE*
| after them (not really needed, unless there are other far constants
| declared in this file). All messages go into a segment named
| _FARMESS
| (the name is, of course, arbitrary, even if meaningful *g*) then,
| one
| after another, without any paragraph alignments between them. (The
| rules of art tell us to enclose these #pragma's with #ifdef
| __BORLANDC__
| ... #endif, of course, maybe even to #define the 'FARness' of these
| constants by means of a macro -- this way we wouldn't affect
| portability to other compilers.)
|
| Anyway, Ray, could you explain, how did you change the messages into
| far strings? Are you sure that you declared only one copy of each
| message (some messages are used in more than one source file)?
| Are you sure that you eliminated all unused messages (there are
| some,
| and the linker isn't smart enough to eliminate them itself)? 
|
| The above questions are to ensure that no opportunity to save some
| space has been lost ;-)
|
| Well, do we really need all this 'FAR MESS'?
|
| (1) Let me remind, that the global data segment is nearly full.
| We can easily overflow it by adding some variables needed for some
| cool new features we'll invent for sure. On the other hand, there
| is a possibility that we can optimize usage of the global data
| segment while reviewing the existing code, but nobody can guarantee
| this. As you can see, there is a risk that we'll face the problem
| sooner or later. Let's be prepared.
|
| (2) If we are going to move something into far data segments, then
| let's move something that can be moved without too big impact on
| code
| size. String constants seem to fulfill this condition. From my
| (limited) experiments, the code generated by BC for modifying far
| variables is rather poor -- too much reloading of segment registers,
| many unnecessary reloads even with optimizations enabled.
|
| (3) It is very nice that we have/will have the first localized
| version
| of Arachne-GPL. We should help the translators by making their work
| easier, and not turning them into coders. The code for localized
| versions should remain fully compatible. If the translators have to
| worry about things like data segment overflow and introduce some
| workarounds in their localized version, we'll encourage a
| development
| fork, which would be a Bad Thing (TM) with our limited human
| resources, even if perfectly within the spirit of GPL.
|
| [ArachneDevelopment list is very quiet recently. This post will be
| quoted there, to reach wider audience (including Vladimir), I hope.]
|
|
| ============================================================================
|
| Sat, Jul 17, 2004 - 10:41:36 am 
|
| ray>
| Yes, the list is too quiet, and it's the best place for
| long messages so ...
|
|
| ============================================================================
|
| Sat, Jul 17, 2004 - 10:56:02 am 
|
| phantom>
| Michal,
|
| Great idea, the -zEname option and moving messages together to
| their own segment is the ideal solution.
|
| I'm off to sleep now (I've just come from a wedding - it's very
| late;-). See you later at ArachneDevelopment.
|
|
|
| ============================================================================
|
| Sat, Jul 17, 2004 - 4:34:18 pm 
|
| ray>
| I've got a nice comfy whole 737 bytes of free room in my data
| segment,
| and that's after taking out the garbage since last November ... I
| hate
| to think how tight it must be with B0 ... I think I'll go and see;
| I'll
| bet MP had it stuffed tight.
|
|
| ============================================================================
|
| Sat, Jul 17, 2004 - 11:07:31 pm 
|
| ray>
| ... not as bad as I thought: 375 free in B0.
|
|
| ============================================================================
|
| Tue, Jul 20, 2004 - 2:40:19 am 
|
| vbm>
| I don't compiled 1.73j4, because haven't yet source.
| moded is ok, problems in line. but sometime it work. ;-)
|
|
| ============================================================================
|
| Wed, Jul 21, 2004 - 7:35:57 pm 
|
| ray>
| Greetings, Commrades!
|
| This is being written with a Russian version of Arachne.
|
| ============================================================================
|
| Thu, Jul 22, 2004 - 9:08:22 am 
|
| ray>
| Help!
|
| I'm stuck in Russia without a passport!
|
|
| ============================================================================
|
| Thu, Jul 22, 2004 - 9:27:12 am 
|
| wondering>
| why Michal not have upload dir at Yahoo group
|
|
| ============================================================================
|
| Thu, Jul 22, 2004 - 12:24:54 pm 
|
| ray>
| Hi Vladimir,
|
| Let me know if those files work for you.
|
| ============================================================================
|
| Fri, Jul 23, 2004 - 9:09:21 am 
|
| mht>
| I just don't have anything big enough to upload. Not yet ;-)
|
|
| ============================================================================
|
| Fri, Jul 23, 2004 - 10:17:08 am 
|
| phantom>
| Vladimir,
|
| For 1.73Jx source, download from ArachneDevelopment the
| files "ar173gpl.zip" and "JDS173*.ZIP".
|
| For 1.73J1, unzip "ar173gpl.zip" then "JDS173A.ZIP.
| For 1.73J2, then unzip also "JDS173B.ZIP".
| For 1.73J3, then unzip also "JDS173C.ZIP".
| For 1.73J4, then unzip also "JDS173D.ZIP".
|
| Cheers!
|
|
|
| ============================================================================
|
| Fri, Jul 23, 2004 - 5:09:06 pm 
|
| wondering>
| Michal,
|
| upload insight.exe fixes ?
|
| ============================================================================
|
| Sat, Jul 24, 2004 - 2:45:57 am 
|
| mht>
| INSIGHT -? says "now being maintained by glennmcc" -- I think it's
| ok
| to submit patches, but uploading new binaries should be consulted
| with
| the maintainer, to avoid chaos. Glenn? ;-)
|
| [Hey, I know I'm old-fashioned, but I prefer when people use their
| real names or nicks clearly derived from them.]
|
|
| ============================================================================
|
| Sat, Jul 24, 2004 - 2:39:55 pm 
|
| glennmcc>
| Hi Michal,
|
| I was just "Wondering" why you don't recognize my newest
| 'alter-ego' <g g g g>
|
| Please.... if you would.... email those insight patches to me.
|
| Thank you.
|
|
| ============================================================================
|
| Sat, Jul 24, 2004 - 4:26:29 pm 
|
| mht>
| Sent (to your @cisnet account).
|
|
| ============================================================================
|
| Sat, Jul 24, 2004 - 5:36:17 pm 
|
| glennmcc>
| Thank you very much sir.
|
| --- clipped from http://www.cisnet.com/glennmcc/ar175gpl/---
| SRC code for Insight.exe (link to insight.zip)
| (ZIP file also contains precompiled INSIGHT.EXE v3.85.7) 
| (Updated on July 24, 2004)
| Recompiled to include another 'fix' by Michal H. Tyc 
| v3.85.7 no longer 'garbles' the names of attached files with LFNs 
| _____________________________________________________________
|
| And 2 screencaps that show how well it works :)
|
| http://www.cisnet.com/glennmcc/images/3856.gif
| http://www.cisnet.com/glennmcc/images/3857.gif
|
|
| BTW,
|
| I had to ask for the 'patch' via e,mail because I have my
| membership at
| Yahoogroups setup so that I do not get emails.
|
| Instead, I just read them online like this.....
| http://groups.yahoo.com/group/ArachneDevelopment/message/940
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 3:55:15 am 
|
| mht>
| Glenn,
| please disable COMDEFs in compiler options for Insight. I have no
| idea
| why they were originally enabled. Many variables were placed in
| separate
| far data segments while they could be near. And, of course, as soon
| as
| the default CORE.EXE will be 386+, INSIGHT.EXE should be 386+ as
| well.
|
| insight.exe 83.872 Original settings
| ins_ncd.exe 82.464 Disabled COMDEFs (they result in lots of far
| data)
| ins_opt.exe 82.192 More optimizations
| ins_186.exe 81.152 Target CPU = 80186
| ins_386.exe 79.360 Target CPU = 80386
|
|
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 3:37:31 pm 
|
| glennmcc>
| Here's my latest compile.
|
|
| Volume in drive C is MCCORKLE_01
| Directory of C:\ARACHNE
|
| INSIGHT EXE 83872 7-24-04 7:04p
|
| Same size so it's probably identical to your "original settings"
| compile.
|
| What are your settings to get those "more optimizations" ?
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 3:57:29 pm 
|
| glennmcc>
| Looks like I found most of the correct settings.
| (only 64 bytes larger than yors)
|
| INSIGHT.EXE 79424 7-25-04 5:52p
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 4:48:54 pm 
|
| glennmcc>
| By George.... I think I've got it ;-)
|
| INSIGHT.EXE 79360 7-25-04 6:45p
|
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 4:52:03 pm 
| URL
|
| glennmcc>
| I'll upload an updated ZIP file to the above URL
| (so that INSIGHT.PRJ will have these new settings)
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 5:21:21 pm 
|
| ray>
| Michal, Glenn,
|
| Good to read the last few posts. To tell the truth, I haven't done
| anything with Michal's last fix since I have no idea what's what
| with Insight.exe, and I don't want to get out of synch with Glenn --
| I'll download the latest .exe from him and not worry about the
| sources untill the next release.
|
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 7:36:43 pm 
|
| glennmcc>
| Too late.
|
| If you downloaded my latest .EXE ....... 
|
| You also download the sources. ;-)
|
| The same holds true if you grab the latest wwwman.exe
| (uploaded v1.77 about 5min ago) <g>
|
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 7:56:48 pm 
|
| ray>
| Well, not that it would hurt to have them ;-)
|
| Anyway that reminds me to go and get them.
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 8:46:28 pm 
|
| glennmcc>
| So... what do think of em ? ;-)
|
| ============================================================================
|
| Sun, Jul 25, 2004 - 11:00:46 pm 
|
| ray>
| ... let you know tomorrow, but I doubt I'll be disapointed!
|
|
|
| ============================================================================
|
| Mon, Jul 26, 2004 - 1:26:36 am 
|
| mht>
| I haven't downloaded and checked yet, but under Compiler
| Optimizations
| I set "Fastest code" first, then change "Optimize for speed" to
| "Optimize for size" and disable inlining of intrinsic functions.
| (At least as far as I remember. I don't have BC++ 3.1 on the machine
| I'm using right now.) "No pointer aliasing" can be checked as well,
| but it can be a bit risky.
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 12:28:05 pm 
|
| ray>
| The new insight lookin' good. Of course I had to make that 
| change to 'pipe.bat'. I wanted '$pipe.bat, but so what? Of all the
| stupid little things one could fight over!
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 12:33:02 pm 
|
| ray>
| Glenn,
|
| I've got Greg M running speed trials on my experimental core (the
| one where all config options are parsed once at startup) due to the
| fact that his machine is a 486/33. He says you made him a custom
| core with the right click feature turned off. I thought that since
| he's helping me it would be nice to hack that into the X core just
| for
| him. Would this be much trouble?
|
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 4:22:12 pm 
|
| ray>
| Glenn,
|
| Possible anomaly with the new insight: When Arachne first connects,
| she goes to arachne.cz even if I have pressed the 'Get Mail' button.
| However on the second press of the button, all is normal. I'm quite
| sure I changed all my roura's properly, so that isn't it.
| Can you confirm or refute?
|
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 5:20:00 pm 
|
| glennmcc>
| Nope...... no such occurence here.
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 5:20:49 pm 
|
| glennmcc>
| Now as to the right mouse button.........
|
|
| Add these 2 sections into onmouse.c
|
| 1 line of original code commented-out,
| 5 lines of new code added
| 7 comment lines
|
| (for this posting, I have added "//original code"
| out to the right of each line to leave untouched)
|
| <snip>
|
| int count=0; //original code
|
| //!!glennmcc: Begin May 17, 2004
| //added to optionally goback or not on rightmouse click
| char *rmgb=0;
| //!!glennmcc: end
|
| if(mousey<p->htscrn_ytop && !customerscreen) //original code
|
| <snip>
|
| nolink: //original code
| if(click==2) //original code
| { //original code
| if(p->activeframe) //right click on frame source //original code
| { //original code
| arachne.target=0; //original code
| return p->htmlframe[p->activeframe].cacheitem.URL; //original code
| } //original code
| else if(!ontoolbar) //original code
|
| //!!glennmcc: Begin May 17, 2004 -- only go back in history if
| //'RightMouseGoesBack Yes' is in arachne.cfg
| rmgb=configvariable(A,"RightMouseGoesBack",NULL);
| if(rmgb && toupper(*rmgb)=='Y') return "arachne:back";
| else
| return 0;
| //return "arachne:back"; //original line had no such option
| //!!glennmcc: end
|
| } //original code
| else //original code
| if(click==4 && !ontoolbar) //middle button //original code
| return "arachne:fullscreen"; //original code
|
| <snip>
| ________________
|
| Also add this section into opt_misc.ah
|
| <INPUT TYPE=RADIO NAME="AutoF3key"
| VALUE=Yes ARACHNECHECKED=AutoF3key> http://if online
|
| <!-- begin new section -->
| <TR>
| <TD>RightMouse
| <td>
| <INPUT TYPE=RADIO NAME="RightMouseGoesBack" VALUE="Yes"
| ARACHNECHECKED="RightMouseGoesBack">
| Goes Back
| <INPUT TYPE=RADIO NAME="RightMouseGoesBack" VALUE="No"
| ARACHNECHECKED="RightMouseGoesBack">
| Ignored
|
|
| <!-- end new section -->
|
| </TABLE>
| <HR>
| <B>XMS for Animated GIFs and fonts</B>
|
|
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 5:24:44 pm 
|
| glennmcc>
| As to pipe.bat
|
| Now I can't remember,
|
| Do we charge for it ?
|
| Or do we make it free of charge ?
|
| <gggggggg>
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 6:36:29 pm 
|
| ray>
| Glenn,
|
| Thanks for the code, I'll see what I can do integrating it into
| core X.
|
| The problem with insight is probably due to the new version not
| synching with my changes to config.c, i.e. it seems to be 
| ignoring the DialPage setting and using the default on the first
| press. Actualy, I'm quite happy with the fact that it works at all
| given that I've made so many radical changes to core X. Since I've
| never even looked into insight, might you have any suggestions on
| where and how insight deals with configuration option settings?
|
| As to pipe ... let's just do the right thing and smoke it -- we'll
| all get to like it after a few drags, I'm sure <g g g 
| g man>
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 7:16:13 pm 
|
| glennmcc>
| And I said:
| No no no no, I don't smoke it no more
| I'm tired of waking up on the floor
| No, thank you, please
| It only makes me sneeze
| And then it makes it hard to find the door
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 7:24:36 pm 
|
| glennmcc>
| Now... back to pipe.bat
|
| How about a name that can't possible get confused with something
| that
| might already be in use by the user for something else ?
|
| pipe.bat might already exist as a file written by the user.
|
| $pipe.bat.... ditto
|
| pipe2.bat.... ditto
|
| $pipe2.bat.... ditto
|
| pipe3.bat.... ditto
|
| $pipe3.bat.... ditto
|
| So........... pipe-ara.bat pipe-ar2.bat and pipe-ar3.bat
|
| Not much chance of those already being there. ;-)
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 7:35:13 pm 
|
| glennmcc>
| By increaseing from 4 characters to 8
| with a 36 possibilities for character (0-9 and A-Z)
|
| The posibilies are now increased from ..... 36 * 36 * 36 * 36 =
| 1,679,616
| possible filenames to a whopping
|
| 36 * 36 * 36 * 36 * 36 * 36 * 36 * 36 = 2,821,109,907,456
|
| 2.8 trillion possibilities should be enough for anyone ;-)
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 7:39:40 pm 
|
| glennmcc>
| And I did not even include in the number the various uctuation
| marks we
| can use.... ~ ! # $ & - (and perhaps others)
|
| With just those 6, we are now up to almost 9.7 trillion
|
| 42 * 42 * 42 * 42 * 42 * 42 * 42 * 42 = 9,682,651,996,416
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 8:03:01 pm 
|
| ray>
| Hey, you're abusing your prisoners -- I surrendered, didn't I?
| .... but if you *want* to fight about it, then I suggest
| 'hashpipe.bat' ... I'm sure MP would aprove <G>
|
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 9:11:27 pm 
|
| glennmcc>
| You win again........ hashpipe.bat it is.
|
| I'll send you the modified SRC files as soon as I'm done. :-)
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 9:23:05 pm 
|
| glennmcc>
| That was easy, here you go .........
|
| __________________ for core.exe 
|
|
| ---------- e:config.c
| [318] strcat(fname,"hashpipe.bat");
| [334] return EXIT_FAST_EXECUTE; //%ARACHNETEMP%\hashpipe.bat
| [336] return EXIT_EXECUTE; //.\hashpipe.bat
|
| ---------- e:urlovrl.c
| [489] //in hashpipe.bat, and if the image is already not
|
|
| ---------- e:main.c
| [741] strcat(mman,"hashpip2.bat");
| [794] strcat(mman,"hashpip2.bat");
|
|
| ___________________________ And for insight.exe
|
| ---------- e:convmsg.c
| [437] sprintf(mman,"%shashpip2.bat",cachepath);
| [451] puts("Can't append to hashpip2.bat...");
|
| ______________________
|
|
| <g g g g g g>
|
|
| ============================================================================
|
| Tue, Jul 27, 2004 - 9:29:47 pm 
|
| ray>
| Hmmmm, I vasilate between calling your bluff and saying:
|
| WHY NOT? 
| After all we're all a little bit outrageous already aren't we?
| What with not giving up on DOS and presuming to navigate the
| Internet with machines that 'should' have been in the dump years
| ago if
| Intel had their way and not leting M$ run our lives for us?
|
| It's your call boss ;--))
|
| ============================================================================
|
| Wed, Jul 28, 2004 - 5:27:03 pm 
|
| glennmcc>
| Naaaaw... I was just continueing the joke for while longer. ;-)
|
| I left it as 'pipe' in my copy of the SRC
|
| I'll change to '$pipe' if either Joe or Michal
| agrees with you that dollar sign should go back in.
|
|
| ============================================================================
|
| Wed, Jul 28, 2004 - 6:03:46 pm 
|
| ray>
| OK man, that reeel cool with me. Like I said before, of all the
| stupid things to fight about.
|
| ...
|
| now, while the pot is boiling there is one other file rename I think
| we should consider. As with 'roura' it's a whole lot of trouble
| in the short term but since we're in this for the long term I sez 
| we should to it:
|
| there are ikon files that are icon files (can't remember the exact
| names right now, is it 'alticon2.ikn'? Anyway since we've
| standardized
| on 'ikon' these should be changed, no?
|
|
|
|
| ============================================================================
|
| Wed, Jul 28, 2004 - 6:07:19 pm 
|
| ray>
| BTW, progress report on some of the suff Michal has me doing:
| Converting the functions to '_fastcall' -- it saves memory (I'm up
| to
| [169+] now, but there are bugs -- some of the functions don't
| behave properly if they are made _fastcall. I'm hoping Dr. T will
| be able to figure out why.
|
| ============================================================================
|
| Wed, Jul 28, 2004 - 6:13:01 pm 
|
| glennmcc>
| What say to ..................
|
| ---------- e:guidraw.c
| [190] DrawIconLater( "alticon1",4,3 );
| [191] DrawIconLater( "alticon2",x_maxx()-146,3 );
| [198] DrawIconLater( "alticon2",x_maxx()-146,4 );
| [199] DrawIconLater( "alticon1",x_maxx()-146,27 );
|
| ---------- e:guitick.c
| [77] DrawIconNow( "ALTICON2",x_maxx()-146,y );
|
|
| Change to ...............
|
|
| ---------- e:guidraw.c
| [190] DrawIconLater( "altIkon1",4,3 );
| [191] DrawIconLater( "altIkon2",x_maxx()-146,3 );
| [198] DrawIconLater( "altIkon2",x_maxx()-146,4 );
| [199] DrawIconLater( "altIkon1",x_maxx()-146,27 );
|
| ---------- e:guitick.c
| [77] DrawIconNow( "ALTIkon2",x_maxx()-146,y );
|
|
| ============================================================================
|
| Wed, Jul 28, 2004 - 6:14:46 pm 
|
| glennmcc>
| As to memory............ I'm at 169[+] green right now. :)))
|
|
| ============================================================================
|
| Wed, Jul 28, 2004 - 8:19:25 pm 
|
| ray>
| Glenn, you're getting mellow in your old age! ;--0
|
|
| As to the memory:
| Yes, but you're probably running Joe's core or a variety of it,
| also,
| This is on top of my having moved all configuration option parsing
| into config.c which causes a memory hit since all the string values
| are now stored in memory. (BTW Greg M reports this is somewhat 
| faster, but the results are tenuous).
|
|
| ============================================================================
|
| Wed, Jul 28, 2004 - 8:24:30 pm 
|
| ray>
| BTW, in guitick.c why is it "ALTIkon2" not just plain old
| "altikon2" ?
| I suspect it don't matter either way, but just asking.
| And I suppose you're going to just let me twist in the wind as far
| as why I have to click twice on 'get mail' to get mail?? ;--((
|
|
| ============================================================================
|
| Wed, Jul 28, 2004 - 11:21:56 pm 
|
| ray>
| Aw nuts, all my problem was is that I neglected to check
| 'sub-pppd.bat'
| for rouras, those pesky criters can hide in the strangest places.
| Sorry for trying to throw the problem in your direction, Glenn.
|
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 2:29:23 am 
|
| Joe>
| Glenn, Michal,
|
| Re '$roura$.bat', I originally wanted '$pipe$.bat', but Michal
| thinks some device names might match/clash (although plain
| alphanumeric device names are at least as likely to match/clash).
|
| Therefore, I propose '!pipe$.bat', etc. This is still a unique
| name that is most unlikely to clash with either an existing batch
| file or device name, yet retains the DOS conventions (used eg. in
| '$roura$.bat') of including a '$' character in temp. file names.
|
| Michal (et. al.), are you happy with '!pipe$.bat', etc.?
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 2:33:19 am 
|
| Joe>
| Ray,
|
| Just to let you know, I'm almost finished converting your B3a code
| to B3b (ie. incorporating my changes in the "cleaned" code). (I've
| been very busy with other stuff, so I'm sorry for the time it's
| been taking ...)
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 8:40:57 am 
|
| ray>
| Joe,
|
| Good to hear from you. Looking forward to getting 3b as time
| permits. (I'm having a 'Dos memory (KB) ...' war with Glenn; we're
| tied right now, but after I get your code I'll *smoke* him)
|
| <G G G>
|
| As far as pipe goes, I'm easy but favour your aproach since it
| makes trouble less likely.
|
| HEY! We could add a new keyword "PipeName" and let folks choose
| their
| own! < g g g>
|
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 10:14:52 am 
|
| glennmcc>
| Smoke me ?!?!?!
|
| HA !!!
|
| That'll be the day. <g g g g>
|
| My next goal is 180
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 10:31:45 am 
|
| ray>
| 180 eh? 
|
| You know, in all seriousness, I think we'll do it eventualy. Infact
| I'd say there's a 50/50 chance we'll do it within the year.
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 10:47:37 am 
|
| glennmcc>
| You're way too pesamistic.....<g>
|
| My time-table says "within the next 30 days"
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 2:24:07 pm 
|
| ray>
| Serious? You've found a way to save 10K? 
|
| Most cool if true.
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 2:25:27 pm 
|
| ray>
| Vladimir,
|
| I tried to send you an email but it bounced. I sent to
| 'arachnerus@mail333.com' but the bounce message said something
| about 'arachnerus@be3.int.mail333'
|
| Am I doing something wrong?
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 4:36:34 pm 
|
| glennmcc>
| Fantastic ! :)
|
| I am currently running Arachne v1.77;UE03
|
| Thats' right my code still works perfectly to create either the GPL
| version or the 'full version'.
|
| UE03 recognizes and reads the registration key
|
| Right now, mine shows "G.W.McCorkle" which was read from arachne.key
|
| The point of my little experiment was to see wheather or not we can
| still send our current code to Michael some day.
|
| Who knows... he just might get back into it once he sees that all
| of our
| improvements _can_ be used in 'full version'.
|
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 5:09:08 pm 
|
| glennmcc>
| Hey Ray,
|
| Dos memory (KB) 173[+] green :))))
|
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 5:58:33 pm 
|
| ray>
| ANIMAL!
|
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 6:02:38 pm 
|
| ray>
| (I'm back down to 167 -- these '_fastcall' changes save memory, but
| they generate one hell of alot of bugs. I'm abandoning that
| thread until such times as Michal can tell me what's wrong -- which
| he will of course).
|
|
|
| ============================================================================
|
| Thu, Jul 29, 2004 - 10:14:07 pm 
|
| ray>
| Clawing my way back up to 168 ... this is a dog eat dog contest,
| Glenn. <g>
|
|
| ============================================================================
|
| Fri, Jul 30, 2004 - 3:13:15 am 
|
| mht>
| Joe,
| you are right that plain character device names cause confusion as
| well.
| Too bad that adding a period and extension doesn't help --
| unfortunately, this is a design limitation from DOS 1.x days that
| must
| be kept for compatibility reasons...
| And "!pipe$.bat" seems to be an acceptable name, of course. However,
| the best way to avoid confusion with some user's program stored in
| another directory is to specify full path to our 'pipe' batch file,
| whatever its name is.
|
|
| ============================================================================
|
| Fri, Jul 30, 2004 - 9:36:13 am 
|
| ray>
| Hmmmm, maybe MP left it as 'roura' precicely because that way there
| was virtualy no chance of any confusion ;-\
| Speaking of full paths, another way to make this problem evaporate 
| would be to start work on something we've discussed before and that
| is a rethinking of Arachne's directory structure. For example, if
| we had a dedicated \SCRATCH directory for pipe.bat and all other
| such
| files used only for Arachne's internal workings, there would be no
| chance of a mixup with another program, no? I realize that a
| directory
| rethink is a big project but sooner or later I think we should do
| it.
|
|
|
|
| ============================================================================
|
| Fri, Jul 30, 2004 - 5:14:09 pm 
|
| glennmcc>
| Good point ......
|
| Since there is virtually 0 chance of anyone having a $roura$.bat
| already on their system.
|
| And virtually 0 chance of having a loaded device named $roura$
|
| My vote is to go back to $roura$.bat $roura2.bat and $roura3.bat
|
| As to a dedicated directory, we already do that.
|
| IIRC, this is where it goes.........
|
| %temp%\arachne.tmp\$roura$.bat
|
| If the env variable %arachnetemp% is set........
|
| %arachnetemp%\$roura$.bat
|
| Upon exiting arachne....... all files in either of those dirs are
| 'purged'
| and the dir itself is removed.
|
|
|
| --- clipped from arachne.bat ---
|
| :execute
| pipe.bat
| goto err
| :fastexecute
| if not (%ARACHNETEMP%)==() %ARACHNETEMP%\pipe.bat
| if exist %TEMP%\ARACHNE.TMP\pipe.bat %TEMP%\ARACHNE.TMP\pipe.bat
| if exist %TEMP%\pipe.bat %TEMP%\pipe.bat
| REM Compatibility with old %TEMP% format maintained
| :err
| echo Abnormal Arachne termination. Please reboot system and/or
| restart Arachne.
| :end
| if exist pipe?.bat del pipe?.bat
| if exist %ARACHNETEMP%\pipe.bat del %ARACHNETEMP%\pipe.bat
| if exist %TEMP%\pipe.bat del %TEMP%\pipe.bat
| if exist *._$b del *._$b
| if exist %ARACHNETEMP%\*._$b del %ARACHNETEMP%\*._$b
| if exist %TEMP%\*._$b del %TEMP%\*._$b
| if exist ip-up.bat del ip-up.bat
| if exist ppp.log echo Type 'arachne -u' to terminate SLIP/PPP
| connection.
| if exist lock del lock
| if not exist %TEMP%\ARACHNE.TMP\*.* goto done
| for %%f in (%TEMP%\arachne.tmp\*.*) do del %%f
| rd %TEMP%\arachne.tmp>NUL
| :done
| echo.
|
|
| ============================================================================
|
| Fri, Jul 30, 2004 - 5:17:50 pm 
|
| glennmcc>
| To clearify.... 
|
| %arachnetemp% is never removed
|
| But rather only 
|
| %temp%\arachne.tmp
|
| Which gets created by core.exe only if the env variable
| %arachnetemp% is not set
|
| ============================================================================
|
| Fri, Jul 30, 2004 - 6:54:16 pm 
|
| ray>
| Glenn,
|
| You have a point vis. roura, maybe the whole idea wasn't worth the
| trouble, if so Me Bad, since it was my idea. On balance I'd still
| opt for !pipe$.bat, like Joe, but I'm easy either way.
|
| But. I understand what you say above, but that's no subsititute for
| the sort of dedicated directory that I'm taling about since as it
| is,
| there could be lots of other stuff in the directory that you
| designate
| as TEMP. The way I think I'd work it (very preliminary just kicking
| ideas around) is that all Arachne's self generated stuff would go
| into
| the TEMP directory, which would be hard coded, just like \SYSTEM.
| And at shutdown, everything there would be left alone just incase
| you
| wanted to have a peek at it, but on startup the whole directory
| would
| be emptied:
|
| ...
| del .\temp\*.*
| ...
|
| To make this work without folks trashing their stuff by stupidly
| being alowed to set their own temp directory (IIRC this was a big
| problem in the old days), the temp directory must be hard coded.
|
| BTW
|
| Glenn, you might be able to help me with this:
|
| I'm get in this message on the status line just after pressing
| "Send mail NOW": "354 Enter message ending with "." on a line by
| itself"
|
| The only place this string exists in in smtp.log, but that is
| regenerated on use, so where the hE|_|_ is that string coming from
| and
| what does it mean?
|
|
|
| ============================================================================
|
| Fri, Jul 30, 2004 - 8:18:12 pm 
|
| glennmcc>
| You misunderstand.
|
| nothing gets deleted from %temp%
|
| But rather from the subdir \arachne.tmp\ which core.exe creates
| under %temp%
|
| %temp%\arachne.tmp\$roura$.bat gets deleted
|
| and then %temp%\arachne.tmp gets removed
|
| all files in %temp% remain untouched and %temp% does not get removed
|
| If the env variable %arachnetemp% is set... pipe.bat in that dir is
| deleted but the dir does not get removed. 
|
| Here's my dir tree when I shell-out when %arachnetemp% is not set
| and %temp% points to l:\temp
|
| --a--- 7,131 2-16-93 9:13a l:\temp\test.txt
| --a--- 36 7-30-04 10:09p l:\temp\arachne.tmp\donotpan.ic!
| --a--- 65 7-30-04 10:09p l:\temp\arachne.tmp\pipe.bat
| total files 3 total bytes 7,232 disk free space 98,107,392
|
|
| Here it is after exiting arachne....
|
| --a--- 7,131 2-16-93 9:13a l:\temp\test.txt
| total files 1 total bytes 7,131 disk free space 98,099,200
|
|
| Here it is when shelled-out and %arachnetemp% set to e:\!arachne
|
| --a--- 36 7-30-04 10:06p e:\!arachne\donotpan.ic!
| --a--- 65 7-30-04 10:06p e:\!arachne\pipe.bat
| total files 2 total bytes 101 disk free space 22,208,512
|
| And then after exiting....
|
| --a--- 36 7-30-04 10:06p e:\!arachne\donotpan.ic!
| total files 1 total bytes 36 disk free space 22,216,704
|
| As we see.... donotpan.ic! remains in-place... only pipe.bat gets
| deleted.
|
| In other words.... it's already doing exactly what you propose.
|
|
| ============================================================================
|
| Fri, Jul 30, 2004 - 8:20:57 pm 
|
| glennmcc>
| "354 Enter message ending with "." on a line by itself"
|
| That is coming from your ISPs SMTP server.
|
| ============================================================================
|
| Sat, Jul 31, 2004 - 12:18:49 pm 
|
| ray>
| Hi Glenn,
|
| Hmmmm, if that's the way it works vis. \temp then the only reason
| I can see that there would ever be a problem with the name of 'pipe'
| is if Arachne has to seach for that file via the DOS path statement.
| It seems to me that Arachne should look for her own files via an
| explicit path, then there would never by any chance of a conflict
| with another file of the same name. IIRC Michal mentioned somehting 
| about this.
|
| Anyway, I don't mean to make waves on this topic, I just think that
| we should look into the subject.
|
| Thanks for the info on "354 ..." I had no idea that Arachne could
| read a message from an ISP and print it on the status line (even if
| via a bug ;-) I found the error, it' because of a nice little
| function
| Michal wrote that simplifies the code, but it must be used with
| care,
| and I allowed an error in implementing it.
|
|
| ============================================================================
|
| Mon, Aug 02, 2004 - 4:34:53 am 
|
| mht>
| Just three words about hard-coding directories: never do it!
|
| About server messages: Arachne prints them on the status line quite
| often, they just change so quickly (unless your connection is
| horribly
| slow) that you have no chance to read them. BTW, I can't reproduce
| this error.
|
|
| ============================================================================
|
| Mon, Aug 02, 2004 - 3:28:01 pm 
|
| glennmcc>
| There we go....
|
| Michal came up with the same 3 words that I agree with.
|
| ============================================================================
|
| Tue, Aug 03, 2004 - 12:56:57 pm 
|
| ray>
| I won't argue the point, since we've more important things to worry
| about, but it seems to me that hard coded subdirectories are
| routine.
| Our \system directory is hard coded, for example, and many other
| software packages that I have won't let you play with their 
| subdirectories, and why would you want to anyway?
|
| But it's not time to do any of this yet, I just want this pot to
| simmer on the back burner.
|
|
|
|
| ============================================================================
|
| Tue, Aug 03, 2004 - 1:40:45 pm 
|
| glennmcc>
| IMO,
|
| It will be simmering for a long, long while. <g>
|
| ============================================================================
|
| Tue, Aug 03, 2004 - 1:18:42 pm 
|
| ray>
| That might not be a bad thing. These changes that affect things
| outside
| of core.exe itself should IMHO, be saved up and all done at once to
| get the agrivation of converting the whole project to whatever new
| filenames, directories, options, etc. etc. are decided on, more than
| once, if it can be helped. Take 'roura' for example: I still think
| the
| change is a good idea but for now it's simply more trouble than
| it's 
| worth (I'm goin' back). If we ever start to think about this
| seriously,
| then I'd say an entire release could be devoted to nothing but
| these filename etc. changes. However, I's say that it would be a
| mistake
| to avoid making structural changes that should be made indefinitely;
| that's like the sad fact that the U.S. still hasn't bitten the
| bullet
| and converted to the metric system. One day the change will happen;
| putting it off just makes the inevitable pain worse.
|
|
| ============================================================================
|
| Tue, Aug 03, 2004 - 4:52:50 pm 
|
| glennmcc>
| <Jar Jar Binks voice>
|
| My give up :(
|
| </Jar Jar Binks voice>
|
|
|
|
| ============================================================================
|
| Wed, Aug 04, 2004 - 4:12:10 am 
|
| mht>
| About hard-coded directories again: I meant hard-coding full paths,
| as C:\ARACHNE\SCRATCH. Some subdirectories may be hard coded, at
| least as a fall-back default values used when suitable environment
| variables or ARACHNE.CFG variables aren't set. Particularly,
| temporary
| directories and any directories that are written to should be made
| as
| configurable as possible -- there are so many possible problems like
| drives with limited free space, slow drives, read-only drives that
| only
| the end user (or a qualified system administrator) can make the
| right
| decision about. Calling SOMEPIPE.BAT always with a full path build
| using
| environment/Arachne variables is good, of course.
|
|
| ============================================================================
|
| Wed, Aug 04, 2004 - 9:11:07 pm 
|
| ray>
| Michal,
|
| Now that you are more specific I can see that I wasn't nearly
| specific enough. I agree with you almost completely. Hard coding the
| full path would be stupid, and of course things like cache, mail
| path,
| and download path should of course be optional. The only things
| I think should be hard coded are 'internal' ones: \system, \doc,
| \bin, and, I think, \aractemp (Xtemp??), because it is only used 
| (or should only be used) for small, short term, internal stuff like 
| halfpipe.bat,
| 'lock', and the other half dozen little files that Arachne creates,
| nothing more. It's debatable whether things like 'page.txt', the
| logs,
| and other such files would go into that directory or into another
| one, 
| called possibly '\usertemp' or something like that. Were 'usertemp'
| to exist, that too should be of optional path.
|
|
|
| ============================================================================
|
| Wed, Aug 04, 2004 - 9:24:20 pm 
|
| ray>
| ... but I didn't think about it properly -- you're right, *any*
| directory that might be written to must have an optional path,
| since,
| as you mentioned, the Arachne main directory might be read only ...
| but
| by that logic, the .cfg files would also need to be in a user
| selectable directory too, no? So maybe we need:
|
| \Xtemp
| \Usertemp
| \Xconfig
|
| ?
|
|
|
| ============================================================================
|
| Thu, Aug 05, 2004 - 2:53:09 am 
|
| mht>
| Maybe a command line switch for CORE.EXE (and ARACHNE.BAT) to use
| another configuration file, possibly in user's directory? This would
| eliminate the need to overwrite ARACHNE.CFG with PROFILE.ACF when
| switching profiles.
|
|
| ============================================================================
|
| Thu, Aug 05, 2004 - 5:02:01 pm 
|
| glennmcc>
| A little FYI for everyone.......
|
| Further to my post about compiling the 'full version' of Arachne.
|
| I have just completed some additional work.
|
| The exact same code works to create all 4 of these.........
|
| 464,728 8-05-04 6:38p
| e:\arachne\asrc\asrc177\chngs177.gpl\!exe.086\040805.exe
| 441,464 8-05-04 6:41p
| e:\arachne\asrc\asrc177\chngs177.gpl\!exe.386\040805.exe
| 467,752 8-05-04 6:31p
| e:\arachne\asrc\asrc177\chngs177.ue\!exe.086\040805.exe
| 444,424 8-05-04 6:34p
| e:\arachne\asrc\asrc177\chngs177.ue\!exe.386\040805.exe
| total files 4 total bytes 1,818,368 disk free space 28,901,376
|
| v1.77;GPL,086-
| v1.77;GPL,386+
| v1.77;UE03,086-
| v1.77;UE03,386+
|
| ============================================================================
|
| Thu, Aug 05, 2004 - 5:02:48 pm 
|
| ray>
| Interesting idea, or perhaps a command line switch just pointing to
| a user cofiguration directory, where arachne.cfg, mime.cfg
| toolbar.cfg
| and whatever else is user configurable would live, along with an
| optional extension to the switch specifying an alternate to 
| arachne.cfg. That would cover all the bases, since the user files
| would be in an (optionaly) separate directory and several different
| configurations could be chosen simply by changing the directory
| in which to find the config files.
|
|
|
| ============================================================================
|
| Sun, Aug 08, 2004 - 2:50:54 pm 
|
| ray>
| Joe,
|
| Well, I have no idea what you're up to vis. base3a>>b, so I've 
| started pulling in your code myself. I've done about 2/3 of the
| work and should be done in a day or two. I must say I'm impressed
| with your code and look forward to studying it in detail, it shows
| a deep understanding of what's going on. I couldn't help but think
| that your *arg*() family of functions might be used as general
| replacements for the *var*() functions, what do you think?
|
|
| ============================================================================
|
| Tue, Aug 10, 2004 - 12:14:22 am 
|
| Joe>
| Hi Ray,
|
| Well ... I've just compiled B3b!!! :-)
|
| I'll be uploading it later today.
|
| See ya!
|
|
|
| ============================================================================
|
| Tue, Aug 10, 2004 - 12:15:51 am 
|
| Joe>
| PS.
|
| I suppose it was implied ... but I'm online with B3b right now ...
| <g>
|
|
|
|
| ============================================================================
|
| Tue, Aug 10, 2004 - 7:33:13 am 
|
| ray>
| Excellento -- now I can integrate Michal's stuff and ... Hmmm, Joe,
| just thinking now -- once I get Michal's stuff rolled in *that*
| was supposed to be B3b, but there would obviously be a name overlap.
| So either you should pull your code after I get it (and once B3b
| is approved, obviously) or I'll call the integrated product B3c.
| The only advatage to the former is that we don't want to get too
| wild with different editions floating around, but that's a trivial
| point, so B3c it is.
|
|
| ============================================================================
|
| Wed, Aug 11, 2004 - 3:01:01 am 
|
| Joe>
| Yep, name overlap (aka. version number clashing) is
| *A Bad Thing* ... therefore next one must be B3c! <g>
|
|
| ============================================================================
|
| Wed, Aug 11, 2004 - 8:18:37 pm 
|
| ray>
| Testing 3c (looks good so far).
|
|
|
| ============================================================================
|
| Wed, Aug 11, 2004 - 11:48:16 pm 
|
| ray>
| Testing Glenn's Auth SMPT (SFSG).
|
|
| ============================================================================
|
| Thu, Aug 12, 2004 - 2:30:15 am 
|
| Joe>
| Speaking of whom (Have I spelt that right? Looks strange) ...
|
| Glenn, when you (or Mr X) get a chance, it would be good to
| check the POSIX stuff in B3b, either as a DJGPP (DOS) or Linux
| build/attempt.
|
| I'd like to know if the POSIX stuff is still intact. It's
| supposed to be, but mistakes can happen, so that's why some
| sort of checking would be prudent ...
|
|
| ============================================================================
|
| Thu, Aug 12, 2004 - 7:04:45 am 
|
| ray>
| Indeed. I went to one hell of a lot of trouble preserving all that
| code up to B3x in the faint hope that it might be of use, so it
| would
| be good to know.
|
|
| ============================================================================
|
| Thu, Aug 12, 2004 - 7:29:40 am 
|
| ray>
| BTW, Glenn,
|
| I've inexplicably lost my source .zip for the 175 release and I
| can't
| find it on your site, may I have a copy please? I'd like to double
| check the importation of your SMTP code into b3c, thanks.
|
|
| ============================================================================
|
| Fri, Aug 13, 2004 - 1:27:45 am 
|
| Joe>
| Ray,
|
| IIRC, it's available in Glenn's directory at ArachneDevelopment.
|
| ============================================================================
|
| Fri, Aug 13, 2004 - 1:10:05 pm 
|
| ray>
| Thanks Joe, for some reason I never even thought to look there <:-/
|
| BTW, I've got all the code integrated and there's two ways I can
| proceed from here -- I can finish editing and documenting it all and
| then upload it, or I can upload it as it is now (functioning but not
| properly edited), wait for any comments and possible changes that
| you guys might suggest as you review each others work, and then
| incorporate those in a finished edition. Any preferences?
|
| BTW Joe, don't I remember that when I was running your own compile
| of J4 I had a 'Dos memory' improvement of 2K? The reason I ask is
| that
| as she is now (b3c) I'm about breaking even on free memory.
|
|
|
| ============================================================================
|
| Sat, Aug 14, 2004 - 11:18:34 am 
|
| Joe>
| Ray,
|
| Linus (I think) said "Release Early, Release Often". <G>
|
| As regards memory, this depends on compiler optimization settings.
| I don't remember what settings I used originally with 1.73J4 - I
| suspect the same as NOKEY.PRJ except for "386, 387 code generation".
| The CORE.PRJ I included with Base3b was nothing special, just what
| I typically use for general purposes, plus the necessary "defines"
| for Arachne Base3b. It didn't include the "386, 387 code generation"
| optimizations.
|
| With that in mind, I did a simple test of free memory while online
| for the above build, then again with maximal code-size optimization
| (except for the potentially dangerous "suppress redundant loads"
| option). The results were as follows :
| "No particular optimization" -> Used 13K more memory than 1.73J4.
| "Maximal optimization" -> Used 2K less memory than 1.73J4.
|
| So, I suspect 1.73J4 and Base3b would have pretty much the same
| memory
| usage if compiled with the same optimizations (including
| recompilation
| of all OBJ files).
|
|
| ============================================================================
|
| Sat, Aug 14, 2004 - 11:36:29 am 
|
| Joe>
| BTW, Glenn,
|
| Has there been any progress recently with DJGPP (by Mr X)
| or Linux (by yourself) builds of Arachne? Even if there
| hasn't been, I for one would still like to know that the
| POSIX stuff in Base3b hasn't suffered accidental damage
| along the line.
|
| Also, let us know if we can assist with either of these
| Arachne ports.
|
|
| ============================================================================
|
| Sun, Aug 15, 2004 - 3:40:08 pm 
|
| ray>
| Joe,
|
| Tx for info about memory. It's hard enough keeping track of
| what the code per se is using across various builds, but throw in
| compiler settings and things get complicated, although I think we've
| just about nailed the latter down. Speaking of memory consumption,
| it
| would be nice if we had some way of obtaining a 'hard' number on
| that; right now we have 'Dos memory(KB) ... ' but of course that
| number changes depending on ones DOS configuration. What do you
| think
| would be involved in getting the inverse? ie the amout of memory
| that
| Arachne is *using* at any given time vs. the amount free?
|
|
|
| ============================================================================
|
| Mon, Aug 16, 2004 - 1:43:42 am 
|
| mht>
| Good idea. Too bad that this number cannot be obtained directly with
| a single function call... but it can be done.
|
| ============================================================================
|
| Wed, Aug 18, 2004 - 2:37:52 am 
|
| vbm>
| Glenn, please, send me script for creating install.exe of arachne.
| I was try created it with rar, but can't run setup.bat after
| unpacking...
| thanks.
|
| ============================================================================
|
| Tue, Aug 24, 2004 - 8:11:40 pm 
|
| ray>
| I'm not sure if this is news to anyone but I just found out that
| if the list of compiler defines in the Borland IDE has spaces in it:
| "MSDOS; WATTCP; etc", there is no problem within the IDE itself,
| but if
| you use 'prj2mak' to create a make file and then try to do a command
| line build, you get an error. It seems that 'make' requires the
| list to be sans spaces.
|
|
|
| ============================================================================
|
| Wed, Aug 25, 2004 - 9:27:45 pm 
|
| ray>
| Joe,
| Joe,
|
| I'm just noticing that 'SearchArg()' in 'htmlstat.c' appears to be
| unused. Can you confirm? Perhaps this function was meant to be
| used to replace 'SearchVar()' at some location but the change wasn't
| made? Or the function is to be used latter?
|
|
|
| ============================================================================
|
| Wed, Aug 25, 2004 - 9:28:45 pm 
|
| ray>
| Hmmmm, did I really enter Joe twice or is that the bug?
|
|
| ============================================================================
|
| Thu, Aug 26, 2004 - 2:16:19 am 
|
| Joe>
| Ray,
|
| Interesting info regarding the "defines", sounds like a minor
| compiler bug ...
|
| The SearchArg() function was added for completeness and there
| isn't a missing change (call to it) in the code. If and when
| the code is converted to replace all the old functions, that's
| when it will be used/required.
|
|
| ============================================================================
|
| Thu, Aug 26, 2004 - 7:12:12 am 
|
| ray>
| Joe,
|
| Cool, thinking ahead, I like that.
|
| Yep, it's too bad Borland isn't still developing their 3x compiler.
| If my counting is correct, that's the 5th bug we've found in it.
| We're
| the ultimate beta testers.
|
|
|
| ============================================================================
|
| Thu, Aug 26, 2004 - 8:13:58 pm 
|
| ray>
| Hey Joe, for what it's worth, I've been importing your code into B5
| today and found something that might be interesting: I had to bump 
| my "IE_MAXWAPLEN" back to Glenn's '33000u' value to avoid geting
| an XSWAP error from 'putikonw.c'. That seems strange given that
| B5 is quite a bit smaller than B3c, but it could just be an
| importation error on my part since I haven't proofed it all yet.
| Just thought you might like to know.
|
|
| ============================================================================
|
| Thu, Aug 26, 2004 - 8:14:52 pm 
|
| ray>
| testing
|
|
| ============================================================================
|
| Fri, Aug 27, 2004 - 1:49:16 am 
|
| Joe>
| Ray,
|
| Sounds like you have a memory leakage problem? Don't use the
| "supress redundant loads" option, I think that's what caused
| a similar situation for you a while ago?
|
| ============================================================================
|
| Fri, Aug 27, 2004 - 4:03:32 am 
|
| mht>
| Ray, Joe,
| The fix for a compiler bug that caused XSWAP error in PUTIKONW.C was
| discussed on AD quite long time ago. Too bad it isn't present in
| B3C.
| Is it on place in B5?
|
|
| ============================================================================
|
| Fri, Aug 27, 2004 - 12:42:48 pm 
|
| ray>
| Michal, Joe,
|
| That's the first thing I though of, too, but that's not it AFAICT.
| The error is on a different line. That fix is in B5 anyway.
| Once I get B5 uploaded you guys can have a look for yourselves,
| there
| is so much new stuff in B5 that anything is possible.
|
|
|
|
| ============================================================================
|
| Sat, Sep 04, 2004 - 9:37:29 am 
|
| ray>
| Joe,
|
| Anything you want changed before I upload 3c?
|
| ============================================================================
|
| Sat, Sep 04, 2004 - 6:54:15 pm 
|
| ray>
| Hey Glenn, was that you who just delete my 'junk' files ;-) ...
| I have a bug some where, and I'm tracking it down so I will be
| uploading dozens (probably) of these, so please leave them until
| I'm finished, then I'll ask you to delete the lot, thanks.
|
|
| ============================================================================
|
| Sat, Sep 04, 2004 - 7:02:17 pm 
|
| glennmcc>
| Oops.... sorry.
|
| To me, 'junk' means 'junk'..... so I delete it. ;-)
|
| I'll leave them alone now.
|
| Your core3c.zip is in there and is intact. :)))
|
|
| ============================================================================
|
| Sat, Sep 04, 2004 - 7:03:29 pm 
|
| glennmcc>
| I updated index.htm to include a link to it.
|
| ============================================================================
|
| Sat, Sep 04, 2004 - 7:11:09 pm 
|
| glennmcc>
| Just remebered that I never mentioned how much storage space
| we have there at Tripod.
|
| It's 20mb
|
| ============================================================================
|
| Sun, Sep 05, 2004 - 9:22:14 am 
|
| ray>
| Glenn,
|
| Yup, that's why they were called 'junk'. If I gave some other name,
| I thought you'd wonder what was going on, but be reluctant to tamper
| lest you killed something important. Anyway, it looks like I have
| a double problem; as far as I've got, the 3c core works OK, but only
| with an older arachne.cfg, so there's a bug in B5 (not surprising)
| and
| also something in arachne.cfg that will disable FTP -- a cursory
| look
| leads me to suspect that it's 8bit encoding.
|
|
|
| ============================================================================
|
| Sun, Sep 05, 2004 - 7:48:50 pm 
|
| ray>
| Glenn,
|
| bug found, please delete all junk :-)
|
|
|
| ============================================================================
|
| Sun, Sep 05, 2004 - 9:30:41 pm 
|
| glennmcc>
| Will do.
|
| What was and where was the 'bug' ?
|
| ============================================================================
|
| Sun, Sep 05, 2004 - 9:35:58 pm 
|
| glennmcc>
| Done.... 'junk' gone.
|
|
| ============================================================================
|
| Mon, Sep 06, 2004 - 2:17:14 am 
|
| Joe>
| Ray,
|
| Re. your next "B3c" upload :
|
| 1. It should be B3d! <g>
| 2. I think it should include the stuff Michal has contributed over
| the past few months.
| 3. Please check for "//ifndef POSIX" in the code --- IIRC, I came
| across at least one instance of commented-out POSIX conditional
| code, based around the "ifndef" conditional. I'll search out
| the specific instances, if you need ...
|
|
|
| ============================================================================
|
| Mon, Sep 06, 2004 - 2:22:49 am 
|
| Joe>
| Glenn,
|
| Have you or Mr X. had a chance to try the B3c code in a POSIX
| context (ie. DJGPP and/or Linux)? I'd like to get some feedback
| on this, so Ray can correct any "inadvertent harm" that may have
| occurred to the POSIX code during the clean-up process, preferably
| for the "Base 3d" iteration.
|
|
| ============================================================================
|
| Mon, Sep 06, 2004 - 2:27:22 am 
|
| Joe>
| Ray,
|
| Good to hear you've tracked down "the bug". I presume this was in
| relation to your apparent memory leakage problem. Like Glenn, I'm
| curious to know what the specifics were. <g>
|
|
|
| ============================================================================
|
| Mon, Sep 06, 2004 - 11:34:53 am 
|
| ray>
| Glenn, Joe,
|
| The bug was just a careless and trivial mistake made in B5
| while fooling with the 'all config variables parsed at startup'
| project:
|
| ˛
| //` strcpy(url->user, ConfigVariable("FTPusername", NULL)); ˛
| //` strcpy(url->password, ConfigVariable("FTPpassword", NULL)); ˛
| //` strcpy(url->host, ConfigVariable("FTPserver", NULL)); ˛
| //` strcpy(url->file, ConfigVariable("FTPpath", NULL)); ˛
| strcpy(url->user, sUI.ftpusername); ˛
| strcpy(url->password, sUI.ftppassword); ˛
| strcpy(url->host, sUI.ftpserver); ˛
| strcpy(url->password, sUI.ftppath); ˛
|
| It isn't surprising that I had trouble with 'junk[x]' as my
| password ;-)
|
|
| ============================================================================
|
| Mon, Sep 06, 2004 - 11:39:31 am 
|
| ray>
| Joe,
|
| As to the name of the next upload, it's 3c not 3d! That's because
| ~3c is for inspection and aproval only and is *replaced* by
| 3c when aproved. 3d will progress from 3c, but I don't want to
| take too big bites by cramming in to much into one release.
| After 3c I want to release 5 and let you guys have a look at that,
| then we'll plot strategy.
|
|
| ============================================================================
|
| Tue, Sep 07, 2004 - 2:41:04 am 
|
| Joe>
| Hi Ray,
|
| Huh? What's ~3c and how is this different from 3c? As far as I
| can tell, what we currently have to look at (play with) is called
| Base 3c, so the next iteration should be called Base 3d, surely?
| Otherwise we will just create confusion. (Or do you expect more
| than 26 iterations of Base 3x?;-)
|
|
| ============================================================================
|
| Tue, Sep 07, 2004 - 7:36:32 am 
|
| ray>
| Joe,
|
| Didn't I explain near the top of BASE~3C.TXT? ... To prevent (well,
| er,
| it seems to have *caused* ;-) confusion and too many versions out
| there, before the 'authorised' edition is released (Bxx), a trial,
| or test or provisional or whatever version is released (B~xx) -- the
| idea is that the contributors will get a peak at what I've done with
| their work and fix any errors or omitions and then the aproved 
| edition is released which replaces the trial version. Michal has
| had a few corrections for me plus the addition of the compiler
| optimizations etc. The nice thing about this system is that it lets
| me trash and forget B~xx once Bxx is out, thus I don't have to
| hang on to too many editions. Think of B~3c as a 'draft'.
|
|
| ============================================================================
|
| Wed, Sep 08, 2004 - 10:28:45 pm 
|
| Joe>
| OK, thanks for the explanation.
|
| Perhaps I overlooked the tilda at some point in the past ... :-(
|
|
| ============================================================================
|
| Thu, Sep 09, 2004 - 7:40:33 am 
|
| ray>
| Joe,
|
| Anyway, so have you no corrections before I upload the 'final'?
| I'm expecting a few corrections and updates from Michal, then she's
| finished. BTW, Greg M. is having strange trouble with cookies with
| the 3c core. His last post on the subject was on the main list.
| I haven't a clue what's wrong, can you shed some light?
|
|
|
|
| ============================================================================
|
| Thu, Sep 09, 2004 - 3:49:43 pm 
|
| glennmcc>
| I think the proble may have something to do with 'expire dates'
|
| Or to be more precise..... the lack there-of in most of the Yahoo
| cookies.
|
| ============================================================================
|
| Thu, Sep 09, 2004 - 8:13:37 pm 
|
| ray>
| Glenn,
|
| God knows. He's getting XSWAP errors too so I may have sent him
| the wrong blasted .exe anyway. I'm getting frazzled, inacurate and
| incompetent.
|
|
| BTW, hey, when's 1.77 due out? Weren't we expecting it some time
| ago?
|
| ============================================================================
|
| Thu, Sep 09, 2004 - 8:40:53 pm 
|
| Joe>
| Glenn,
|
| Cookies without an expiry date are temporary cookies that are
| supposed to be deleted by the client (ie. Arachne) when the
| current session is closed. This is a functionality that I was
| thinking of adding to Arachne, however there is no pressing
| need, as usually Arachne still interacts fine as is.
|
|
| ============================================================================
|
| Thu, Sep 09, 2004 - 8:46:49 pm 
|
| Joe>
| Ray,
|
| Before uploading ...
|
| 1. I'd still like you to re-consider the "3c" naming thing. <g>
| 2. It sounds like you are already adding Michal's changes - good.
| :-)
| 3. Double-check the "ifndef POSIX" stuff (Glenn, how's the POSIX
| stuff in "3c" behave compared to earlier code?).
| 4. Greg's stuff worries me ... :-/
|
|
| ============================================================================
|
| Thu, Sep 09, 2004 - 10:14:18 pm 
|
| ray>
| Joe,
|
| The changes from Michal were at least started and pending back when
| the synch with 1.75 was planned so the door is open on that stuff,
| but the subequent changes should be saved for another edition -- I
| don't want 3c to be open ended. Most of what Michal has done
| subsequently was infact done by me, albeit with his
| guidance/cooperation, and is in B5 which I've got to get back to
| since
| it's full of loose ends. I've got too much stuff dangling right now
| so the door is closed on 3c as it is. Once you've seen 5, then 
| either it will stand as such (with whatever mods are negotiated, or
| the good stuff in it can be moved into what will then be B3d ...
| we'll
| see.
|
| As for Greg -- insuficient data.
|
|
| ============================================================================
|
| Sun, Sep 12, 2004 - 5:41:33 pm 
|
| ray>
| Ahh ... Dos memory (KB) 174[+] ... :-)
|
|
|
| ============================================================================
|
| Thu, Sep 16, 2004 - 8:42:40 pm 
|
| ray>
| Core.exe: 407,088 bytes, and 176[+] free!
|
| Good work Michal -- I was predicting we'd break the 400K .exe size
| barrier by the end of the year, but it looks like we're almost there
| now : - )
|
|
| ============================================================================
|
| Thu, Sep 16, 2004 - 9:42:00 pm 
|
| ray>
| Michal,
|
| I finaly had time to implement that change to "Dos memory" we
| talked 
| about. Mine now reads "Free low memory 179088" -- I figured that
| the unabridged string is the most informative, and it saves
| 80 bytes too.
|
|
| ============================================================================
|
| Fri, Sep 17, 2004 - 7:22:00 am 
|
| ray>
| Most interesting:
|
| My Free low memory stat bounces around quite a bit. Session 
| started with 179,408, now it's down to 179,280.
|
|
|
| ============================================================================
|
| Fri, Sep 17, 2004 - 5:05:44 pm 
|
| ray>
| Yo! 181[+] with 3c and new hixlov.lib!
| We'll get to 200[+] or I'm a WinAddict ;-)
|
|
| ============================================================================
|
| Mon, Sep 27, 2004 - 4:17:51 pm 
|
| ray>
| Boy, things sure are dull on the board these days ...
|
| ============================================================================
|
| Thu, Sep 30, 2004 - 8:53:49 pm 
| URL
|
| glennmcc>
| Let's see if this 'perks thing up' a little bit. <g>
|
| Latest news.... Sep 30, 2004 21:37:00 
| v1.77;GPL,386+ will be available 
| for download in just a few more hours. 
|
|
| ============================================================================
|
| Thu, Sep 30, 2004 - 10:04:10 pm 
| URL
|
| glennmcc>
| She's ready for primetime. ;-)
|
| ============================================================================