DOS Coders1 Message Board Thu, 01 Sep 2005 01:33:58 ______________________________________________________________________________ | ============================================================================ | | 1) Fri, Jul 01, 2005 - 12:06:06 am | | gregy> | Okay...new-style board up, and I'll have the latest archive up in a few | more minutes... | | ============================================================================ | | 2) Fri, Jul 01, 2005 - 1:11:18 am | | gregy> | Okay | The | Archives | are | up | and | it | looks | like | we're | in | business....... | | ============================================================================ | | 3) Fri, Jul 01, 2005 - 8:28:38 am | | ray> | Thanks | Gregy | | | | ============================================================================ | | 4) Sat, Jul 02, 2005 - 1:09:41 am | | gregy> | No...thank Glenn...he's the guy who wrote/rewrote/rerewrote these board | scripts.... | | | | ============================================================================ | | 5) Sat, Jul 02, 2005 - 2:35:54 pm | URL | | glennmcc> | See .... http://www.hi-line.net/~gfeig/brd/misc/mindex.php (post #5) | | ============================================================================ | | 6) Sun, Jul 03, 2005 - 2:27:23 am | | gregy> | Yeah, and go see (post #6)... | | | | ============================================================================ | | 7) Sun, Jul 03, 2005 - 2:28:38 am | | gregy> | ..but, you've really done a nice job, Glenn, with your "tweaking" I | mean... | | | ============================================================================ | | 8) Sun, Jul 03, 2005 - 2:33:01 am | | gregy> | Oh, and BTW Coders, I have consistently Archived this board in .TXT by | doing an "ArachnePconvert" ...and, if you guys grab the Archives, you | will find that, while this board has not, previously, displayed your | indentation formatting, the actual post-DATA file has stored it, and, | the "P" convert function gives a .TXT file with the indentation shown. | ... | so, all your hard work hasn't been lost.... | | | ============================================================================ | | 9) Sun, Jul 03, 2005 - 8:25:00 am | | ray> | Cool, the board isn't supposed to be for things that need saving, but | it doesn't hurt, that's for sure. | | | | ============================================================================ | | 10) Sun, Jul 03, 2005 - 8:25:46 am | | ray> | Hey Gregy, supposing you put a link to the other board at the bottom | of this one, and visa versa? | | ============================================================================ | | 11) Sun, Jul 03, 2005 - 7:07:23 pm | | Joe> | Honeysuckle Creek, we have a problem. | | I've selected "Show 32 messages" here and it don't show any more | that the default. | | | ============================================================================ | | 12) Sun, Jul 03, 2005 - 7:37:56 pm | | glennmcc> | It can't show any more than the nuber that exists. | | When you chose 32... only 10 existed | | | ============================================================================ | | 13) Sun, Jul 03, 2005 - 9:28:41 pm | | ray> | LOL | | ============================================================================ | | 14) Sun, Jul 03, 2005 - 11:32:56 pm | | gregy> | Joe....don't mind them laughing at you.... | .. | Anyway...I think you'd like to see "previous posts"...?? | .. | The link to the "Archives Directory" is right below the "Show 11 | messages" button... | ... | | | ============================================================================ | | 15) Sun, Jul 03, 2005 - 11:34:59 pm | | gregy> | Ray....Re: Cross-links....hmmmmm. I'll think about it... | ..or, maybe a link to a "Boards Links" page would be better. That way | I can change the .HTM easier that changing/adding/etc on the board's | code... | | | | ============================================================================ | | 16) Mon, Jul 04, 2005 - 2:37:29 am | | gregy> | Ray....Re: putting a link to this board on the other (or any other) | is not likely... | WHY..?? | | You remember, that this board is a "private" board, BY INVITATION ONLY | for you Coders....and I invite NO one. If you want someone to know | about this board, the YOU guys must give them the link, and I suggest | that you keep it to a close group. Otherwise, you'll get "spam posts" | here, and a lot of unnecessary traffic....and, someday, you'll be | complaining to me, and I'll hafta fix it.... | | | ============================================================================ | | 17) Mon, Jul 04, 2005 - 6:56:53 am | | ray> | I'd risk it, but you're the boss. | | | ============================================================================ | | 18) Mon, Jul 04, 2005 - 11:11:29 pm | | Joe> | Glenn, Gregy, | | Thanks for the tip about the "less than X messages" problem ... I | hadn't encountered this before, do messages disappear from the board | when they're archived? | | | ============================================================================ | | 19) Tue, Jul 05, 2005 - 6:29:29 am | | gregy> | Joe, yes, message-post files are "zero-byted" whenever the current | board is archived. | | Glenn uses an automated process, which archives/displays his boards | posts each month. It's really nifty.... | | But, I don't do that (for a variety of reasons - mostly lack of post | volumne). | | Instead, I manually archive each Quarter (hence, my archives have a | 05Q2, etc, part in their filename), by grabbing "All Messages" and | pressing "P" to convert them to .TXT, then upload them to the /archives/ | directory, which is linked just below the "Show xxx messages" button. | | | ============================================================================ | | 20) Tue, Jul 05, 2005 - 6:34:54 am | | gregy> | Ray, Re: "...but your the boss".....nope...this is YOUR (Coders) board, | and it's for your use. I'll configure it/conform it to any way that | is reasonable/within-my-capabilities (pretty limited, I must confess) | ... | .. | I have decided to not cross-link because of the "privacy" considerations | I just stated, above, but...you guys can passout the URL like confetti | if you want.... | | Just, my experience has been, that in not too long a time, you'll be | overwhelmed, and asking me to "fix that"....which I am prepared to | do...but it's a secret until that occurs.....(if I told you, now, I'd | hafta kill you).... | | | ============================================================================ | | 21) Fri, Jul 08, 2005 - 8:47:59 pm | | ray> | I built myself a little automated bat file that rebuilds after | having modified core.mak to move all the overlayed modules one by one | to static link and then do a speed test and log the result. Let her | rip for a hour and the results: Untill the point where she chokes | for lack of memory, and with every module that could even possibly | be useful static, there's a 2% improvement over stock. 2%. | | It seems that all the improvements are in the other direction ;-) | | | | ============================================================================ | | 22) Fri, Jul 08, 2005 - 8:58:56 pm | | Joe> | Hi Ray, | | Don't forget, stuff like RAM disk and CPU and disk speeds vary | from one user to the next, so don't jump to conclusions | | ============================================================================ | | 23) Fri, Jul 08, 2005 - 9:18:41 pm | | ray> | core.exe: 380,592 bytes, free DOS mem 250152 :-))) | | | | ============================================================================ | | 24) Fri, Jul 08, 2005 - 9:21:11 pm | | ray> | Joe, | | Yeah, I don't take these numbers seriously yet, but they do raise | some interesting questions. | | | | ============================================================================ | | 25) Sat, Jul 09, 2005 - 2:54:15 pm | | glennmcc> | The one thing that is it _does_ "strongly suggest".... | | The major difference is in memory usage.... not in speed. | | But, there is one very major problem that CAV found by experimenting | with mine. | | More modules overlayed == even worse problems with frames than already exist | | ============================================================================ | | 26) Sat, Jul 09, 2005 - 5:03:59 pm | | ray> | Yup, it's a puzzlement -- more questions than answers. It seems that | the overlay manager does such a good job that there's no point in | having most modules static -- until something goes busted. But why | should it go busted and not just slow down? The interupt handlers | need to be static IIRC but other than that it should be a question of | speed not one of reliability, in theory anyway. | | | ============================================================================ | | 27) Sat, Jul 09, 2005 - 5:06:31 pm | | ray> | BTW, try cutting down your stack size, that's how I got 250K free. | Mind most of my cleaning was 'stack biased' so it may be that I can | get away with less than you. I'm using 0x6000 now and it seems OK, | but I hardly expect I'll get away with it. Let me know how small you | can get. | | | | | ============================================================================ | | 28) Sat, Jul 09, 2005 - 5:52:09 pm | | glennmcc> | unsigned _stklen=32750u;//!!glennmcc: July 09, 2005 -- cut stack size in hal | //unsigned _stklen=65500u; | | CORE.EXE memory usage.... 384,400 bytes | | Current stats inside of 'tinyest core' ;-) | | Dos memory 259,168[+] green | | How ya like them apples ??? | | ============================================================================ | | 29) Sat, Jul 09, 2005 - 6:07:55 pm | | glennmcc> | Catch me if you can. | | 'tinyer tinyest' has the stack cut down to your size of 24000 | | CORE.EXE memory usage .... 376,096 bytes | | Dos memory 268,032[+] green | | ============================================================================ | | 30) Sat, Jul 09, 2005 - 6:10:38 pm | | glennmcc> | 20000 will not run at-all... crashes almost instantly and is locked-up-tight | | Had to power-down. | | | ============================================================================ | | 31) Sat, Jul 09, 2005 - 6:39:22 pm | | ray> | Right, I locked up at 0x5000. Let's see how stable things are | with a half stack -- I've been ok for two days, but there's bound to | be something. | | Oh, and I can 'catch you' any time by just using more overlays like | you but I'll let you break trail with that. | | | ============================================================================ | | 32) Sun, Jul 10, 2005 - 10:30:16 pm | | ray> | Breakthrough! | My core is now small enough that I can run it | in debug mode in Borland. Just barely though, my free mem is | deep into the red, but it does load and trace. | | | ============================================================================ | | 33) Mon, Jul 11, 2005 - 3:03:35 am | | Joe> | Holy cow, Ray! | | You guys are sure pushing things to the limit! | | This raises an interesting option ... Gregy, IIRC, you've played | with OS/2 ... does Arachne work there? If so, IIRC, OS/2 can give | DOS environments with > 700k conventional memory ... perhaps this | would make debug mode viable??? | | | ============================================================================ | | 34) Mon, Jul 11, 2005 - 3:24:59 am | | gregy> | Yes, it does work on OS/2. I don't think it can get >700Kb conventional | ...but, should be able to get at least 640....and...OS/2 has debuggers, | that should be usable for debugging running x86VirtualMachines (which | is what the OS/2 "DosBox" actually is. | ... | And, one OS/2 debugger (which I have, somewhere around) is ICAT, which | runs a special debug kernal on the target box, with complete debug | monitoring/information continuously transmitted out a serial port to | a second monitoring box. I think it can also pump info out thru an | ethernet NIC, but I'm not sure. | ... | I'm just setting up OS/2 permanent RAMdisk on my installed Warp3Connect | system, and install ArachneRAMdisk installation, and tweak it to see | what maximum I can get for conventional memory.. | .. | back to you later... | | | ============================================================================ | | 35) Tue, Jul 12, 2005 - 2:56:52 am | | gregy> | As promised, I'm checking into maybe doing debug under OS/2 DOS VDM. | .. | To further that, I'm blowing the dust off my rusty OS/2-->>DOS apps | skills.... | .. | Now posting here using ArachneOS2, which is a standard v.1.83 installed | on a native OS/2 RAMdisk. I'm getting 115,584bytes DOS mem free while | ONLINE. | That's not too bad, and I think I can improve that by loading EPPPD | HIGH, and a couple other things. If I could get ArachneOS2 to use the | standard OS/2 DUN in "Connection Ready Mode" that would free up even | more memory. Maybe we can do it.... | | | ============================================================================ | | 36) Wed, Jul 13, 2005 - 3:10:33 am | | Joe> | From memory (no pun intended), the >700k for DOS apps trick | in OS/2 involved loading device drivers and TSR's separately | to the main DOS environment, yet providing the services of | said device drivers and TSR's to that environment. | | Never used OS/2, this is simply one of those things I remember | that was claimed for OS/2. Not sure if that's enough of a clue | to go on, but it's all I can tell you ... | | | ============================================================================ | | 37) Wed, Jul 13, 2005 - 10:06:26 pm | | gregy> | Joe...yes...right now, in OS/2 I'm getting about 3Kbytes more DOS mem | (checking with the ALT-M) than I do with straight DRDOS6.0...I'm sure | this is purely because OS/2 is providing the mouse driver, so I don't | have to load it, when I fire up Arachne in there. | ... | Now, I haven't even tweaked my DRDOS6.0 for max...which I'm going to do, | and also the MSDOS6.22 which is on this same minimal-box (which makes | this whole setup very useful for testing). Once I can do that, such | as loading EPPPD high, and, if I can manage it, making Arachne use the | OS/2 PPP packet driver (which it says it is providing at 0x69hex...then | I can free up even more space. | Anyway, I'm twiddling, and also asking on some of my OS/2 lists.. | .... | again...get back to you all, later.. | | | ============================================================================ | | 38) Thu, Jul 14, 2005 - 4:31:01 am | | Joe> | Hi Gregy, | | FWIW, DR DOS 6.0 won't give you as much free conventional memory | after tweaking as 7.0x, not sure how it compares with MSDOS 6.22 | or 7.10. I know PC-DOS 7 and 2000 are supposed to be better for | this than the M$ product. Basically, that's because later DOSes | had more tricks added to allow for such tweaking. So, I'd be | surprised if you couldn't do better with OS/2. | | However, note that I'm talking about a fundamental difference here | between OS/2 and normal DOSes, not the usual tweaking. I'm talking | >700K free conventional memory, that's more than the whole 640K | that DOS might give if it consumed zero conventional memory! That's | what I remember some IBM guy telling us incredulous DOS users when | OS/2 Warp was released, saying this was the way to go for DOS app's. | | It may be that >700K only applied to text-mode app's, so perhaps | graphical app's like Arachne can only achieve 640K (655k) free | conventional memory. I didn't know enough about graphics stuff at | the time to ask. If so, and if OS/2 can avoid the need for any | OS data/code below 640K, perhaps also with device drivers and TSR's | in another virtual machine, then the ultimate free conventional | memory possible for Arachne under OS/2 is 639K, since the interrupt | vectors must be at 0000 for compatibility. The best I've achieved | under DR-DOS 7.02+ is about 629K, I think (that's with mouse driver, | disk cache, EMS, and a few odds and ends). So I guess that implies | the ultimate free conventional memory under OS/2 should be about 10K | more than can be achieved under DR-DOS 7.02+, even more when compared | to any M$ product. I would make this easier too, if OS/2 can make the | BIOS "disappear". | | Sorry for rambling here, I'm just trying to recall what that IBM guy | said years ago and to make sense of it all, to see if OS/2 can be of | help to provide a debugging environment for Arachne. One thing we need | to know from Ray is just how much free conventional memory he has to | begin with. | | ============================================================================ | | 39) Thu, Jul 14, 2005 - 1:07:36 pm | | ray> | Joe, | | I'll post some hard numbers shortly. Every byte will count since | it seems I can't run the whole core in debugging mode at the same time, | I've got to build some modules without it or she hangs. Obvioulsy more | memory would reduce that problem. Anyway, it sure is cool tracing | through the program. | | | | ============================================================================ | | 40) Thu, Jul 14, 2005 - 11:59:42 pm | | gregy> | Joe...hey..you gotta good memory..... I'm re-digging into my "OS/2 | Warp Unleashed" volume, and finding approximately the same values | you're quoting... | .. | FWIW...that >700K spec was if VideoRestriction==CGA was implemented. | With straight VGA, it's about 639-642 (the value varies depending upon | other settings, some of which we need, to loadhigh stuff, etc.) | | But, I'm tweaking all my OS-partitions, including the MSDOS6.22 one, | just to see what's possible. However, the OS/2 looks like it will | ultimately yield the most. Plus, I'm remembering, and running into, | some info about "remote debugging" meaning that you install a remote | debugger kernal, install a DebuggingKernalClient on another box. | Connect them together thru a SerialPort Null Modem Cable, and run the | target app on the DebuggerBox, watching the actual app's performance, | which simultaneously viewing the debug/run info on the screen of the | second box. I'm gonna try using my laptop as the second box.. | | | ============================================================================ | | 41) Fri, Jul 15, 2005 - 12:02:25 am | | gregy> | Ray, Re: the SecondMonitor in OS/2....yes, there are brief, vague | allusions to that....but...the information available is scarce to nil, | and nobody I have contacted has any experience with them. But, it's | supposed to work with two cards in there. Or, an alternative appears | to be using a VidCard with dual-head (i.e. two VGA outputs, inputs, | processors, etc.) | | | ============================================================================ | | 42) Fri, Jul 15, 2005 - 2:40:51 am | | Joe> | Hey, that's an interesting possibility, Gregy ... there are | indeed some debuggers that can do the debugging on a separate | terminal, although I've never played with such an animal. | | Now, I wonder if this is a capability of Borland's BC tools??? | It would surely take much less memory to run such a debugger | than one that has an integrated IDE. In other words, if the | debugger IDE ran on another machine, with just the essentials | and some comm's taking up memory on the test machine, along | with the application being tested (Arachne) of course. In the | case of OS/2, I wouldn't be surprised if the second machine | couldn't simply be "virtual", so that you could achive this | with just one box. Most interesting ... need to figure out | what debugging goodies Borland have for BC ... | | Oh, about two video cards ... from memory, using a separate | video card for the debugger means two different types of | video card - one type (usually colour graphics) for the | test application, the other type (usually text monochrome) | for the debugger. They have to be differnt types, else they | will conflict (try to occupy the same memory space). I think | this is viable/exists. OTOH, a debugger that runs through a | serial port can avoid this constraint, since then there is only | one video card per machine. | | | ============================================================================ | | 43) Fri, Jul 15, 2005 - 12:56:13 pm | | ray> | Gents, | | The video card thing is done on the hardware level, it's OS | independant. As Joe said, VGA/EGA and mono/CGA cards address | different memory, thus they can live together with no conflict. | Dual head cards are a totaly different thing as they in effect | just take one monitor and spread the screen out over two. The ability | of OS/2 to give over 700K free if there is no monochrome monitor | is due to the fact that the mono address space is contiguous with | the lower 640, and can thus be appended to it with very little trouble. | 386MAX and other memory managers did the same trick. | | | ============================================================================ | | 44) Fri, Jul 15, 2005 - 5:36:51 pm | | glennmcc> | OpenDos can do dat too via MEMMAX.COM :)) | | -------------- | Format | | MEMMAX [/?|/H] [-|+L] [/L] [-|+U] [/U] [-|+V] [/V] | | Explanation | | Use MEMMAX to control extended memory at the system prompt. | MEMMAX allows you to temporarily enable and disable memory | areas. Some applications may not work properly when certain | memory types are enabled or disabled. | | When you load the DOS software into high or upper memory, the | lower (conventional) memory normally used by DOS is made | available to applications. Some applications, however, will not | load into lower memory and may require lower memory to be | disabled. Use the MEMMAX L switches to disable and re-enable | lower memory. | | Because some applications are sensitive to upper memory, it is | disabled by default when you start the operating system. Unless | you are using HILOAD or LOADHIGH (which automatically enables | and then disables upper memory), you may need to enable upper | memory yourself in order to load device drivers and programs | high. Use the MEMMAX U switches to enable and disable upper | memory. | | When you are using an EGA or VGA video display adapter, and you | reserved video memory space by configuring the memory manager, | you will need to enable the video memory before it can be used | by applications. Use the MEMMAX U switches to enable and disable | video memory. | | If you want a MEMMAX command to take effect every time you start | the operating system, place it in AUTOEXEC.BAT rather than | running it every time at the system prompt. | | Command Switches | | /? or /H Displays help text. | | -L Disables lower memory. | | +L Enables lower memory. | | /L Displays the current status of lower memory. | | -U Disables upper memory. | | +U Enables upper memory. | | /U Displays the current status of upper memory. | | -V Disables use of video memory by applications. | | +V Enables use of video memory by applications. | | /V Displays the current status of video memory. | | ---------------------------------------------------------------- | | Example | | The following lines from AUTOEXEC.BAT show how you can use | MEMMAX to control lower memory when requested: | | REM**Disable lower memory, it confuses LOGIN** | MEMMAX -L | I: | CD LOGIN | REM**Re-open lower memory** | MEMMAX +L | -------------- | | BRB with some memory stats. | | ============================================================================ | | 45) Fri, Jul 15, 2005 - 5:48:35 pm | | glennmcc> | Before using memmax to map video ram for use by applications..... | | | Memory Type Total Bytes ( Kbytes ) Available For Programs | | Conventional 655,360 ( 640K ) 581,568 ( 568K ) | Upper 61,440 ( 60K ) 10,032 ( 10K ) | High 65,520 ( 64K ) 0 ( 0K ) | Extended 67,043,328 ( 65,472K ) 0 ( 0K ) | Extended via XMS -------- 65,478,656 ( 63,944K ) | EMS 33,554,432 ( 32,768K ) 33,554,432 ( 32,768K ) | | Largest executable program: 581,184 ( 568K ) | Total Free DOS memory: 591,600 ( 578K ) | | After running memmax.com +v | | Memory Type Total Bytes ( Kbytes ) Available For Programs | | Conventional 655,360 ( 640K ) 679,872 ( 664K ) | Upper 98,288 ( 96K ) 98,288 ( 96K ) | High 65,520 ( 64K ) 0 ( 0K ) | Extended 67,043,328 ( 65,472K ) 0 ( 0K ) | Extended via XMS -------- 65,478,656 ( 63,944K ) | EMS 33,554,432 ( 32,768K ) 33,554,432 ( 32,768K ) | | Largest executable program: 679,488 ( 664K ) | Total Free DOS memory: 679,872 ( 664K ) | | | ============================================================================ | | 46) Fri, Jul 15, 2005 - 5:52:15 pm | | glennmcc> | Not too shabby. | | An additional 96kb conv and an aditional 86kb upper for a total of | 182kb more memory for text based apps to use. | | ============================================================================ | | 47) Fri, Jul 15, 2005 - 6:07:42 pm | | ray> | How do you get 96kb when the mono memory that can be stolen is | only 64k? Anyway, with that much extra memory, you've recovered | much more that the difference between our core sizes so I can't see | any reason why you wouldn't be able to debug your core too! | | If you aren't sure what to do: | | -Options/Advanced Code Generation/Debug info in OBJ's [X] | /Debugger/Source Debugging On [X] | /Display swapping -- as you like. | /Program Heap Size -- 160K works, 256 seems better. | | Now, in the project files window, ^O to options and turn off | source debugging as needed module by module to get your core small | enough to run. | | | | | | ============================================================================ | | 48) Fri, Jul 15, 2005 - 7:01:53 pm | | glennmcc> | re:"so I can't see any reason why you wouldn't | be able to debug your core too!" | | Try this as possible reason why not | | 182kb more memory for text based apps to use. | | These 'tricks' in both OpenDos and in OS/2 remap the video memory | for use by text apps. | | You can't run a graphics app with these things enabled. | | | | ============================================================================ | | 49) Fri, Jul 15, 2005 - 9:59:39 pm | | ray> | Right! But I think you've got it backwards: Seems to me that | The reason for all that memory is that you've stolen all *but* | the text mode VGA. Hmmmm, but you can still keep all the | mono memory no? Nuts I used to know every damned thing about this | and now I kint remember but that's still 64K no? But perhaps | your manager won't give you just the mono space? | | | | ============================================================================ | | 50) Sun, Jul 17, 2005 - 2:43:22 pm | | gregy> | Good grief...!!! I have been tweaking with my DRDOS load environment, | and I just achieved 182,704bytes ONLINE... For all the past 5-6 years | running Arachne, I have ONLY best been running at 112,???bytes ONLINE, | and now, finally getting EPPPD to HILOAD, I'm ONLINE with that big | 182,704bytes...sheesh..!!! | .. | I don't know what affect this will have on my surfing performance, but | it doesn't seem to be hurting it any..... | ... | Now, I'm sure that some of you "super-gurus" have even more DOS mem | free, but this is a gigantic increase for me. And, I'm sure, I can | achieve even more when I get done tweaking my DOS VDM in OS/2... | | | ============================================================================ | | 51) Sun, Jul 17, 2005 - 2:55:04 pm | | gregy> | Re: The "stolen video memory" Glenn's right. It will only be | available for use "by text based apps"....VGA apps need not apply.. | ... | This is, by the way, subject to my understanding of this being correct. | As always, THAT is NOT a given..... | .. | But, the "Dual Head" VGA cards do give OS/2 the ability to spread out | across two monitors, BUT...and this is important....I am almost 100% | certain that each head can be used for a separate window (including | a "Full Screen Window"...which means that it can be app in one session, | and debugger in the other (both VGA apps)..... | .. | And, furthermore, Ray's/Joe's suppositions that the OS/2 "Dual Monitor" | capabilities is for one VGA and one HERC (or other text card) card | would give the capabilities of running Arachne in a VDM using the VGA | card, and the debugger (in text mode) in another VDM using the HERC | card. | | | | ============================================================================ | | 52) Mon, Jul 18, 2005 - 7:32:23 pm | | gregy> | NOTICE - my ISP's backbone connections (by QWEST, who else???) are, and | have been, flakey. They're working on it, but you probably will find | these boards up-and-down intermittantly. | | | ============================================================================ | | 53) Tue, Jul 19, 2005 - 1:32:45 pm | | ray> | Gregy, | Wealth is a relative thing no? ... I've got 251K free and I think | Glenn has even more with tinyfoot or a slightly less radicle | version of it. | | | ============================================================================ | | 54) Tue, Jul 19, 2005 - 10:03:00 pm | | gregy> | Ray...you're RIGHT...!!!........I, right now, have 182,688bytes | steady, and am surfin away, but...before I could get that EPPPD to | HILOAD, I was comfortable with only 112,???bytes ONLINE....and don't | think I ever noticed the missing 70Kbytes.... | ... | And, on my ArachneOS2, I cannot get that EPPPD to loadhigh, so I'm | surfin away on it, with only 115,???bytes...and it works superbly... | ... | | | | ============================================================================ | | 55) Wed, Jul 20, 2005 - 5:11:15 pm | | glennmcc> | From my experience, anything from 100,000[+]green and on-up is just fine. | | | ============================================================================ | | 56) Wed, Jul 20, 2005 - 5:32:35 pm | | ray> | Glenn, | | Scotty's dead. | | | ============================================================================ | | 57) Wed, Jul 20, 2005 - 5:33:29 pm | | ray> | ... and didn't we agree that you need something like 150K to run | ASF's? | | | ============================================================================ | | 58) Wed, Jul 20, 2005 - 7:29:44 pm | | glennmcc> | If I'm not mistaken, Gregy runs ASFs with only 112k free. | | So I guess we were wrong on that figure. | | As to Scotty. | | Nope, not dead.... | | Simply has been transported to another dimension. | | | ____ | | RIP James DooHan | | | ============================================================================ | | 59) Wed, Jul 20, 2005 - 7:35:30 pm | URL | | glennmcc> | Ashes of famous actor James Doohan to be sent into space | | LOS ANGELES, July 20 (Xinhuanet) -- | The ashes of Star Trek's James Doohan, who died Wednesday at his home | in Redmond, Washington, will be sent into space to meet his wishes, | his agent Steven Stevens said here. | | The 85-year-old actor, who played engineer Montgomery "Scotty" Scott | in the original "Star Trek" TV series in the 1960s and the motion pictures | that it spawned, | died of pneumonia and complications from Alzheimer's disease, said Stevens. | | The actor's ashes are to be sent into space, as Doohan had wanted, | by a company called Space Services Inc., according to the agent. | | The Houston-based company teams with commercial launch organizations | in offering out-of-this-world memorial services. It has sent into space | the ashes of "Star Trek" creator Gene Roddenberry and drug guru Timothy Lear | | Meanwhile, the Hollywood Chamber of Commerce is offering a more earthly | tribute to Doohan on Wednesday afternoon as representatives will place | flowers on Doohan's star on the Hollywood Walk of Fame. | | "I don't think anyone loved his fans and appreciated his fans more than he d | said Stevens. | | ============================================================================ | | 60) Thu, Jul 21, 2005 - 4:41:58 am | | gregy> | yeah....makes me feel old...kinda....:(( | | | | ============================================================================ | | 61) Thu, Jul 21, 2005 - 4:45:07 am | | gregy> | Re:...running .ASFs....hmmmm...yes...but, AFAIR, I was running them | OFFLINE....not ONLINE.....and I had sumpin like 146K DOS mem free. | ... | hmmm....I'll hafta try running some of them ONLINE, with this ....tada.. | ...tada....check it out.....188,000bytes DOS mem free ONLINE, right now | .... | .. | This is with my new OpDOS7.01 OperatingSystem install... | | | ============================================================================ | | 62) Thu, Jul 21, 2005 - 5:10:04 am | | gregy> | Yep...just as I suspected....XMSDSK works FINEST KIND with OpDOS7.01.. | ... | Now running Arachne from a 16Mbyte XMSDSK RAMdisk...... | ...gee, progress certainly is encouraging..... | | | ============================================================================ | | 63) Fri, Jul 22, 2005 - 2:57:37 am | | Joe> | Hi all, | | Well, gregy, why not use LSPPP instead of EPPPD? Much smaller, hence | much easier to load high, more memory free ... | | As for getting less free memory with OS/2 than OpenDOS 7.01, that's | dissapointing. Why can't you load EPPPD/LSPPP high with O/S2? For | that matter, why can't you load EPPPD/LSPPP in its own virtual machine, | to get heaps more free memory? Or is that trick only available in O/S2 | for device drivers, not TSR's? I'm guessing that's why, so perhaps | magical memory for Arachne debugging won't be possible with OS/2 after | all. Sigh! | | | ============================================================================ | | 64) Fri, Jul 22, 2005 - 2:59:26 am | | Joe> | Sad to hear about Scotty ... great actor, will be missed. | | ============================================================================ | | 65) Fri, Jul 22, 2005 - 7:09:15 am | | ray> | Joe, | | Speaking of lsppp, have you any idea why it doesn't work for Vladimir | as his dialer? | | ============================================================================ | | 66) Fri, Jul 22, 2005 - 1:10:23 pm | | glennmcc> | I suspect that Vladimir has one of the 'buggy' releases of Lsppp in | which the dialer was broken. | | I don't remeber which versions are broken, | but I can say that it is fixed now. | | | ============================================================================ | | 67) Sat, Jul 23, 2005 - 2:34:49 am | | gregy> | "...dissapointing. Why can't you load EPPPD/LSPPP high with O/S2? For" | | Joe, because I have never gotten LSPPP, from it's earliest days, to | work as my dialer. I'm planning to make a major assault on that | objective just as soon as I get my new setup(s) stabilized. | | "that matter, why can't you load EPPPD/LSPPP in its own virtual machine, | to get heaps more free memory? Or is that trick only available in O/S2 | for device drivers, not TSR's? I'm guessing that's why, so perhaps..." | | Joe....you know...that possibility never even occurred to me.... | ... | AFAIK, I know nothing about such a concept. But, now that you've | mentioned it, I'm gonna check it out with some of the gurus on my OS/2 | lists..... | | | ============================================================================ | | 68) Sat, Jul 23, 2005 - 2:39:39 am | | gregy> | Joe...Re: loading EPPPD high with OS/2....I've tried that, several ways, | and it keeps stepping on my video-RAM....so far I haven't found the | proper place to make it load into so it won't do that. How do I know | it's stepping on my video...??? | I have this nifty 17" DELL-branded Sony Trinitron, with nifty control | stuff inside, and when I load EPPPD high, the screen goes blank, the | "lost signal" led on the front lights up, and it pops up a nice little | white box with "Out of Range" on it..... | ... | So, for right now, it's being loaded in conventional...until I can find | the exact UMB to load it into, that will not stomp my hardware... | | | ============================================================================ | | 69) Wed, Jul 27, 2005 - 1:49:16 pm | | Joe> | Hi all, | | Well, Iīm now on holidays in Europe ... so contact will probably | be sporatic ... | | Anyway, regarding dialling with lsppp - AFAIK, there are some | modems that wonīt dial with lsppp, _any_ version. Thatīs one of | the two reasons I use miniterm for dialling in my lspppcfg package. | | Now, if I had Vladimirīs modem to test with, Iīd probably find | out whatīs going wrong without much trouble. Trouble is, I donīt | have his modem, so thereīs no way I can tell whatīs going wrong. | Perhaps if he has the means, he can get a log of the dialogue | between lsppp and his modem, this may be enough to sort things | out - maybe. | | | ============================================================================ | | 70) Wed, Jul 27, 2005 - 2:01:00 pm | | Joe> | Gregy, | | Whatīs wrong with OS/2īs hi-loading code??? It should prevent any | TSR which doesnīt fit in a UMB from attempting to load high, right? | Thatīs what normally happens in XX-DOS. Perhaps you have a strange | configuration which bypasses some safety checks or something? Else | the TSR (ie. EPPPD) is simply crashing severely and trashing I/O | space, in particular, the video stuff (seems unlikely, but who knows?) | | As for not using lsppp because it doesnīt dial with your modem - why | not use miniterm for dialling and lsppp for the packet driver - thatīs | what Iīve done, virtually from day one. Why forgo the memory advantages | of lsppp simply because of a dialling problem? Remember, EPPPD came | with itīs own dialler, yet MP chose not to use it, developing miniterm | for this purpose instead. | | ============================================================================ | | 71) Wed, Jul 27, 2005 - 4:37:16 pm | URL | | glennmcc> | Dear fellow Arachne fans....... | | I have just now uploaded v1.85;GPL,386+ | http://www.cisnet.com/glennmcc/ | | --- on the front page --- | Arachne v1.71, v1.73, v1.75, v1.77, v1.79, v1.81 | and v1.83 are no longer available on this site. | | Available Jul 27, 2005...... Arachne v1.85;GPL,386+ | ___________________________________________________ | | Enjoy !!! :) | | ============================================================================ | | 72) Wed, Jul 27, 2005 - 8:19:49 pm | | ray> | Will grab that tommorow Glenn. | | | ============================================================================ | | 73) Thu, Jul 28, 2005 - 4:59:04 am | | gregy> | Got it installed on both my Op701 and my Warp3C...and it's working just | fine... | | | ============================================================================ | | 74) Fri, Jul 29, 2005 - 4:56:01 am | | Joe> | Cool, new code to play with! :-) | | Glenn, when convenient, could you upload the changed | source files to AD? Donīt delete the 1.83(-) stuff, of | course (I presume you delete old versions from your | own web site due to space constraints, however space | isnīt much of an issue at AD yet, particularly if we | avoid duplication). | | See ya. | | PS. BTW, I have arranged with our IT people for my | work e-mail password not to expire while Iīm on | holidays. The bad news is that I think the Base64 | problem has returned, so I wonīt be able to post | anything ledgible to freelists for the present. | | ============================================================================ | | 75) Fri, Jul 29, 2005 - 7:10:28 am | | ray> | Nuts I'm falling behind again, code from both Joe and Glenn to | look at. And I want to have some fun while it's summer; Arachne | can be a full time job. | | | | ============================================================================ | | 76) Fri, Jul 29, 2005 - 12:07:01 pm | | glennmcc> | My 4th attempt to upload the 1.85 SRCs has just failed. | | Tried twice with Linux FireFox... both attempts resulted in | "Document contains no data" | | Rebooted into DOS and tried twice with Lynx 2.85.1 | Both attempts resulted in ... "Unexpected network write failure" | | ============================================================================ | | 77) Fri, Jul 29, 2005 - 12:09:03 pm | | glennmcc> | Both attempts at Yahoo whith FireFox were with JS enabled just in case | that might be needed. | | I _have_ uploaded files in the past using Lynx in DOS | | Will try again later. | | ============================================================================ | | 78) Fri, Jul 29, 2005 - 1:49:10 pm | | ray> | My latest attempts to use AD were similar failures :( | | | | ============================================================================ | | 79) Fri, Jul 29, 2005 - 3:03:51 pm | URL | | glennmcc> | Yahoo _is_ accepting data from me via the 'create text file' option. | | Just now did this one using Arachne 1.85 | | http://f2.grp.yahoofs.com/v1/0JjqQtKoz18D-X3zts2kVSofEFRZM1ZNTA4O5fCYS6JgdgY | | | ============================================================================ | | 80) Fri, Jul 29, 2005 - 3:04:53 pm | | glennmcc> | But my 5th atempt to upload ar185src.zip has also failed :( | | ============================================================================ | | 81) Fri, Jul 29, 2005 - 3:57:39 pm | | Joe> | Hi Glenn, et al. | | Hmmm ... | | As I recall, we had a similar problem a few months ago. | I think it was OK a day or two later, so hopefully itīs | one of these temporary glitches again. | | | ============================================================================ | | 82) Fri, Jul 29, 2005 - 7:44:22 pm | | ray> | Hey can anyone get to: | | http://www.bbc.co.uk/wiltshire/entertainment/days_out/hpv_castlecombe.sht | | Eric says he had a hang on that page using Build7 but I can't load | the page at all, get "404 page not found" then get shunted to: | | http://www.bbc.co.uk/ | | which loads just fine. | | | ============================================================================ | | 83) Sat, Jul 30, 2005 - 12:23:59 am | | gregy> | Nah...that link 404s on me, too....so Eric gave you a wrong link...get | the correct one from him.... | | | ============================================================================ | | 84) Sat, Jul 30, 2005 - 8:24:07 am | | ray> | Thanks Gregy | | | ============================================================================ | | 85) Sat, Jul 30, 2005 - 8:41:40 am | | ray> | Glenn, | | Just to avoid duplicated effort, I was thinking of changing the | mail download code so that if you press ESC or the big X, it will | delete any sucessfuly downloaded emails from the server so's you | don't start again at the next download. In your work on the spam | bug did you contemplate any such thing? I ask since during my | investigations, it looked like a bug fix and the above would be | very closely related, one riding on the other so to speak. | | | ============================================================================ | | 86) Sat, Jul 30, 2005 - 10:18:28 am | | Joe> | Ray, | | The 404 is simple - you`re missing "ml" at the end, as in "shtml". | | And ... I can tell you that the page loads perfectly in 1.83J1. | | | ============================================================================ | | 87) Sat, Jul 30, 2005 - 1:08:28 pm | | ray> | Joe, | | Thanks, that's rather obvious with hindsight. Renders fine here too. | | | | ============================================================================ | | 88) Sat, Jul 30, 2005 - 2:01:58 pm | | ray> | BTW, possible bug report: Going to that page, pressing F6 then F7, | entering search word, then 'enter' hangs with my core and with 1.83. | Confirmations? | | ============================================================================ | | 89) Sat, Jul 30, 2005 - 5:20:11 pm | | ray> | Joe, | | 1J183 Up and running. Speed check is as before: yours maxes out at | 3K, but mine and Glenn's at 4.8K. This is the .exe as downloaded, not | build here. | | | ============================================================================ | | 90) Sat, Jul 30, 2005 - 6:04:57 pm | | ray> | Joe, | | Grabbed src183c1.zip, now if I merge that with the 04-09-12 release | of B3c will I be OK? It looks that way but better to check. | First impressions are good. I'm esp. interested in that breakout | code that you've attached to ctrl-alt-esc. I tried something | with the same goal, but it resulted in a huge slowdown. | | | | ============================================================================ | | 91) Sat, Jul 30, 2005 - 10:16:41 pm | | glennmcc> | Tried again to no availe... Upload attempts 6 and 7 also failed :( | | ============================================================================ | | 92) Sun, Jul 31, 2005 - 8:35:37 am | | ray> | Glenn, I'll take your silence as a negative as per normal. | | ============================================================================ | | 93) Sun, Jul 31, 2005 - 2:03:02 pm | | glennmcc> | Silence about what ???? | | Hold on a sec, scrolling back to which question of yours I overlooked. | | | ============================================================================ | | 94) Sun, Jul 31, 2005 - 2:04:12 pm | | glennmcc> | OK, got it now.... | | This one ? | | "mail download code so that if you press ESC or the big X, it will" | | BRB | | | | | ============================================================================ | | 95) Sun, Jul 31, 2005 - 2:16:11 pm | | glennmcc> | Do you mean this, that I was working on 2 months ago ? | | ____________________________________________________________________________ | To: "Arachne4DOS" | From: "Glenn McCorkle" | Date: Thu, 26 May 2005 01:10:33 -0400 | Reply-To: arachne4dos@coollist.com | Subject: [a4dos-list] Re: bad mail | Precedence: list | Mailing-List: arachne4dos@coollist.com | X-Coollist-User: glennmcc@cisnet.com | Content-Transfer-Encoding: 7bit | MIME-Version: 1.0 | Content-Type: text/plain; charset=ISO-8859-1 | | On ?, Bastiaan T.Edelman wrote: | | > On 23 May 2005 13:38:38 +0200, G J Feig wrote: | | >> On Sun, 22 May 2005 11:12:16 -0600 (MDT), Glenn Gilbreath Jr. wrote: | | >>> Ray, | >>> I have been under the impression that Arachne downloads emails | >>> 1 at a time ever since I first fired up version 1.1 back in 1997...but I | >>> couldn't use her "online" since the only real internet connection I had | >>> available at the time was AOHell...BUT, when I did get her online with | >>> FreeWWWeb in 98, it sure looked like Arachne would fetch each | >>> email individually, number them, | | >> Wizzy, et al, | | >> If what I'm seeing on my bottom status bar, during email download, | >> is what is really happening, then I confirm that Arachne downloads | >> one message, sends a delete, downloads the next, sends a delete, | >> and so forth. | | > Nope, the most anoying lies in the fact that Arachne is not deleting the | > messages one by one! | > If a bad mail, say message #6, is blocking the download and you stop the | > downloading for the removal of message #6 with telnet or another browser | > the next time you download the mail... messages #1 to #5 are still at | > the provider. | > So after the second download the first messages are now twice in your | > inbox. And if you are not accustomed to the e-mail blocking you may find | > them 3 or more times. | > A few weeks ago #68 out of 80 e-mails was the offending one. | > They were download 3 times before I could kill #68 => 204 e-mails to | > sort!!! | | There seems to be some confusion as to "who's decision it is" (so to speak) | as to wheather to 'delete when told to' or 'wait till later'. | | Some SMTP servers are setup to 'delete now', while others are setup to | 'mark for deletion, but do not delete till receiving the "QUIT" command'. | | The server at my ISP (mail.cisnet.com), is now setup to ... | 'mark for deletion, but do not delete till receiving the "QUIT" command'. | | In the past (2 years or so ago), it was setup to 'delete now'. | | Arachne and all other pop3 clients, can't change this. | All they can do is to send this command... | | DELE message_number | | It now depends upon the server settings as to wheather it gets deleted | now or simply marked for deletion awaiting the QUIT command. | | Now, be that as it may...... | | It might be a good idea to have the option of having already downloaded | messages to be deleted when we abort a pop3 session via the ESC key. | | I will look into adding the 'QUIT command' | to the ESC key section of the code. | | -- | Glenn | http://www.delorie.com/listserv/mime/ | http://www.cisnet.com/glennmcc/ | http://www.law.cornell.edu/constitution/constitution.table.html | | | ============================================================================ | | 96) Sun, Jul 31, 2005 - 2:19:40 pm | | glennmcc> | If that's what you're refering to. | | I worked on it for a little over a week, | | had no success at-all, | | abandoned the whole idea, | | deleted everything I tried from the SRC code files. | | | ============================================================================ | | 97) Sun, Jul 31, 2005 - 2:33:31 pm | | glennmcc> | 2 more failed attempts just now to upload ar185src.zip to Yahoo :(( | | ============================================================================ | | 98) Sun, Jul 31, 2005 - 6:06:18 pm | | ray> | Ya, that's what I was refering to, I'll give it a try and post my | efforts, as I'm imagining it, it should be easy, but .... ;-) | But first I should look at your bad email code to make sure that I | don't do something that would bugger that up. | | BTW, I still can get on to Yahoo neither, except via W$ | | ============================================================================ | | 99) Sun, Jul 31, 2005 - 6:08:07 pm | | ray> | Damn, so much to do and I've gotta go diving while the weather's nice. | | | ============================================================================ | | 100) Sun, Jul 31, 2005 - 7:15:29 pm | | glennmcc> | re"BTW, I still can get on to Yahoo neither, except via W$ " | | ???? | | I can get onto Yahoo just fine with Arachne. (and all other browsers) | | Uploading of files is the problem. | | Tried uploading ar185src.zip to our group with all of these to no avail.... | | -- in DOS --- | Lynx 2.85.1 | | -- in Linux --- | FireFox | Mozilla | | --- in Win98 --- | | IE | NS | FireFox | Mozilla | Opera | | ___________ | | No go with any of them. :( | | ============================================================================ | | 101) Sun, Jul 31, 2005 - 9:55:06 pm | | ray> | That's what I was trying to say actualy <:-) -- I can log on, just | not move any files. | | | ============================================================================ | | 102) Sun, Jul 31, 2005 - 9:56:29 pm | | ray> | We need a better upload site, yours would be fine or perhaps Jake's. | | | | ============================================================================ | | 103) Sun, Jul 31, 2005 - 10:13:30 pm | | gregu> | We have 150Mbyte upload site, if you guys will use FTP on it...but, | unfortunately, hTTP-upload is out... | ... | | | ============================================================================ | | 104) Sun, Jul 31, 2005 - 10:17:19 pm | | gregy> | And, I know Joe is behind routers, etc, so he can't do FTPupload...but.. | ..hmmmm. | .. | Maybe we could do it this way: | 1. You guys upload to the FTP site. | 2. Both FTP and HTTP downloads are possible from there | 3. Joe can upload to yahoo, then someone can HTTPdownload it and | FTPupload to the FTPupload site. | ... | At least there would be an alternate for posting code. | .. | You could use this site, too, but some people have had trouble FTP | uploading to this site, and, my ISP only gives me 5Mbytes storage, | so that would fill up pretty fast. | | | ============================================================================ | | 105) Mon, Aug 01, 2005 - 7:51:16 am | | Joe> | Hi all, | | I think I`ve got the AD problem sorted - there seems to be an upload | size limit, perhaps 512K. I`ve uploaded an abridged version of | Glennīs 1.85 sources (just quickly selected the newest files) and | that uploaded fine (the full original ZIP failed for me too). | | | ============================================================================ | | 106) Mon, Aug 01, 2005 - 9:01:22 am | | ray> | Joe, | | So what about my question re your latest sources? Do I merge them | with the last 3c upload? | | ============================================================================ | | 107) Tue, Aug 02, 2005 - 4:40:53 pm | | Joe> | Hi Ray, | | Sorry, canīt be definitive about merging with 04-09-12 B3C | code ... working with limited tools at the moment, so Iīm not | clear about the version numbers. However, just check for any | 1xxCx (eg. 177C1, 183C1) uploads at AD, these all are updates | of the original B3C code, so these all need to be applied, | unless the 04-09-12 designation implies some of these updates | are already implemented. Heck, I thought your plan was to merge | the 1xxCx stuff with your present B5 code, in which case you would | certainly have a better idea of what youīve merged already. If | you plan to go back to the B3C series, then itīs already done, | itīs the 1xxCx files. | | What does all this mean? Basically, go ahead and merge, unless thereīs | older 1xxCx that you need to merge, in which case you should do that | first. Hope that helps some. Sorry, but my computer is at the other | end of the globe and Iīm doing what I can with my sisterīs computer, | some tools Iīve installed on it, the rest being the standard Windoze | stuff. | | BTW, I am sad to say that my sister only knows some Windoze stuff, | being totally ignorant of DOS. The fact that Iīve installed Arachne | on her PC, is simply because I like to test Arachne on as many PCīs | as I can - also I feel much safer without having to worry about Windoze | viruses and worms (when I got here, my sisterīs PC was infected with | four different viruses, affecting almost 2400 files!). | | Note that I will probably be moving around a bit more in the near | future, so even my sisterīs PC will be out of reach, and contact | will become much more difficult. | | ============================================================================ | | 108) Tue, Aug 02, 2005 - 5:09:17 pm | | ray> | Joe, | | Thanks for that, I understand your inability to 'say for sure'. | I'm synchronised with you up to 1.79 so if I understand your system, | that's all there is. This file synch, stuff can get out of hand, sheesh | I have trouble remembering what's going on in my own uploads. I'm | becoming more disiplined with that, however. | | | | ============================================================================ | | 109) Tue, Aug 02, 2005 - 5:10:04 pm | | ray> | BTW, I've found two bugs, not big ones, but bugs NTL. More later. | | | ============================================================================ | | 110) Tue, Aug 02, 2005 - 7:44:26 pm | | glennmcc> | In your stuff, or mine ? | | ============================================================================ | | 111) Tue, Aug 02, 2005 - 7:47:14 pm | | ray> | I think its old code one is ... no both are in files I'm playing with | ongoing so ... na, I'd better show you now, but first leme check | your files BRB. | | | ============================================================================ | | 112) Tue, Aug 02, 2005 - 8:05:24 pm | | ray> | Yup, it's old code. Too long for here, I'll post it to AD | | ============================================================================ | | 113) Tue, Aug 02, 2005 - 8:09:09 pm | | ray> | Opps, I forgot the fix for bug #2 <:-) | | ============================================================================ | | 114) Tue, Aug 02, 2005 - 8:12:17 pm | | ray> | But it's easy: | | ArachneDial(char *buffer) // takes a pointer, not void as before. | And it returns "return buffer". | Then, forget 'buf' in that function, instead create 'buf' below, | where ArachneDial() is called by | | This sux ... BRB | | | | ============================================================================ | | 115) Tue, Aug 02, 2005 - 8:17:06 pm | | ray> | | Base6 code: | (pardon my strange defines like 'DeF' and 'Eif' | | static char *ArachneDial DeF (char *buffer) // BUG! converted from '(void)' | { | // BUG! MHT reports that automatic arrays shouldn't be used to pass strings | // since they are not protected once their function returns! | // char buf[5 * IE_MAXLEN]; // Original | | ... | | return buffer; | } | | ... lower down: | | int ProtocolArachne DeF (struct HTTPrecord *sCacheItem, struct Url *url, | int *returnvalue) | { | | ... | | Eif (!strcmpi(url->file, "dialer")) | { | // RAY: 05-08-01: BUG! fix explained above: "MHT reports" ... | char buffer[80]; | *returnvalue = RouraExecute(ArachneDial(buffer)); | return GOTO_END; | } | | ============================================================================ | | 116) Tue, Aug 02, 2005 - 9:12:10 pm | | ray> | Ha! Almost got it, after an "X" she won't start from scratch on the | next download, one problem remaining is that mail in progress at the | time of the chop is chopped. Now, that would kill it if a spam I'm | quite sure, but suppose you've just decided to discontinue a download | but don't want to cut anything in half?? Hmmmm .... | | | ============================================================================ | | 117) Wed, Aug 03, 2005 - 8:14:38 pm | | ray> | How do you send the command to abort a RETR ? | On user break, when she goes to send QUIT, there is garbage | returned from the socket sometimes, apart from that it seems to work. | | | | ============================================================================ | | 118) Thu, Aug 04, 2005 - 6:11:02 pm | URL | | glennmcc> | Sorry to say it don't look good. :( | | Just now sent this reply to a4dos mailing list. | | From what I've been able to find in the RFCs.... no can do. | | As we can see here.... http://rfc1939.x42.com/ | | There is no provision in the POP3 RFC for the server to accept a command | while in the middle of a RETR | | | | ============================================================================ | | 119) Thu, Aug 04, 2005 - 6:17:25 pm | | ray> | testing | | | ============================================================================ | | 120) Thu, Aug 04, 2005 - 6:19:14 pm | URL | | glennmcc> | 9. POP3 Command Summary | | Minimal POP3 Commands: | | valid in the AUTHORIZATION state | | USER name | PASS string | QUIT | | valid in the TRANSACTION state | | STAT | LIST [msg] | RETR msg | DELE msg | NOOP | RSET | QUIT | | Optional POP3 Commands: | | APOP name digest valid in the AUTHORIZATION state | | TOP msg n valid in the TRANSACTION state | | UIDL [msg] | | POP3 Replies: | +OK | -ERR | | ______________________ | | | That's all there is, there ain't no more. | | No 'ABORT' command exists. | | | ============================================================================ | | 121) Thu, Aug 04, 2005 - 6:19:17 pm | | ray> | Hmmm, can't attach a file here, why not? | | | ============================================================================ | | 122) Thu, Aug 04, 2005 - 6:19:57 pm | | glennmcc> | | | Hi Ray ;-) | | | | ============================================================================ | | 123) Thu, Aug 04, 2005 - 6:20:46 pm | | glennmcc> | How and what, are you trying to 'attach' ? | | | ============================================================================ | | 124) Thu, Aug 04, 2005 - 6:20:47 pm | | ray> | Leme shorten it: | | Gregy, all: | | Speaking of 'won't connect', thought I'd bounce an idea off you: | I've replaced my last experimental dialer scheme with what I think | is the final word on the subject of alternate dialer schemes. | | | ============================================================================ | | 125) Thu, Aug 04, 2005 - 6:22:40 pm | | ray> | Sheesh, what's our size limit here? | | ============================================================================ | | 126) Thu, Aug 04, 2005 - 6:23:05 pm | | glennmcc> | 20kb | | | ============================================================================ | | 127) Thu, Aug 04, 2005 - 6:23:23 pm | | glennmcc> | Are you using Arachne ? | | ============================================================================ | | 128) Thu, Aug 04, 2005 - 6:23:49 pm | | ray> | ... | Once you cut through all the layers, at bottom, Arachne doesn't | dial or connect, she creates a batchfile that does it by calling | external programs. She attempts to make this process user friendly by | insulating you from what's really going on via arachne.cfg keywords | which are then used as raw material for the construction of the | batchfile. As with Windows, this is fine as long as your needs and | choices fall within the range of what the batchfile construction | code thinks you should do. If not, tough. So my alternate | code replaces all this with the direct aproach: you create a batchfile | by hand, and then the keyword 'connection' does nothing but | execute that batchifle. Period. No smoke, no mirrors. The batchfile | can do anything you want, use any dialer/ppp driver combination that | you want. | | | ============================================================================ | | 129) Thu, Aug 04, 2005 - 6:25:05 pm | | ray> | 20kb? Yeah, using Arachne as always. So I should be able to post | 20K in one bite? | | | ============================================================================ | | 130) Thu, Aug 04, 2005 - 6:26:24 pm | | ray> | It can include various tests and try different phonenumbers | automaticaly. Vladimir is testing it and his 'connect.bat' is now | about 20 lines long and among other things, it tries several DNS's | if the first one fails. | | I like it, it's DOSey not Windowsey. | | ============================================================================ | | 131) Thu, Aug 04, 2005 - 6:27:27 pm | | glennmcc> | I do it simlarly, but much simpler. | | No changes to core.exe.... just to arachne.cfg | | Connection @lsppp.exe /?>NUL | Dialer @lh lsppp.exe>ppp.log | | | --- in lsppp.cfg --- | /M:at&fs7=90s11=50 | /x | /i:4 | /b:bx800 | /d:330-423-4005 | /B:115200 | /V:60 | /U:glennmcc@cisnet.com | /P:******** | ____________________ | | | Simple, straight forward... works perfectly. :) | | | ============================================================================ | | 132) Thu, Aug 04, 2005 - 6:29:40 pm | | glennmcc> | If you can't seem to get your core to post something that big here. | | Try the realease version of mine. | | --- in changes.txt --- | | Jul 16, 2005 | Fixed an intermitant problem with posting of forms on some web pages. | (postings and deletions of messages at mail.yahoo.com now work perfectly) | | _______________________ | | | | | ============================================================================ | | 133) Thu, Aug 04, 2005 - 6:30:04 pm | | glennmcc> | Out dinner .... back in 45min | | ============================================================================ | | 134) Thu, Aug 04, 2005 - 6:30:20 pm | | ray> | That's exactly how I do it too, but you make my point for me: you and | I can do things the 'normal' way with no problems, but for folks | who need a bit more customisation, all the stuff in ArachneDial() | is just in the way, since it forces $roura$.bat to do things the | user might not like. | | ============================================================================ | | 135) Thu, Aug 04, 2005 - 6:30:38 pm | | glennmcc> | oops, .... out _to_ dinner | | | ============================================================================ | | 136) Thu, Aug 04, 2005 - 6:31:23 pm | | ray> | OK, try that file via 1.83: | | Gregy, all: | | Speaking of 'won't connect', thought I'd bounce an idea off you: | I've replaced my last experimental dialer scheme with what I think | is the final word on the subject of alternate dialer schemes. | Once you cut through all the layers, at bottom, Arachne doesn't | dial or connect, she creates a batchfile that does it by calling | external programs. She attempts to make this process user friendly by | insulating you from what's really going on via arachne.cfg keywords | which are then used as raw material for the construction of the | batchfile. As with Windows, this is fine as long as your needs and | choices fall within the range of what the batchfile construction | code thinks you should do. If not, tough. So my alternate | code replaces all this with the direct aproach: you create a batchfile | by hand, and then the keyword 'connection' does nothing but | execute that batchifle. Period. No smoke, no mirrors. The batchfile | can do anything you want, use any dialer/ppp driver combination that | you want. It can include various tests and try different phonenumbers | automaticaly. Vladimir is testing it and his 'connect.bat' is now | about 20 lines long and among other things, it tries several DNS's | if the first one fails. | | I like it, it's DOSey not Windowsey. | | | ============================================================================ | | 137) Thu, Aug 04, 2005 - 6:33:01 pm | | ray> | Hmmm, no problem with 1.83. | | Sorry guys, gotta post this a hundred times till I isolate it ... | | | Gregy, all: | | Speaking of 'won't connect', thought I'd bounce an idea off you: | I've replaced my last experimental dialer scheme with what I think | is the final word on the subject of alternate dialer schemes. | Once you cut through all the layers, at bottom, Arachne doesn't | dial or connect, she creates a batchfile that does it by calling | external programs. She attempts to make this process user friendly by | insulating you from what's really going on via arachne.cfg keywords | which are then used as raw material for the construction of the | batchfile. As with Windows, this is fine as long as your needs and | choices fall within the range of what the batchfile construction | code thinks you should do. If not, tough. So my alternate | code replaces all this with the direct aproach: you create a batchfile | by hand, and then the keyword 'connection' does nothing but | execute that batchifle. Period. No smoke, no mirrors. The batchfile | can do anything you want, use any dialer/ppp driver combination that | you want. It can include various tests and try different phonenumbers | automaticaly. Vladimir is testing it and his 'connect.bat' is now | about 20 lines long and among other things, it tries several DNS's | if the first one fails. | | I like it, it's DOSey not Windowsey. | | | ============================================================================ | | 138) Thu, Aug 04, 2005 - 7:09:47 pm | | ray> | testing | | | ============================================================================ | | 139) Thu, Aug 04, 2005 - 7:12:24 pm | | ray> | Gregy, all: | | Speaking of 'won't connect', thought I'd bounce an idea off you: | I've replaced my last experimental dialer scheme with what I think | is the final word on the subject of alternate dialer schemes. | Once you cut through all the layers, at bottom, Arachne doesn't | dial or connect, she creates a batchfile that does it by calling | external programs. She attempts to make this process user friendly by | insulating you from what's really going on via arachne.cfg keywords | which are then used as raw material for the construction of the | batchfile. As with Windows, this is fine as long as your needs and | choices fall within the range of what the batchfile construction | code thinks you should do. If not, tough. So my alternate | code replaces all this with the direct aproach: you create a batchfile | by hand, and then the keyword 'connection' does nothing but | execute that batchifle. Period. No smoke, no mirrors. The batchfile | can do anything you want, use any dialer/ppp driver combination that | you want. It can include various tests and try different phonenumbers | automaticaly. Vladimir is testing it and his 'connect.bat' is now | about 20 lines long and among other things, it tries several DNS's | if the first one fails. | | I like it, it's DOSey not Windowsey. | | ============================================================================ | | 140) Thu, Aug 04, 2005 - 7:13:10 pm | | ray> | testing \!5c\.e30: | | Gregy, all: | | Speaking of 'won't connect', thought I'd bounce an idea off you: | I've replaced my last experimental dialer scheme with what I think | is the final word on the subject of alternate dialer schemes. | Once you cut through all the layers, at bottom, Arachne doesn't | dial or connect, she creates a batchfile that does it by calling | external programs. She attempts to make this process user friendly by | insulating you from what's really going on via arachne.cfg keywords | which are then used as raw material for the construction of the | batchfile. As with Windows, this is fine as long as your needs and | choices fall within the range of what the batchfile construction | code thinks you should do. If not, tough. So my alternate | code replaces all this with the direct aproach: you create a batchfile | by hand, and then the keyword 'connection' does nothing but | execute that batchifle. Period. No smoke, no mirrors. The batchfile | can do anything you want, use any dialer/ppp driver combination that | you want. It can include various tests and try different phonenumbers | automaticaly. Vladimir is testing it and his 'connect.bat' is now | about 20 lines long and among other things, it tries several DNS's | if the first one fails. | | I like it, it's DOSey not Windowsey. | | | ============================================================================ | | 141) Thu, Aug 04, 2005 - 7:15:12 pm | | ray> | Testing above core in \6\.e30: | | Gregy, all: | | Speaking of 'won't connect', thought I'd bounce an idea off you: | I've replaced my last experimental dialer scheme with what I think | is the final word on the subject of alternate dialer schemes. | Once you cut through all the layers, at bottom, Arachne doesn't | dial or connect, she creates a batchfile that does it by calling | external programs. She attempts to make this process user friendly by | insulating you from what's really going on via arachne.cfg keywords | which are then used as raw material for the construction of the | batchfile. As with Windows, this is fine as long as your needs and | choices fall within the range of what the batchfile construction | code thinks you should do. If not, tough. So my alternate | code replaces all this with the direct aproach: you create a batchfile | by hand, and then the keyword 'connection' does nothing but | execute that batchifle. Period. No smoke, no mirrors. The batchfile | can do anything you want, use any dialer/ppp driver combination that | you want. It can include various tests and try different phonenumbers | automaticaly. Vladimir is testing it and his 'connect.bat' is now | about 20 lines long and among other things, it tries several DNS's | if the first one fails. | | I like it, it's DOSey not Windowsey. | | | | ============================================================================ | | 142) Thu, Aug 04, 2005 - 7:20:12 pm | | ray> | tesing | | ============================================================================ | | 143) Thu, Aug 04, 2005 - 7:21:35 pm | | glennmcc> | re:" since it forces $roura$.bat to do things the user might not like." | | What do you see here that the user might not like ??? | | --- $roura$.bat 'undeleted' after dialing with miniterm.exe --- | | @MINITERM.EXE>PPP.LOG | @if errorlevel 1 goto skip | @lh epppd.exe>>PPP.LOG | :skip | @if exist IP-UP.BAT call IP-UP.BAT | @echo PPPD status: IP address set to %MYIP%>>PPP.LOG | @arachne.bat -r | _____________________________________________________ | | | ============================================================================ | | 144) Thu, Aug 04, 2005 - 7:24:28 pm | | ray> | Hard to say until one bumped into some problem, maybe the errorlevel | test? Anyway, the point isn't so much that the existing bat is 'bad' | but that it's limited. I'll send you Vladimir's last dialer.bat if | I can still dig it up and you'll see what I mean. | | | | ============================================================================ | | 145) Thu, Aug 04, 2005 - 7:31:55 pm | | glennmcc> | Ray, my man........... | | As stated in changes.txt | | Jul 16, 2005 | Fixed an intermitant problem with posting of forms on some web pages. | _________^^^^^^^^^^^___<-- the problem is intermitant | | 100 repostings or a thousand won't isolate it. | | It does not happen every time. | And it does not happen in a reproducable manner. | | We can do the exact same post 100 times under exactly the same | circumstances and it will work fine. | On attempt #101 it fails and #102 works fine again, #103 to 10,562 | all work perfectly but #10,563 fails. | | It took me 3yrs and many, many thousands of failed posts | to finally isolate the problem on July 16th. | | That's what fixes it.... my July 16th change in v1.85 | | ============================================================================ | | 146) Thu, Aug 04, 2005 - 7:33:53 pm | | glennmcc> | re:"...I'll send you Vladimir's last dialer.bat " | | Where did 'dialer.bat' come from ? | | Is it 'custom' batch file that Vladimir constructed himself ? | | | | ============================================================================ | | 147) Thu, Aug 04, 2005 - 7:37:52 pm | | ray> | Ah, well thanks, I think you just saved me about 100 hours of | this sort of thing. God Damn all intermitent errors! | I was just noticing that a build that worked once wouldn't work the | next time, so I'd then start looking at other files which would take | a week :-(. Let me run off and see what you did ... | | | ============================================================================ | | 148) Thu, Aug 04, 2005 - 7:49:01 pm | | ray> | The way this experimental code works is that 'connection' points | to the batchfile of your choice ... come to think of it, that's | no change ... it's just that you do everything in that batchfile, | the IP-UP stuff and the line to write to the log are all gone, and | there's no parsing of 'dialer' since all that stuff is done in the | one file. Next thing is to make a dialer page that let's you pick one | of possibly several dialer batchfiles and maybe even edit them on | screen within arachne. | | ============================================================================ | | 149) Thu, Aug 04, 2005 - 7:51:13 pm | | ray> | Hey, I just looked at 1.85, searched for "July 16," and saw nothing | about frames, which file? ... never mind why don't I just go and | get changes.txt? Hell, I need it anyway ;-) | | ============================================================================ | | 150) Thu, Aug 04, 2005 - 7:58:42 pm | | glennmcc> | as to the error level | | Miniterm exits with errorlevel 1 upon successfully dialing and connrecting. | | errorlevel 0 == failed | | Try this as 'manually' run bat file. | | --- dailtest.bat --- | @MINITERM.EXE | @if errorlevel 0 echo failed | @if errorlevel 1 echo succeeded | @pause | @if errorlevel 1 goto skip | @lh epppd.exe>>PPP.LOG | :skip | @if exist IP-UP.BAT call IP-UP.BAT | @echo PPPD status: IP address set to %MYIP%>>PPP.LOG | @arachne.bat -r | _______________________ | | | ============================================================================ | | 151) Thu, Aug 04, 2005 - 8:01:30 pm | | ray> | Na, still can't find it, you comment the fix but not the file. | (It's not that stuff in HTML.C is it? .. :-/ ... na, that's color | stuff). | | ============================================================================ | | 152) Thu, Aug 04, 2005 - 8:02:00 pm | | glennmcc> | re:".... searched for "July 16," and saw nothing..." | | --- in changes.txt --- | Jul 16, 2005 | | | 'Jul 16' not 'July 16' | | ============================================================================ | | 153) Thu, Aug 04, 2005 - 8:04:24 pm | | ray> | Sure, that's fine IF you're using miniterm. Anyway this new code | won't replace the old, it's just an alternate way of connecting that | give the user total control. Speaking of miniterm, someone must have | those sources dag-nabit. Clarence, Bernie? Ivan Polack? | | | ============================================================================ | | 154) Thu, Aug 04, 2005 - 8:05:22 pm | | ray> | I found it in 'changes.txt' cronologicaly, but which source file? | | ============================================================================ | | 155) Thu, Aug 04, 2005 - 8:11:08 pm | | ray> | Hey, I see you gave up on the smaller stack. Can't you get even a bit | smaller? I'm still using 0x6000 and it seems fine. I did lighten up | on some automatic variables but not *that* much, you should still be | able to get close to me. Which reminds: no comment on those small bugs? | | | ============================================================================ | | 156) Thu, Aug 04, 2005 - 8:16:56 pm | | ray> | Anyway, so here's the alternate connetion code, it's nothing more | than the old stuff stripped down to the essentials: | | | static char *ArachneDial DeF (char *buffer) // BUG! converted from '(void)' | { | if (sUI.altconnection) | { | if (!sUI.connection) | exit(137); | | sprintf(buffer, "%Ns\n", sUI.connection); | | if ((!strcmpi(sUI.dialpage, p->htmlframe[p->activeframe].sCacheItem.URL) | || strstr(p->htmlframe[p->activeframe].sCacheItem.URL, "err_")) | && sArachne.scriptline == 0) | strcat(buffer, "@arachne.bat -o\n"); // Ends up at homepage, or shows er | else // see INIT.C c. L 630 and 735. | strcat(buffer, "@arachne.bat -r\n"); // Reload previous page. | | Outsf(cDIAL, cESC); | // unlink("PPP.LOG"); // Must delete to reset online timer. | return buffer; | } | | | ============================================================================ | | 157) Thu, Aug 04, 2005 - 8:17:56 pm | | ray> | Hmmm, no problem there. Anyway, it's coffee time, G'nite Glenn. | | | ============================================================================ | | 158) Thu, Aug 04, 2005 - 8:28:27 pm | | glennmcc> | SHIT !!! | | Just now back from looking. | I forgot to comment the change in HTTP.C :( | | while(postindex+16 | OK, try again..... there were 2 changes, mnot just one :( | | //!!glennmcc Jul 16, 2005 -- fix intermitant posting problem | while(postindex+16 | Ray, Glenn, et al...Re: Ray's call to Gregy, all....about separating out | the Arachne dialup code, so that we have a "free-standing" DUN rather | than the, now, automagical inter-twined Arachne dialup... | .. | I've been griping about this for forever. I could NOT get miniterm to | dialup, pass to EPPPD, then negotiate, then load the packet driver, then | HOLD THE CONNECTION, so I could launch whatever DOS-Inet proggy I might | want to (including Arachne).... | ... | Now that Ray has explained what is going on, I think I see how to do it, | and I think I see more clearly the "inter-mingling" of the two different | functions. | ... | As far as "standard" techniques goes, the separation of the DUN function | from the actual browser function is more standard than the way we now | have it. | ... | If...IF...IF...Ray's direction can be implemented WITHOUT breaking | everything else (or, even, ANYTHING else), then good...go for it... | ... | But...I will complain if it is not backward compatible....or, at least, | as understandable as what we now have. | ... | If Ray's "improved" cores bugger things about, so that I (or any other | upgrading user, or any newUser) have to fiddle around with my .CFG | settings (as was the case with his initial two cores he send for | testing)...then I would be "not in favor"....as I don't intend to keep | on resetting stuff in my .CFGs, just because the core-builder decided to | set some defaults to whatever he thought might be cool.... | | | ============================================================================ | | 161) Fri, Aug 05, 2005 - 7:22:20 am | | ray> | Gregy, | | What I'm proposing would be in addition to the standard method, it | would not replace it unless it turned out to be simple enough to | implement that there was no longer any need for the (what I think | Sam H. would call the 'dumbo' method). | | As for config changes, combatablitity is always a virtue but sometimes | changes gotta happen. And remember, my keywords and Glenn's keywords | can live together without fighting. It's unfortunate that changes of | this sort are so anoying. | | | | ============================================================================ | | 162) Fri, Aug 05, 2005 - 7:23:45 am | | ray> | Glenn, | | Rodger those changes, thanks. | | | | ============================================================================ | | 163) Fri, Aug 05, 2005 - 7:42:35 am | | gregy> | yes, especially if they turn the user-base OFF.... | | | ============================================================================ | | 164) Fri, Aug 05, 2005 - 1:21:32 pm | | glennmcc> | Ray, | | Re: the keywords living together. | | Don't forget.... there is a 256 line limit to arachne.cfg | | And that line limit includes blank lines and commented-out lines. | | Anything over 256 and this is what happens........ | | Error reading arachne.cfg: | File not found, file is longer than 256 lines or out of memory | Type 'arachne -u' to terminate SLIP/PPP connection. | | | | ============================================================================ | | 165) Fri, Aug 05, 2005 - 3:00:08 pm | | ray> | Glenn, rodger the 256 -- of course that can be changed with no problem. | Now to try your fix: | | | | | ============================================================================ | | 166) Fri, Aug 05, 2005 - 3:02:38 pm | | ray> | Nuts, she doesn't work | | ============================================================================ | | 167) Fri, Aug 05, 2005 - 3:04:52 pm | | ray> | But instead of hanging, it posts and reloads the page, but there's | no new text. | | ============================================================================ | | 168) Fri, Aug 05, 2005 - 3:05:47 pm | | ray> | However, I noted in the 185 sources the 3d change: "postindex+=16" | should that be? | | | ============================================================================ | | 169) Fri, Aug 05, 2005 - 3:07:27 pm | | ray> | Gregy, all: | | Speaking of 'won't connect', thought I'd bounce an idea off you: | I've replaced my last experimental dialer scheme with what I think | is the final word on the subject of alternate dialer schemes. | Once you cut through all the layers, at bottom, Arachne doesn't | dial or connect, she creates a batchfile that does it by calling | external programs. She attempts to make this process user friendly by | insulating you from what's really going on via arachne.cfg keywords | which are then used as raw material for the construction of the | batchfile. As with Windows, this is fine as long as your needs and | choices fall within the range of what the batchfile construction | code thinks you should do. If not, tough. So my alternate | code replaces all this with the direct aproach: you create a batchfile | by hand, and then the keyword 'connection' does nothing but | execute that batchifle. Period. No smoke, no mirrors. The batchfile | can do anything you want, use any dialer/ppp driver combination that | you want. It can include various tests and try different phonenumbers | automaticaly. Vladimir is testing it and his 'connect.bat' is now | about 20 lines long and among other things, it tries several DNS's | if the first one fails. | | I like it, it's DOSey not Windowsey. | | ============================================================================ | | 170) Fri, Aug 05, 2005 - 3:08:33 pm | | glennmcc> | Nope... that's a 'while loop' | | while(postindex+16 | Ha! on a hunch I made a 3d change "sock_fastwrite ... 16" (from 512) | and she's good. Hey Glenn, hows about some comments there! What | does this stuff do? Looks to me like your just uploading stuff in | smaller bites. | | | Gregy, all: | | Speaking of 'won't connect', thought I'd bounce an idea off you: | I've replaced my last experimental dialer scheme with what I think | is the final word on the subject of alternate dialer schemes. | Once you cut through all the layers, at bottom, Arachne doesn't | dial or connect, she creates a batchfile that does it by calling | external programs. She attempts to make this process user friendly by | insulating you from what's really going on via arachne.cfg keywords | which are then used as raw material for the construction of the | batchfile. As with Windows, this is fine as long as your needs and | choices fall within the range of what the batchfile construction | code thinks you should do. If not, tough. So my alternate | code replaces all this with the direct aproach: you create a batchfile | by hand, and then the keyword 'connection' does nothing but | execute that batchifle. Period. No smoke, no mirrors. The batchfile | can do anything you want, use any dialer/ppp driver combination that | you want. It can include various tests and try different phonenumbers | automaticaly. Vladimir is testing it and his 'connect.bat' is now | about 20 lines long and among other things, it tries several DNS's | if the first one fails. | | I like it, it's DOSey not Windowsey. | | | | ============================================================================ | | 172) Fri, Aug 05, 2005 - 3:10:23 pm | | ray> | Anyway it works, Tx Glenn. | | | ============================================================================ | | 173) Fri, Aug 05, 2005 - 3:14:18 pm | | glennmcc> | You broke it. :( | | It works fine on this end. | | --- here's a test of posting XSWAP.TXT --- 3,881 bytes ___ | | | 1) "XSWAP contexts" are my new "invention" for Arachne 1.60. They allow | instant garbage collection, and I feel this may be new memory allocation | strategy, although it has probably been invented a thousand times before. | In case it is really my invention, I am publishing the idea (not the | code) under the GNU public license (www.gnu.org) so no one else will have th | right to patent it ;-). So, just in the case I havn't reinvented the wheel.. | Anyway, I don't want the story of the LZW algorithm in GIF files to be | repeated. | | "Open patent": Memory allocation contexts | ----------------------------------------- | | Traditional memory allocation strategy is to let programmers call | "malloc(...)" style functions, which allocate memory pages without telling | the memory allocation system any details about future usage and deallocation | of the memory. This means that ANY memory swapping strategy can only produce | random results, not the best possible garbage collection algorithm, since it | is invoked by the "free(...)" function, needs extra CPU time to make the mem | re-usable again, and merges blocks of randomly de-allocated memory. | | The XSWAP memory allocation strategy lets programmers specify the context | when allocating memory; at a later time, all memory chunks belonging to a | certain context can be de-allocated at once at very little CPU time cost | (it depends only on number of used memory pages, not on the number of items | being de-allocated). Additionaly, different contexts can be given different | priority when storing pages of memory using media with different access spee | (e.g., RAM vs. disk vs. networked memory). | | I declare this algorithm to be publicly known, so no one is allowed | to protect this algorithm by patent laws of any country worldwide. | If you use this algorithm, you are welcome to mention that it is a | re-implementation of XSWAP memory allocation strategy used in the Arachne | WWW browser. Please contact xchaos@arachne.cz if you are interested. | | - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - | | 2) When running Arachne, the '?' key dumps XSWAP debugging info. | This feature is a kind of Xmas present for Clarence Verge and other | "advanced Arachnologists", who like to compare page load times, etc. ;-) | | The debugging information has the following format: | | *[s=???, n=???, c=???] (repeated many times) r=???, w=??? | | where: | | * (if present) is the number of the area (1 or 2) where buffer is currently | s is the size of buffer in bytes | n is the number of items located in this buffer (tables, lines of text, etc. | r is the number of buffers reads (from XMS, EMS or disk) | w is the number of written buffers (both r and w will overlow after some tim | c is the context; -1 is an undefined value, other values follow: | | #define CONTEXT_HTML 0 | #define CONTEXT_SYSTEM 1 | #define CONTEXT_TABLES 2 | #define CONTEXT_ICONS 3 | #define CONTEXT_TMP 4 | | The XSWAP debugging info can help to analyze some types of problems, e.g., | very slow Arachne operation, etc. You should never actually see "CONTEXT_TMP | in this listing, and the number of items (n) in buffers with CONTEXT_TABLES | should be equal to the number of tables in the document (for 0 tables, this | shouldn't be present at all). Number of HTML atoms Arachne reports | when Alt+M is pressed is NOT identical to the number of items in CONTEXT_HTM | | - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - | | 3) If Arachne crashes with an "Illegal xSwap operation" message, please | send the details (module name and line number) to xchaos@arachne.cz, but | ONLY IF IT IS A 100% REPRODUCIBLE CRASH !!! - i.e., you are able to | install a clean Arachne, configure it, go online with empty an cache if cras | was caused by an online page, and then reproduce the crash exactly at any ti | | | ============================================================================ | | 174) Fri, Aug 05, 2005 - 3:15:01 pm | | glennmcc> | Oops, you beat me to it... you fixed your break. | | | ============================================================================ | | 175) Fri, Aug 05, 2005 - 3:17:57 pm | | glennmcc> | re:"Looks to me like your just uploading stuff in smaller bites." | | Yep, 100% correct in that assumption. | | That's all it took to fix the posting problem was to upload in 16 byte | pieces instead of 512 byte pieces. | | Now go over to mail.yahoo.com and do some web based email. | | All of iot now works perfectly. | | | ============================================================================ | | 176) Fri, Aug 05, 2005 - 3:29:42 pm | | ray> | | So what happens if we try other 'bitesizes'? ... | | // !!glennmcc/RAY Jul 16, 2005 -- fix intermitant posting problem | // by changeing BITESIZE to 16 from 512) | #define BITESIZE 16 | for (; postindex + BITESIZE < ql; postindex += BITESIZE) | { | // this is needed only for WATTCP | while (sock_tbleft(gSocket) < BITESIZE) // SDL | { | SOCK_TICK // Shift TCP/IP | XChLogoTick(1); // Animate logo | if (GUItick()) | goto post_aborted; | } | MALLOCIFNULL(querystring = IEgetSwap(sGlobal.postdataptr)) | SOCK_TICK // Shift TCP/IP | SFW(&querystring[postindex], BITESIZE); | } | if (postindex < ql) | { | while (sock_tbleft(gSocket) < strlen(&querystring[postindex])) // SD | | ============================================================================ | | 177) Fri, Aug 05, 2005 - 3:31:30 pm | | ray> | Sorry, line wraps ... | | | So what happens if we try other bite sizes? ... | | // !!glennmcc/RAY Jul 16, 2005 -- fix intermitant posting problem | // by changeing BITESIZE to 16 from 512) | #define BITESIZE 16 | for (; postindex + BITESIZE < ql; postindex += BITESIZE) | { | // this is needed only for WATTCP | while (sock_tbleft(gSocket) < BITESIZE) // SDL | { | SOCK_TICK // Shift TCP/IP | XChLogoTick(1); // Animate logo | if (GUItick()) | goto post_aborted; | } | MALLOCIFNULL(querystring = IEgetSwap(sGlobal.postdataptr)) | SOCK_TICK // Shift TCP/IP | SFW(&querystring[postindex], BITESIZE); | } | if (postindex < ql) | { | while (sock_tbleft(gSocket) < strlen(&querystring[postindex])) // SD | | | | ============================================================================ | | 178) Fri, Aug 05, 2005 - 3:53:20 pm | | glennmcc> | Ray, | | Yep, your hunch was right on the fastwrite change as-well. | | Here's the entire block of code which is now commented correctly like I | _should_ have done to begin with on Jul 16th | | | //!!glennmcc Jul 16, 2005 -- fix intermitant posting problem | // by sending 16 byte chunks instead of 512 bytes | while(postindex+16 | I tried all 'bitesizes' from 512 to 16 stepping-down in /2 increments. | | 512, 256, 128, 64, 32, 16 | | 16 is where the fix 'took hold' and has not failed me yet. :) | | ============================================================================ | | 180) Mon, Aug 08, 2005 - 5:56:01 pm | | ray> | Glenn, thanks, as you probably figured, I set that #define so's | I could play with some other sizes but if you've already done it I won't | bother, 16 it is. | | BTW, I understand your mortification at having missed commenting that | change since you have always been faultless in that regard. I was | air tight up to base 5 but have since fallen off the straight and narrow | path -- I've got lots of undocumented changes to sort thru and it's a | bit of a mess. That's what happens when a body is working on half a | dozen things at once -- you close a file remembering you have to get | back to it and then forget :( | | | | ============================================================================ | | 181) Mon, Aug 08, 2005 - 10:55:32 pm | | Eric> | Forgetting is agrivating in itself....but I remember that I | forgot....but I forget what it is that I'm remembering | that I forgot. | | ============================================================================ | | 182) Thu, Aug 11, 2005 - 9:01:59 pm | | ray> | Joe, | | Tried to upload base5c just now, got "The document contains no data" | after about 5 minutes of what looked like uploading. The zip file is | <512K and it does contain data, I checked and it's OK. | | | ============================================================================ | | 183) Thu, Aug 11, 2005 - 9:06:05 pm | | ray> | Gregy, | | Wana upload B5c to AVA or to Jake's site or something? Or Glenn could | take it or whatever. I'm not designed to handle administrative | stuff like this, I need some nice old foo ... sargeant ... who can | look after all this sort of stuff. An upload site maintainer I aint. | | | | | ============================================================================ | | 184) Fri, Aug 12, 2005 - 10:46:14 am | | glennmcc> | Re:".... Or Glenn could take it or whatever....." | | If you can dish it out.... I can take it. | | Email it to me if you'd like. | | ============================================================================ | | 185) Fri, Aug 12, 2005 - 11:30:43 am | | gregy> | Ray....you can email it to me, also....if it's source files, though, | why don't you break it into: | base5c01.zip | base5c02.zip | base5c03.zip | | Then I'll upload it to a /pub/ or /coresrc/ directory on AVA... | ... | After all, we've got 150Mbyte of space there, with 1Gbyte traffic | per month, or sumpin... | ... | | | ============================================================================ | | 186) Fri, Aug 12, 2005 - 1:08:19 pm | | ray> | Thanks Gentlemen, I'll send it to both of you. Gregy, why does it have | to be small files, the total size is about 480K? | | | ============================================================================ | | 187) Fri, Aug 12, 2005 - 2:13:14 pm | | glennmcc> | Ray, | | Just now got home and got your eamil with base5c.zip attached. | | Thank you. | | I assume post #182 means you trying to upload to your directory at Yahoo. | | I'll try it right now and see if I have any better results. | | BRB | | | | ============================================================================ | | 188) Fri, Aug 12, 2005 - 2:21:18 pm | | glennmcc> | Nope... no go with Lynx v2.85.1 (DOS) | | ============================================================================ | | 189) Fri, Aug 12, 2005 - 2:22:10 pm | | glennmcc> | I'll try it in Linux | | BRB | | | ============================================================================ | | 190) Fri, Aug 12, 2005 - 2:37:02 pm | | glennmcc> | Still no go with FireFox in Linux :( | | There is definately something buggered-up at Yahoo | | According to this.... | http://help.yahoo.com/help/us/groups/files/files-11.html | | File size limit is not 512kb.... it's 5 megs | | ============================================================================ | | 191) Fri, Aug 12, 2005 - 3:11:30 pm | | ray> | Rodger that with Yahoo, I tried again using Opera in windows with no | luck. | | | ============================================================================ | | 192) Fri, Aug 12, 2005 - 4:39:33 pm | URL | | gregy> | It's up here, 484,754 bytes.......and...complaints, complaints... | ...it doesn't hafta be in small bits...just ordinary bytes... | ...anyway, I just thought your SMTP might go smoother with smaller | emails...my download doesn't matter, since my ISP's POP3 servers don't | seem to mind (at least, I've gotten emails up to 1.5Mbytes)... | | http://arachne.virtualave.net/source/base5c.zip | | | | ============================================================================ | | 193) Fri, Aug 12, 2005 - 4:42:11 pm | URL | | gregy> | And, doggone it..!!! I still can't EVER get my Arachne to FTP upload. | ...I've gotta figure that out, since it's getting to be a pain to | dump OpDOS, reboot to Win3.1, fireup WS_FTP Lite, then, then...etc | ..just to upload a single file... | .. | | | ============================================================================ | | 194) Fri, Aug 12, 2005 - 5:13:48 pm | | ray> | Ah good, the right size, so what's that ~600K~ size you mentioned in | your email? Nevermind. As for FTP, now that I'm starting to get some | Idea how these things work, I'm going to have a look at that one of | these months. I've got Vladimir chewing on B6 and that mad Cossack | doesn't miss a thing ;-) also gotta finish tinkering with this | resart download stuff, then integrate Glenn's 1.85 and Joe's C1. | Should be a couple of months. | | | ============================================================================ | | 195) Fri, Aug 12, 2005 - 5:31:50 pm | | gregy> | Well, if you guys are looking for "projects" to do, you've got one | hanging over "from the old days"....I mean that "runaway download | counter loop" thingee. | ... | Glenn did a partial fix/kludge, but it still runs away, only now it | doesn't keep on to infinity, all the time...only some of the time... | I just had two of them on Yeehaw (it's not Yeehaw, it just DOES it | once in a while, won't stop at Glenn's kludge).... | ... | But, that one is/should-be a fairly high priority.. | | | | ============================================================================ | | 196) Fri, Aug 12, 2005 - 5:35:47 pm | | ray> | Nuts, I thought I might have the ultimate answer to the bad email | thing -- instead of looking for the final dot, look for the final | '+OK' trouble is, there ain't no final '+OK' Why the devil not? | Every other thing recieved from the server end in +OK except for | the text of an email? I don't like it. | | | ============================================================================ | | 197) Fri, Aug 12, 2005 - 5:37:11 pm | | ray> | So what happens if you put: | | text | . | text | | in an email on purpose? | | ============================================================================ | | 198) Fri, Aug 12, 2005 - 6:02:52 pm | | glennmcc> | It gets changed to.... | | text | .. | text | | | | | ============================================================================ | | 199) Fri, Aug 12, 2005 - 6:14:46 pm | | gregy> | Are you rambling on...??? Well, ramble-on over to the ctst. I just | left you a post, Re: your docs.... | ... | You asked for it..... | | | ============================================================================ | | 200) Fri, Aug 12, 2005 - 6:15:27 pm | | gregy> | hi Glenn, ..... | ... | I'm just harrassing Ray... | | | ============================================================================ | | 201) Fri, Aug 12, 2005 - 6:38:48 pm | | ray> | Where does it get changed to ".." ? and how does Arachne know that | it isn't time for 'done = 1' ? | | | | ============================================================================ | | 202) Fri, Aug 12, 2005 - 6:53:19 pm | | ray> | BTW, my resume mail download code seems ok but obviously in needs | some beating up on, Glenn, want a look? | | ============================================================================ | | 203) Fri, Aug 12, 2005 - 6:57:06 pm | | ray> | BTW, stratagy wise, I've got it so that any partialy downloaded emails | are retained in incomplete form even though the full email might then | be redownloaded, this is both easier to do and I'd say much safer, | thoughts? | | | ============================================================================ | | 204) Sat, Aug 13, 2005 - 1:05:14 pm | | gregy> | Re: retains...ata boy......safer for your health.....I mean, | from irrate users' retaliation.... | | | | ============================================================================ | | 205) Sat, Aug 13, 2005 - 4:14:53 pm | | glennmcc> | re:"....... Glenn, want a look? ......" | | Ray, | | Both you and Gregy seem to be in the mood for rethorical questions lately. | | | | Yes, I wanna' see it. ;-) | | ============================================================================ | | 206) Sat, Aug 13, 2005 - 4:36:25 pm | | ray> | Ok, I'll post it to AD. Even if it works as it should, there's still | improvements esp where the 'goto' points. As you will quickly see, | all I've done is add a new return value from TcpIdleFunc() to | add '-2' in case of user break vs. -1 in case of timeout. Then, | at the 'sock_err' label I just point back up the code, reestablish | the socket and send the DELE code for the apropriate messages using | the value of 'process'. I rather suspect that the new DELE code | could be better positioned esp. so that the ppp.log could be kept | open to record the deletions. Kick it around. | | | ============================================================================ | | 207) Sat, Aug 13, 2005 - 4:39:05 pm | | ray> | Oh, I've commented out some code for experimental reasons, that | 'lastchar' stuff looks very strange, it seems not to bee needed at | least for 'good' emails, I'd like to know more about this stuff, | uncomment as needed. | | | | ============================================================================ | | 208) Sat, Aug 13, 2005 - 4:40:38 pm | | ray> | Nuts, I've got a ferry to catch, till Sunday ... | | ============================================================================ | | 209) Sat, Aug 13, 2005 - 7:16:50 pm | | Eric > | Hey Ray, Another rhetorical question, what's the fairy's name, | Tinkerbell ?? | | ============================================================================ | | 210) Mon, Aug 15, 2005 - 4:01:32 pm | | ray> | Ha Ha ... hey that's such an obvious pun yet we never use it up here. | If you want to make puns with 'ferry' it's always vis. a vis. | homosexuals: | | What's the difference between Horseshoe Bay and a Queer with AIDS? | | One is a ferry terminal, the other is a terminal fairy. | | Heh ... heh ... heh ... | | Now some code for McC on AD ... | | | | ============================================================================ | | 211) Mon, Aug 15, 2005 - 4:56:38 pm | | glennmcc> | AD line-wrapped your post making it totally unusable. :( | | If you would please, email that stuff to me. | | ============================================================================ | | 212) Mon, Aug 15, 2005 - 5:11:50 pm | | ray> | Wha? AD hasn't done that before has it? :( | Just as well, since I"ve already found a small bug, I used a macro | for the write of the message text to log, which is incorrect. | And ... oh joy, oh bliss ... I've got a spam! :-) and I was swift enough | to change my settings to leave it on my server, so I can now | play with it to my heart's delight. Steenkin' Outlook, they all | come from Outlook no? Anyway, I'll zip 'em off to you shortly, just | let me play with this spam for a while. | | | | ============================================================================ | | 213) Mon, Aug 15, 2005 - 6:27:14 pm | | ray> | Glenn, | | It's in the wire, modified just a few minutes ago, so handle with care. | | | ============================================================================ | | 214) Tue, Aug 16, 2005 - 8:49:48 am | | Joe> | Hi all, | | Just a very quick drop-in ... havenīt been near the īnet for over a | week now, I think. | | Just saw a comment above about line-wrapping at AD ... If you use | the web interface for posting, thereīs a check-box to enable/disable | wrapping. If you use e-mail, not sure how this is controlled, it | _might_ be an account option, otherwise, Iīve no idea. | | However, your best bet is to either upload code to the Files area or, | if itīs just a short code snippet, post as a text attachment. AFAIK, | such attachments are sent to all list members, but are not archived. | So, for the best of both worlds (?), you can e-mail code snippets as | a plain message, with the snippet also included as a text attachment. | That way, list members can receive the snippet in a form they can | use directly, without any line-wrapping issues, yet there remains the | plain message version, archived for posterity, which members can also | reply to, with suitable context provided by the original message text. | | Have fun! | | | ============================================================================ | | 215) Tue, Aug 16, 2005 - 1:26:44 pm | | ray> | Hi Joe :) | | | ============================================================================ | | 216) Tue, Aug 16, 2005 - 7:51:31 pm | | glennmcc> | Ray, | | Just now looked through your latest popget.c | | Your code is so totally diferent from mine that it is imposible to | import your changes into my popget.c | | My recomendation is that you keep testing a while longer and when you | are satisfied that it all works correctly..... | | Make your own distro of Arachne and release it to the public. | | | ============================================================================ | | 217) Tue, Aug 16, 2005 - 9:04:35 pm | | ray> | Yeah it's a streach to import back to base 0. But this is the first | I've heard about making my own distro, I thought you were planning | to convert to cleaned code after 1.85 wich would obviously remove | any code translation problems. Have you looked at 5c? Like? Don't | like? Let me know what you're thinking and we can make a plan from | there but I'm no where near ready or even willing to make a seperate | distro. Almost everything I've learned is in core, beyond that I'm | still a bit wet behind the ears. Or, if you want, I could reverse | engineer my popget.c stuff into base 0, I've had lot's of practice | with that, no problemo. | | | ============================================================================ | | 218) Tue, Aug 16, 2005 - 9:07:16 pm | | ray> | Another possibility is that I send you B6 as is and you can work | with that vis. this popget stuff. Mind you don't have to commit to | it, just use it as a testbed. | | | ============================================================================ | | 219) Wed, Aug 17, 2005 - 1:15:42 pm | | glennmcc> | Have not looked at 5c 'in-depth' as-of-yet. | | Does it include all of my changes which became my v1.85;GPL ? | | If so.... can I just drop-in your current popget.c and compile it | to the create the actual release version of 1.87;GPL ? | | If not.... then might I please impose upon you to make the needed | changes to get it ready for said release ? | | IMO, | Working together we can have v1.87;GPL ready to go by Sep 1st | | | ============================================================================ | | 220) Wed, Aug 17, 2005 - 3:23:13 pm | | ray> | 5c is synched with 1.81, I wanted to wait for a full debugging of that | and 6 before starting the synch with 1.85 -- first things first. | I can start this any time but first it would be good to sort out | what you think of 5c vs. 6, since if you like 6, there's no need | to bring 5c up to speed, just leave her as she is and concentrate on | 6. At present I'd say my popget experiments a too green to include | in any stable release but that could change by Sept. | | ============================================================================ | | 221) Wed, Aug 17, 2005 - 3:30:08 pm | | ray> | Also, the current popget won't work with 5c since (by definition) | modules shouldn't be mixed across bases without expecting some | probably minor compatablity glitches. I should get the base6 docs | done, that way you can see what I've been up to there. 6 is basicaly | just a final global edit, saved about 1400 lines mostly by using | some global macros for very common expresions. | | | ============================================================================ | | 222) Wed, Aug 17, 2005 - 3:35:27 pm | | ray> | The way I figure it, for you the jump from 0 to 6 would hardly be | more difficult than the jump from 0 to 5c, so you may as well go | for the gusto, once you find your way around the newer layout, I'm | sure you'll find that up close, 80% of the functions work just as | before even if they are cosmeticaly different. | | | | ============================================================================ | | 223) Wed, Aug 17, 2005 - 3:38:15 pm | | ray> | Oh, if you want to have a look at 6 that would be nice, since I'd | feel more confident in it if you had given it a good thrashing before | I start incorporating new code. | | ============================================================================ | | 224) Wed, Aug 17, 2005 - 4:31:37 pm | | ray> | Where's Gregy? ... | | ============================================================================ | | 225) Wed, Aug 17, 2005 - 5:21:44 pm | | glennmcc> | So if I understand correctly, that stuff you sent me for popget.c is for 'Ba | | If so..... base6, here I come. | | OK my friend ........ send the whole shibang on over this way. | | | ============================================================================ | | 226) Wed, Aug 17, 2005 - 5:26:49 pm | | ray> | OK, let me hack around with popget some more and I'll send it after | dinner, docs to follow in days or weeks (a lot of work to do :() | Today I managed to bust something. | | | ============================================================================ | | 227) Wed, Aug 17, 2005 - 7:11:13 pm | | ray> | It's in the pipe. | | | ============================================================================ | | 228) Wed, Aug 17, 2005 - 9:35:05 pm | | glennmcc> | Two points ......... | | 1) All ... and I do mean _ALL_ of the keywords that existed before | _MUST_ be put back the way they were. | | 2) All ... and I do mean _ALL_ of the hotkeys that existed before | _MUST_ be put back the way they were. | | | Both points NON-NEGOTIABLE | | | ============================================================================ | | 229) Wed, Aug 17, 2005 - 10:54:36 pm | | ray> | Well that seems heavy, does it include the changes you yourself suggested? | Why so intransigent? Compromise is the soul of cooperation naturaly, | so why such a ridgid position? As far as keywords goes, I'd need | a very good reason to undo SysFont and the color keywords could only | be reversed with very great difficulty so what's the reason? | | As far as hotkeys, I added several and changed a very few, but all | of that is easy to change, infact I'd like to make an entirely new | alternate set of hotkeys, thus the old set would obviously be | best left exactly stock. | | | ============================================================================ | | 230) Wed, Aug 17, 2005 - 11:00:39 pm | | ray> | BTW, not to belabour the obvious, but you are of course free to | do anything with my code you like, and even if I don't agree, I'll | help you do it, but in so far as we try for code acceptable to both, | then we either compromise or we conditionalise differences. | | | ============================================================================ | | 231) Thu, Aug 18, 2005 - 5:42:09 am | | Joe> | Hi again, | | Well, Iīm at my sisterīs place for a couple of days at present. | | Iīve made a very interesting discovery - her conmputer and/or | ISP exhibit that "2/3 wattcp speed" effect, so maybe Iīll get | to track down where that effect comes from after all! :-) | | | ============================================================================ | | 232) Thu, Aug 18, 2005 - 5:47:49 am | | Joe> | Iīve just sent Yahoo a query re the 5M vs. 0.5M file size thing. | Maybe theyīll get back to me, weīll see. | | My guess is that the limit was once 5M, latter changed to 0.5M | and someone forgot to update the help info. I know they used to | archive message attachments but changed that policy to reduce | server space requirements for Yahoo Groups, perhaps at that time | the file size limit was also reduced. The savings made may have | contributed to their being able to offer 1G space for their | web-mail accounts. | | ============================================================================ | | 233) Thu, Aug 18, 2005 - 5:53:58 am | | Joe> | Re "clean code" ... Note that 1.83C1 is synched to 1.83, not | just 1.81 as the "5c" code as mentioned by Ray. It still has | all the POSIX and Linux stuff, so Glenn, why not consider this | as a starting point if you plan to move to the "clean code"? | Iīm sure Ray would have little trouble moving his "6" code | changes into this code base. Just a suggestion. | | L8R | | ============================================================================ | | 234) Thu, Aug 18, 2005 - 7:34:41 am | | ray> | Joe, | | Yeah, the 3c line has its merrits, esp. as you say, the POSIX stuff | is all there, the big drawback is that 6 uses so much less memory than | 3c. | | | | ============================================================================ | | 235) Thu, Aug 18, 2005 - 1:40:31 pm | | joe> | Hey, if your code is synched up to 1.83 that must mean that you have | imported most of Glenn's stuff yourself, that's excelent news, because | it means that I can use your sources for both his and your work up to | 1.83 right? Infact, if you'd import his 1.85 stuff to, then that | would make my next job much easier, any plans to do so? Or, if I did it, | subject to your final approval, that could be '1.85C1' -- very nice | to have the 3x line brought right up to date! | | | ============================================================================ | | 236) Thu, Aug 18, 2005 - 1:42:00 pm | | ray> | Ooopps, that was Ray posted that, not Joe. | | | ============================================================================ | | 237) Thu, Aug 18, 2005 - 5:10:48 pm | | glennmcc> | Joe, | | re:" the 5M vs. 0.5M file size thing." | | It's not even working for a file of only 0.4M | | Both Ray and I have tried several times to upload this file to no avail. | | | Volume in drive E is MCCORKLE_03 | Directory of E:\ARACHNE\TEAM\RAY_A | | BASE5C ZIP 484754 8-12-05 3:08p | 1 File(s) 8863744 bytes free | | | ============================================================================ | | 238) Thu, Aug 18, 2005 - 5:29:43 pm | | glennmcc> | Ray, | | I am willing to compromise on many, many things. | | However, I will not compromise on those 2 points. | | All keywords and all hot keys which exist in v1.85 must be put back | in-order for me to switch over to your B6 | | My mind is made up and it is not open for discussion nor debate. | | ============================================================================ | | 239) Thu, Aug 18, 2005 - 6:22:32 pm | | glennmcc> | BTW, | | I still can't figure-out where you came up with the notion that 'NoHTT' | ever was a keyword in any previous version. (mine or anyone else's) | | --- from your arachne.cfg in the ZIP you sent to me --- | | [GLENN ONLY] | NoHTT Y | | [BASE6 ONLY] | ; KeepHTT N ; Replaces Glenn's "NoHTT" -- same function but you | ; don't answer yes to a negative. | ____________ | | Searched my HDD using this command line .... find "nohtt" e:\arachne.cfg /s | | ---------- e:\arachne\team\ray_a\ra050817\config\arachne.cfg | NoHTT Y | ; KeepHTT N ; Replaces Glenn's "NoHTT" -- same function but you | | | There never has been such a keyword as 'NoHTT' in all the history of Arachne | | It is and always has been 'KeepHTT' | | ---- from keywords.txt ---- | "KeepHTT" Yes|No (Controls caching of HTTP headers) | __________________________ | | | --- from arachne.cfg in all of these versions --- | --- using this command line..... find "keephtt" arachne.cfg /s | | ---------- h:\!\!ar161sr\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar162\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar164\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar165\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar166\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar167\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar168\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar169\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar170\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar171\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar173g1\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar175g1\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar177\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar179\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar181\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar183\arachne\arachne.cfg | KeepHTT Yes | | ---------- h:\!\!ar185\arachne\arachne.cfg | KeepHTT Yes | | _________________________________ | | --- now serching for all files _not_ containing 'NoHTT' using this | command line ..... find /v /f "nohtt" arachne.cfg /s | | h:\!\!ar161sr\arachne\arachne.cfg | h:\!\!ar162\arachne\arachne.cfg | h:\!\!ar164\arachne\arachne.cfg | h:\!\!ar165\arachne\arachne.cfg | h:\!\!ar166\arachne\arachne.cfg | h:\!\!ar167\arachne\arachne.cfg | h:\!\!ar168\arachne\arachne.cfg | h:\!\!ar169\arachne\arachne.cfg | h:\!\!ar170\arachne\arachne.cfg | h:\!\!ar171\arachne\arachne.cfg | h:\!\!ar173g1\arachne\arachne.cfg | h:\!\!ar175g1\arachne\arachne.cfg | h:\!\!ar177\arachne\arachne.cfg | h:\!\!ar179\arachne\arachne.cfg | h:\!\!ar181\arachne\arachne.cfg | h:\!\!ar183\arachne\arachne.cfg | h:\!\!ar185\arachne\arachne.cfg | | _____________________________________ | | All of the same files which _do_ contain 'KeepHTT' _do not_ contain 'NoHTT' | | ============================================================================ | | 240) Thu, Aug 18, 2005 - 7:33:54 pm | | ray> | Glenn, | | Re: nohtt, I'll look into that, maybe I got scrambled somewhere, anyway | if it always has been keephtt, then that's one less thing to fight | about < g g g >. | | As for the keywords etc. it you are not even willing to discuss it, | then I guess we can't discuss it, too bad because there are several | options which might be explored. | | | | | | ============================================================================ | | 241) Thu, Aug 18, 2005 - 8:00:22 pm | | ray> | | Right you are: | | I got the keyword and the structure member scrambled. In B5 I was | matching the names of keywords exactly to their coresponding member for | obvious housekeeping reasons and somehow I got it into my head that | the *keyword* not the *member* was named 'nohtt'. | | BASE 0: CONFIG.C: | | value=configvariable(&ARACHNEcfg,"KeepHTT",NULL); | if(value && toupper(*value)=='N') | user_interface.nohtt=1; | else | user_interface.nohtt=0; | | Thanks. | | | ============================================================================ | | 242) Thu, Aug 18, 2005 - 8:01:21 pm | | ray> | Where's Gregy? | | | ============================================================================ | | 243) Thu, Aug 18, 2005 - 8:21:23 pm | | glennmcc> | re:"one less thing to fight about" | | Sheesh....... you din't take the bait. | | I was _trying_ to start a fight. | | Don't you remember ? | | You and I start working together both 'well and hard' after fighting | about something. | | Remember how well we did in fixing that damned bug with CSS ? | | Ahhh forget it.... if you ain't gonna fight.... I might as well just | call it a night. | | CU tomorrow ;-) | | | ============================================================================ | | 244) Thu, Aug 18, 2005 - 8:21:55 pm | | glennmcc> | Gregy seems to have gone AWOL again | | ============================================================================ | | 245) Thu, Aug 18, 2005 - 8:57:47 pm | | ray> | Ah, but there will be serious things to fight about anyway, you know it. | Yeah, I do remember that CSS stuff, 'twas fun and efective, I'd love | to get your input on what I'm doing in popget. | G'nite. | | | ============================================================================ | | 246) Fri, Aug 19, 2005 - 6:43:40 am | | gregy> | no...not AWOL....conscripted for other off-site duties for the past | week... | ... | now, I need to go thru 2.5+Mbyte in my Inbox.... | | | ============================================================================ | | 247) Sat, Aug 20, 2005 - 3:18:57 am | | Joe> | Hi all, | | Roger that about 0.5M uploads not working at AD. That particular | value was just a guess on my part. Clearly there is a limit and | itīs clearly not 5M as claimed in the help, but without doing | some trial-and-error tests, the exact limit is unknown at this | stage. You were able to upload all the sources in two parts, so | that gives some idea of what is small enough to upload. If Yahoo | get back to me with further details, Iīll inform the group ... | | ============================================================================ | | 248) Sat, Aug 20, 2005 - 3:21:11 am | | Joe> | Ray, | | Yep, 1.83C1 is synched to Glennīs 1.83, and yes, obviously I did | the synching. | | ============================================================================ | | 249) Sat, Aug 20, 2005 - 8:56:04 am | | ray> | Joe, | | Glad to hear it, now what about a synch to 1.85? | It would save me a lot of time if you did it, but like I said, | if you want, I'll do it and subject to your final edit, that could | continue your code line. | | | | ============================================================================ | | 250) Sat, Aug 20, 2005 - 1:08:52 pm | | ray> | Had a good idea: the pop3.log records all emails from one Arachne | session instead of being overwritten each time you restart a download. | This is especialy nice with my ability to abort a download without | starting again next time. | | ============================================================================ | | 251) Sun, Aug 21, 2005 - 11:18:35 am | | ray> | Glenn, | | In mime.cfg you have lots of stuff with the "|" character being | used to separate commands. This is pretty abominable, I'd say | you should be using '\n' which is more commonly used anyway. | Not that you'd care, but 4DOS chokes on the former, flagging it as | a bad command, (which it is). | | | ============================================================================ | | 252) Sun, Aug 21, 2005 - 4:22:10 pm | | glennmcc> | Changes made. | | file/bmp2zbm.dgi >TXT|zip -m -j $s _4prt.bmp \n del _4prt.bmp \n if exist | file/export.dgi >TXT|copy /b _4prt.txt $s>NUL \n del _4prt.txt \n if exis | file/exportps.dgi >TXT|copy /b _4prt.ps $s>NUL \n del _4prt.ps \n if exist | file/exportbmp.dgi >TXT|copy /b _4prt.bmp $s>NUL \n del _4prt.bmp \n if exis | file/bmp2jpg.dgi >TXT|cjpeg -q 95 _4prt.bmp \n copy _4prt.jpg $s \n del _4 | converted to $s>$2 | file/exportbmp.dgi |[100]COPY /b _4prt.bmp $w$s>NUL \n del _4prt.bmp | file/edithelp.dgi >htm|copy textarea.tmp mail.tmp \n copy edithelp.htm $2 | file/.skn >HTM|copy $1 skin.css \n copy /v chngdone.htm $2 | | | ============================================================================ | | 253) Sun, Aug 21, 2005 - 4:32:21 pm | | ray> | Hey thanks! Never thought I'd see you change something just because | 4DOS didn't like it :-)) ... just when you think you've got someone | figured out < g g g > | | | | ============================================================================ | | 254) Sun, Aug 21, 2005 - 5:39:27 pm | | glennmcc> | Oh.... you are quite welcome. :) | | And if there's one thing that you should have figured out about me by now... | | You'll _never_ figure me out. | | | | ============================================================================ | | 255) Sun, Aug 21, 2005 - 5:41:25 pm | URL | | glennmcc> | BTW, | | 1460 failed miserably. | | 576 is where she stays. | | | ============================================================================ | | 256) Sun, Aug 21, 2005 - 7:08:49 pm | | glennmcc> | Ray, since you are having some fun with popget.c | | I gots an idea.......... | | ESC == abort current download, delete all previouisly D/Led | messages which were 'marked' for deletion _except_ | the current one which was aborted. | | Ctrl+ESC == abort current download, delete all previouisly D/Led | messages which were 'marked' for deletion _including_ | the current one which was aborted. | | Why ? | | For the past 10 days I have been getting between 30 and 40 copies per day | of this new variant of the 'Surila' backdoor trojan. | | It started-out that all of them were 'Surila.M' | Now most of them are 'Surila.N' wioth just a few of the Ms mixed-in. | | All of the trojan infected messages are about 112kb in size. | | If I see a message coming-in that's 112kb........ I could hit Ctrl+ESC | to remove it from the server without waiting for the D/L complete to | send the DELE command. | | If it turns-out through examining the partial D/L that this particular | one was legit.... I would then simply request that the sender re-send it | to one of my other email accounts that does not get those trojans. | | My CisNet address is so 'visible' that its the only one that gets all of | this crap. | | My Yahoo address and the one at MyRealBox.com never get any that stuff. | | Therfore I would never need to use Ctrl+ESC at MyRealBox | | And since the free address at Yahoo can only be checkked via the webmail | interface..... both ESC and Ctrl+ESC _can't_ be used on that one. | | ============================================================================ | | 257) Sun, Aug 21, 2005 - 7:50:20 pm | | ray> | Uh-oh, we're thinking alike again ;-) | | ESC already works as you suggest, and I was about to throw in a | delete key that would work exactly as you describe ... but ... | the best answer would be to coble up an HTML page that lets you | make one of several choices, see that popget I sent you, I can think | of at least 4 things you might want to do. | | | ============================================================================ | | 258) Sun, Aug 21, 2005 - 7:55:19 pm | | ray> | Nuts, I'm all out of spams again so I can't test this stuff :( | | Oh, you should check out my on-the-go font shift code, you can | make it work in your sources easily. | | | | ============================================================================ | | 259) Sun, Aug 21, 2005 - 8:01:17 pm | | glennmcc> | Lemme see dat | | ============================================================================ | | 260) Sun, Aug 21, 2005 - 8:13:57 pm | | glennmcc> | On that note though..... | | IMO, | There is never any reason what-so-ever to need a fontshift of more than +1 | | +2 is way too big | | and that setting of +4 you mentioned is just totally ridiculous. | | ============================================================================ | | 261) Sun, Aug 21, 2005 - 8:19:11 pm | | glennmcc> | Even at +1 the time-of-day seconds are now off-the-edge and the last | letter of my last name now encroaches over into the online-timer. | | http://www.cisnet.com/glennmcc/my-stuff/shift_1.gif | | +2 gets even worse.... | http://www.cisnet.com/glennmcc/my-stuff/shift_2.gif | | | | ============================================================================ | | 262) Sun, Aug 21, 2005 - 8:30:04 pm | | glennmcc> | I'd hate to think just how badly things will get mangled-up at +3 and +4 | | ============================================================================ | | 263) Sun, Aug 21, 2005 - 8:37:11 pm | | glennmcc> | Now we can see why Michael had opt_misc.ah setup to only allow +1 | as the max setting via the setup screens. | | IMNSHO, | | Since even +1 begins to cause 'encroachment' of one thing over into another. | +1 should be eliminated as a posibilty and the allowed settings should be | hard-coded as follows.... | | if >0 == 0 | if <-2 == -2 | | Just for the hell of it, I'm going to try -3 and -4 | | BRB | | ============================================================================ | | 264) Sun, Aug 21, 2005 - 8:47:02 pm | | glennmcc> | -2 is already the limit. | | No change with settings of -3 -4 -5 -6 | | So it looks like my 2nd 'if' up-above is already in there. | | Now all we need to do is add that 1st one to keep it from ever going above 0 | | | | ============================================================================ | | 265) Sun, Aug 21, 2005 - 8:50:09 pm | | ray> | I would have said the same things ... infact I did say them ... but now | that I have this as an instant response toggle, I can ramp up to +2 | in those few places where I might want it then just ramp down again | as needed. I never thought to try the negatives, is there any point? | BTW, the 'propnn6.fnt' error: it seems to me that this font just | doesn't work so I've limited my maximum font to '4', can you | confirm? | | | ============================================================================ | | 266) Sun, Aug 21, 2005 - 8:55:08 pm | | ray> | Here's the guts of it, how it would look in B0 I can't remember but | the functionality is simple. | | GUIEVENT.C: | | // RAY: 050821: Toggle sUI.fontshift up two or back to normal. | case CTRLF: // 6 | { | sUI.fontshift++; | if (sUI.fontshift > 2) | sUI.fontshift = 0; | return Repaint(); // NB not RedrawAll()! | } | | ============================================================================ | | 267) Sun, Aug 21, 2005 - 9:09:55 pm | | glennmcc> | Confirmed... but even worse here. :( | | Anything higher that +2 crashes my system when trying to load propnn6.fnt | | | ============================================================================ | | 268) Sun, Aug 21, 2005 - 9:12:14 pm | | glennmcc> | Where do you 'set' the value for 'case CTRLF:' ? | | | | ============================================================================ | | 269) Sun, Aug 21, 2005 - 9:14:37 pm | | glennmcc> | It was easy-as-pie to set the limts between 0 and -2 | | --- in config.c --- | | if(cgamode) | user_interface.fontshift=-2; | else | { | value=configvariable(&ARACHNEcfg,"FontShift",NULL); | if(value) | //!!glennmcc: Aug 21, 2005 -- limit to these 3 values.. 0, -1 and -2 | { | user_interface.fontshift=atoi(value); | if(user_interface.fontshift <-2) user_interface.fontshift=-2; | if(user_interface.fontshift >0) user_interface.fontshift=0; | } | // user_interface.fontshift=atoi(value); // original single line | //!!glennmcc: end | else | user_interface.fontshift=0; | } | | | ============================================================================ | | 270) Sun, Aug 21, 2005 - 9:17:03 pm | | glennmcc> | To change yours so that she goes from -2 through 0 and starts over at -2 | | sUI.fontshift++; | if (sUI.fontshift > 0) | sUI.fontshift = -2; | | | | ============================================================================ | | 271) Sun, Aug 21, 2005 - 9:31:23 pm | | glennmcc> | My version of your idea ;-) | | -- in guievent.c --- | | //!!glennmcc: begin Aug 21, 2005 | //idea stolen from Ray ;-) | //cycle through fontshift settings --- Ctrl+F | else if(key==8454) | { | user_interface.fontshift++; | if (user_interface.fontshift >0) | user_interface.fontshift=-2; | } | return repaint(); | //!!glennmcc: end | | | ============================================================================ | | 272) Sun, Aug 21, 2005 - 9:36:00 pm | | glennmcc> | oops.... return needs to be inside the brackets. | | return repaint(); | } | //!!glennmcc: end | | | | ============================================================================ | | 273) Sun, Aug 21, 2005 - 9:45:54 pm | | glennmcc> | 3 more copies of Surila.N just arrived :(( | | New idea for popget.c | | Alt+ESC == abort D/L of current message, mark it for deletion even if | 'keepOnServer Yes', continue-on and begin D/L of next message. | | | ============================================================================ | | 274) Sun, Aug 21, 2005 - 10:05:29 pm | | ray> | Yeah that's the spirit, one can obviously choose the range of values | one wants, but do we really need negative shifts? | | | ============================================================================ | | 275) Sun, Aug 21, 2005 - 10:10:49 pm | | ray> | And there's nothing wrong in principal with fontshifts bigger than 2 | so long as the final font size calculated in svgastat is < 6. See where | the range checking is? Just set the limit to '4' any your OK. The +2 | shift could be nice for some people for reading mail, it looks quite | good if you ask me. It looks to me like size 6 fonts just don't work, | but they must have at one time or they wouldn't be there no? Can you | test some old cores and see if they ever did work? | | | ============================================================================ | | 276) Sun, Aug 21, 2005 - 10:11:54 pm | | glennmcc> | It's the pos shift we don't need. | | The neg shifts don't cause any problems. | | Even a shift of only +1 causes the problem of the final digit in the | time-of-day clock being off-the-edge. | | http://www.cisnet.com/my-stuff/shift_1.gif | | +2 makes it even worse by causing the #of atoms / page load time | over in the Alt+M box to be shifted down on top of the online timer. | | http://www.cisnet.com/my-stuff/shift_2.gif | | | | ============================================================================ | | 277) Sun, Aug 21, 2005 - 10:12:08 pm | | ray> | Maybe some memory saving stuff but the boots to the size 6 fonts. | | | ============================================================================ | | 278) Sun, Aug 21, 2005 - 10:14:00 pm | | ray> | But remember, with this new code you can toggle it on and off instantly | so if you need it to read some fine text in some html page, you've | got it, then just turn it off. Sure the neg shifts cause no trouble, but | what use are they? Show me any situation where it'd be useful. | | | ============================================================================ | | 279) Sun, Aug 21, 2005 - 10:20:31 pm | | glennmcc> | To get more lines on-screen at once. | | BRB | | | ============================================================================ | | 280) Sun, Aug 21, 2005 - 10:31:19 pm | | glennmcc> | Reset00 site at bottom.... | http://www.cisnet.com/glennmcc/my-stuff/shift_0.gif | | 10 additional links past Reset00... | http://www.cisnet.com/glennmcc/my-stuff/shift_-2.gif | | ============================================================================ | | 281) Mon, Aug 22, 2005 - 12:57:46 pm | | ray> | Hmmmm, OK, I see the point, I'll adjust my range to -2 >> +2 | | | ============================================================================ | | 282) Mon, Aug 22, 2005 - 3:16:14 pm | | ray> | I can't get propnn6.fnt to work nohow but the '3D' size six fonts | are fine. BTW where are these 3D fonts? they don't seem to be | stored with the others. I just added a hack such that if the font | won't load it tries again with the next size down, seems to work but | I don't like it. | | | ============================================================================ | | 283) Mon, Aug 22, 2005 - 7:55:55 pm | | ray> | Right, I know where the 3D fonts come from :) | | | ============================================================================ | | 284) Mon, Aug 22, 2005 - 9:52:05 pm | URL | | glennmcc> | Got it. :) | | >1 still causes the crash on propnn6.fnt | (we'll work that one out eventually) | | Look for these changes in all 6 files contained in this ZIP | http://www.cisnet.com/glennmcc/my-stuff/aug22.zip | | ---------- h:config.c | [688]//!!glennmcc: Aug 22, 2005 -- legal values... -2 through +1 | | ---------- h:drawtime.c | [303]//!!glennmcc: Aug 22, 2005 | | ---------- h:guidraw.c | [454]//!!glennmcc: Aug 22, 2005 | [692]//!!glennmcc: Aug 22, 2005 -- maintain size independant of fontshift | | ---------- h:guievent.c | [868]//!!glennmcc: begin Aug 22, 2002 | [1313]//!!glennmcc: begin Aug 22, 2005 | | ---------- h:guitick.c | [209]//!!glennmcc: Aug 22, 2005 -- prevent fontshift >1 from messing-up memi | | ---------- h:outs.c | [26]//!!glennmcc: Aug 22, 2005 -- maintain size independant of fontshift | | | ============================================================================ | | 285) Mon, Aug 22, 2005 - 10:09:39 pm | | ray> | I'll have a look tomorow, but from the above you might be on to | something -- we need a way of insolating some stuff from the fontshift | while allowing it to have effect elsewhere. But I won't rest until | we know what the hell is wrong with propnn6.fnt. Glenn you should be | able to trace back to the stone age and see when that font stopped | working. | | | | ============================================================================ | | 286) Mon, Aug 22, 2005 - 11:01:35 pm | | glennmcc> | I never stopped working. | | It still works fine today. | | We just can't fontshift it more than +1 | | So long as fontshift is limited to amx of +1........ everything is fine. | | Including propnn6.fnt | | ============================================================================ | | 287) Tue, Aug 23, 2005 - 7:17:14 am | | ray> | Hmmm, it seems to me that the propnn6.fnt problem only happens because | in some page a proportional size 4 'nn' font is called for and it | then gets shifted up propnn6 if fontshift is 6, but only up to propnn5 | if fontshift is 5. I'd be surprised if propnn6.fnt works anywhere | anytime. If it does then I'm barking up the wrong tree. | | works anywhere | | ============================================================================ | | 288) Tue, Aug 23, 2005 - 7:18:44 am | | ray> | BTW I can fontshift up to 4 (not that it has any real use ;-) as long | as I add a guard in HTMLfont() against propnn6, so it's that font file | that's the culprit. | | | | ============================================================================ | | 289) Tue, Aug 23, 2005 - 7:19:56 am | | ray> | Oppps, above I ment 'but only up to propnn5 if fontshift is 1. | | | ============================================================================ | | 290) Tue, Aug 23, 2005 - 8:23:58 pm | | ray> | I've got a better idea :-) | | ============================================================================ | | 291) Tue, Aug 23, 2005 - 9:10:36 pm | | ray> | I just ran a batchfile test of every one of the 48 html fonts and | they all load except the four 'propxx6.fnt' the respective fixedxx6.fnt | are fine. These fonts appear to be busted. | | | | ============================================================================ | | 292) Tue, Aug 23, 2005 - 9:24:34 pm | | ray> | And none of 'em work with 1.71 neither, but that's as far back as I | go. Glenn be a sport and dig back further and see if they work with | any core you've got. Nuts they must have worked at some time. Unless | at some point these fonts became corrupt, so it's not a fault with | our cores, but with the fonts themselves. | | | | ============================================================================ | | 293) Wed, Aug 24, 2005 - 2:12:59 pm | | glennmcc> | Send me that batchfile, would you please ? | | That way we'll both be running the same test for sure. :) | | | | ============================================================================ | | 294) Wed, Aug 24, 2005 - 9:04:31 pm | | ray> | for %a in (system\*.fnt) (copy %a system\8x14.fnt ^ pause ^ call arachne) | | 4DOS allows multple commands per line and bracketing as you see, but | you get the idea, I just use each font in turn as the system font | and see if HTMLfont() swallows it or not: | | | | | ============================================================================ | | 295) Wed, Aug 24, 2005 - 9:08:23 pm | | ray> | Here's my fix for the faulty font: | | SVGASTAT.C | | s[6] += size; // replace '0' with 'size'. | } | CurrentSize = size; | CurrentStyle = style; // Remember current size/style | | if (x_fnt_load(string, 25, 1) != 1) // load font from disk | { | // RAY: 050821: four 'propxx6.fnt's seem not to work with my code nor with 1 | // nor with 1.71, but the four 'fixedxx6.fnt's' are fine, why? Perhaps becau | // The problem fonts are > 16K in size? This hack prevents the problem. | --size; | if (size > 0) | goto trysmallerfont; | FontErr(string); | } | } // block | } // End: void H TMLfont(int fnum, char style) | | | ============================================================================ | | 296) Wed, Aug 24, 2005 - 10:06:58 pm | | glennmcc> | I've just completed comparing the fonts. | | All fonts except 'size 1' have not changed since version 1.41 | | Comparing files in... | L:\ARACHNE\SYSTEM\141 | with files in... | L:\ARACHNE\SYSTEM\187 | | Comparing 8X14.FNT with 8X14.FNT - Files compare ok. | Comparing FIXEDBI1.FNT with FIXEDBI1.FNT | Files are not the same size. Compare them anyway (Y/N) ? N | Comparing FIXEDBI2.FNT with FIXEDBI2.FNT - Files compare ok. | Comparing FIXEDBI3.FNT with FIXEDBI3.FNT - Files compare ok. | Comparing FIXEDBI4.FNT with FIXEDBI4.FNT - Files compare ok. | Comparing FIXEDBI5.FNT with FIXEDBI5.FNT - Files compare ok. | Comparing FIXEDBI6.FNT with FIXEDBI6.FNT - Files compare ok. | Comparing FIXEDBN1.FNT with FIXEDBN1.FNT | Files are not the same size. Compare them anyway (Y/N) ? N | Comparing FIXEDBN2.FNT with FIXEDBN2.FNT - Files compare ok. | Comparing FIXEDBN3.FNT with FIXEDBN3.FNT - Files compare ok. | Comparing FIXEDBN4.FNT with FIXEDBN4.FNT - Files compare ok. | Comparing FIXEDBN5.FNT with FIXEDBN5.FNT - Files compare ok. | Comparing FIXEDBN6.FNT with FIXEDBN6.FNT - Files compare ok. | Comparing FIXEDNI1.FNT with FIXEDNI1.FNT - Files compare ok. | Comparing FIXEDNI2.FNT with FIXEDNI2.FNT - Files compare ok. | Comparing FIXEDNI3.FNT with FIXEDNI3.FNT - Files compare ok. | Comparing FIXEDNI4.FNT with FIXEDNI4.FNT - Files compare ok. | Comparing FIXEDNI5.FNT with FIXEDNI5.FNT - Files compare ok. | Comparing FIXEDNI6.FNT with FIXEDNI6.FNT - Files compare ok. | Comparing FIXEDNN1.FNT with FIXEDNN1.FNT | Files are not the same size. Compare them anyway (Y/N) ? N | Comparing FIXEDNN2.FNT with FIXEDNN2.FNT - Files compare ok. | Comparing FIXEDNN3.FNT with FIXEDNN3.FNT - Files compare ok. | Comparing FIXEDNN4.FNT with FIXEDNN4.FNT - Files compare ok. | Comparing FIXEDNN5.FNT with FIXEDNN5.FNT - Files compare ok. | Comparing FIXEDNN6.FNT with FIXEDNN6.FNT - Files compare ok. | Comparing PROPBI1.FNT with PROPBI1.FNT | Files are not the same size. Compare them anyway (Y/N) ? N | Comparing PROPBI2.FNT with PROPBI2.FNT - Files compare ok. | Comparing PROPBI3.FNT with PROPBI3.FNT - Files compare ok. | Comparing PROPBI4.FNT with PROPBI4.FNT - Files compare ok. | Comparing PROPBI5.FNT with PROPBI5.FNT - Files compare ok. | Comparing PROPBI6.FNT with PROPBI6.FNT - Files compare ok. | Comparing PROPBN1.FNT with PROPBN1.FNT | Files are not the same size. Compare them anyway (Y/N) ? N | Comparing PROPBN2.FNT with PROPBN2.FNT - Files compare ok. | Comparing PROPBN3.FNT with PROPBN3.FNT - Files compare ok. | Comparing PROPBN4.FNT with PROPBN4.FNT - Files compare ok. | Comparing PROPBN5.FNT with PROPBN5.FNT - Files compare ok. | Comparing PROPBN6.FNT with PROPBN6.FNT - Files compare ok. | Comparing PROPNI1.FNT with PROPNI1.FNT | Files are not the same size. Compare them anyway (Y/N) ? N | Comparing PROPNI2.FNT with PROPNI2.FNT - Files compare ok. | Comparing PROPNI3.FNT with PROPNI3.FNT - Files compare ok. | Comparing PROPNI4.FNT with PROPNI4.FNT - Files compare ok. | Comparing PROPNI5.FNT with PROPNI5.FNT - Files compare ok. | Comparing PROPNI6.FNT with PROPNI6.FNT - Files compare ok. | Comparing PROPNN1.FNT with PROPNN1.FNT | Files are not the same size. Compare them anyway (Y/N) ? N | Comparing PROPNN2.FNT with PROPNN2.FNT - Files compare ok. | Comparing PROPNN3.FNT with PROPNN3.FNT - Files compare ok. | Comparing PROPNN4.FNT with PROPNN4.FNT - Files compare ok. | Comparing PROPNN5.FNT with PROPNN5.FNT - Files compare ok. | Comparing PROPNN6.FNT with PROPNN6.FNT - Files compare ok. | | If propnn6.fnt is bad (as you theorise), then it's been bad for 7 years. | | | | ============================================================================ | | 297) Wed, Aug 24, 2005 - 10:13:22 pm | | glennmcc> | BTW, | | I don't think it's possible to get a 'size 6' proportional font to work | in-place-of the 8x14 fixed-size system font. | | | | ============================================================================ | | 298) Wed, Aug 24, 2005 - 10:17:30 pm | | glennmcc> | Oh yeah... in addition to that..... after altering a font, you need to | re-write fontinfo.bin with 'arachne.bat -f' at the command prompt. | | | | ============================================================================ | | 299) Thu, Aug 25, 2005 - 7:36:15 am | | ray> | Your right, the 'html' fonts can't be used as system fonts, but | the font loading function doesn't know that, this is just to see | if they will load at all and they all will except for the size six | proportionals. | | So if the fonts are unchanged, then the next question becomes, | will they load in the older versions? | | | ============================================================================ | | 300) Thu, Aug 25, 2005 - 7:13:19 pm | | glennmcc> | If I'm not mistaken,........ Not in place of 8x14.fnt | | Try simply copying propnn6.fnt over-top-of propnn5.fnt | | BRB | | | ============================================================================ | | 301) Thu, Aug 25, 2005 - 7:20:15 pm | | glennmcc> | I just renamed propnn5.fnt as 6, and 6 as 5 | | No problem so-long-as I don't fontshift up-to a point of using it via | fontshift. | | | | ============================================================================ | | 302) Thu, Aug 25, 2005 - 7:26:50 pm | | glennmcc> | is now actually displaying propnn5.fnt renamed to 6 | is now actually displaying propnn6.fnt renamed to 5 | | If I fontshift 1 while using .... crash! | | | ============================================================================ | | 303) Thu, Aug 25, 2005 - 7:30:08 pm | | glennmcc> | Now renamed back to the original filenames. | | No problem using fontshift 1 with | | But try using fontshift 2 with ........ crash! | | | ============================================================================ | | 304) Thu, Aug 25, 2005 - 7:36:51 pm | | glennmcc> | Everything is just fine so-long-as we stick to the original situation | of a max fontshift of +1 | | It is only when we attempt to exceed +1 that any kind of problem happens. | | ============================================================================ | | 305) Thu, Aug 25, 2005 - 9:31:13 pm | | ray> | I just turned off my little guard hack, renamed propnn6.fnt to 8x14.fnt | and restarted with fontshift 0. Sure enough it wouldn't load the | font. It doesn't matter what you rename them, those fonts don't | work, you can avoid having them called by limiting fontshift to one | if and only if the font size specified in the document is 4 or less. | Make a document that calls a proportional size 5 font, put fontshift | to 1 and I'll eat my hat if it loads, 'cause that will then translate | to a size 6, which won't load. | | ============================================================================ | | 306) Thu, Aug 25, 2005 - 9:34:16 pm | | ray> | But ... | | It turns out that you are right to limit fontshift to one, but not | for the reason above. I'm allowing fontshifts up to 2, and for some | things its quite nice. Alas, unless I step down again before hitting | ESC or "R" or "F9" it hangs. It seems that the image of the page in | cache can't handle a shift of over one. God know why it will handle | even one, but it does. | | | | ============================================================================ | | 307) Thu, Aug 25, 2005 - 9:36:21 pm | | ray> | It bugs me. This sort of thing is why I'm suspicious of the whole | fontshift show, I don't thing Michael thought it through completely. | | | ============================================================================ | | 308) Thu, Aug 25, 2005 - 10:04:01 pm | | gregy> | ray....okay..be suspicious...but...some of us (myself, included) need | to be able to shift, when we get a page that is eensy...so....as long | as you guys can make it work, we need it.. | | | ============================================================================ | | 309) Thu, Aug 25, 2005 - 10:09:38 pm | | ray> | Ya, well I'll stick to a toggle of +1 or -1 for now, but this problem | with higher shifts is a bug for sure. And as for the size 6 fonts, | I wish Glenn would let us know if that problem is there for versions | before 1.71. ... | | Hey Gregy, you go back a ways, can you check it? If you've got an | earlier version, like good ol' 1.66, see you can set fontshift to | '2' and then load Glenn's most recent dial page. | | | | ============================================================================ | | 310) Thu, Aug 25, 2005 - 10:11:00 pm | | ray> | If you get the 'propnn6.fnt' error, then it is one heck of a mystery | when and where those fonts ever did work. | | | ============================================================================ | | 311) Thu, Aug 25, 2005 - 10:20:44 pm | | ray> | Here's my way of protecting against fontshift where it's not wanted: | | void HTMLfont DeF (char size, char style) | { | // RAY: 05-04-11: BUG: not "size == !SYSFONT" ... IDIOT. | if (size != SYSFONT) // not system font | { | int f; | // USE NEGATIVE NUMBERS TO INDICATE: NO FONTSHIFT! | // { RAY: 050825: If font size is negative, that indicates that fontshift | // is to be ignored. | if (size < 0) | f = size * -1; | else | // } | f = size + sUI.fontshift; // scale it: | RANGELO(f, 1) // too small, keep in range | RANGEHI(f, 6) // too big, keep in range | size = f; | | GUITICK.C========================================================= | | gSecondsSleeping = 0L; | }// end if screensaver | if (!FULLSCREEN) | { | x_setfill(0, xSILVER); | if (MOUSEY > x_maxy() - 30 && MOUSEX > x_maxx() - 230) | MouseOff(); | | // Blank out the clock. | x_bar(x_maxx() - 206, x_maxy() - 13, x_maxx() - 156, x_maxy() - 2); | x_setcolor(xBLACK); | // SO DONW HERE, ALL I'VE GOTA DO IS ADD A MINUS SIGN :-) | HTMLfont(-1, 0); // RAY: 050825: After glennmcc, ignore fontshift. | x_text_ib(x_maxx() - 206, x_maxy() - 15, (uchar *)cas2); | | if (MOUSEY > x_maxy() - 30 && MOUSEX > x_maxx() - 230) | MouseOn(); | } | | ============================================================================ | | 312) Thu, Aug 25, 2005 - 10:22:34 pm | | ray> | Glenn, | | That change you made in guidraw "htmlfont(3-user_interface...." | Where does that take effect, I'd like to confirm my fix above for | all places you identify. | | | | ============================================================================ | | 313) Fri, Aug 26, 2005 - 3:46:12 pm | | glennmcc> | propnn6.fnt works just fine with fontshift 1 and | | But you can not rename it as 8x14.fnt and expect it to work. | | ============================================================================ | | 314) Fri, Aug 26, 2005 - 4:07:36 pm | | glennmcc> | http://www.cisnet.com/glennmcc/images/5_0.gif | http://www.cisnet.com/glennmcc/images/5_1.gif | http://www.cisnet.com/glennmcc/images/6_0.gif | http://www.cisnet.com/glennmcc/images/6_1.gif | | ============================================================================ | | 315) Fri, Aug 26, 2005 - 4:09:43 pm | | glennmcc> | As you can see from those screen caps.. 5_0 is using the size 5 fonts. | | All of the rest of them are using the size 6 fonts. | | | ============================================================================ | | 316) Fri, Aug 26, 2005 - 4:10:39 pm | | glennmcc> | The size 6 fonts are too big to be used in place of the 8x14 system font. | | | | ============================================================================ | | 317) Fri, Aug 26, 2005 - 4:14:43 pm | | glennmcc> | But as those 3 screen caps prove.... the size 6 fonts ARE getting used | and they ARE working just fine. | | | | ============================================================================ | | 318) Fri, Aug 26, 2005 - 4:47:45 pm | URL | | glennmcc> | Just now uploaded this test page, http://www.cisnet.com/glennmcc/my-stuff/fo | | | | ============================================================================ | | 319) Fri, Aug 26, 2005 - 4:50:23 pm | URL | | glennmcc> | Cycle-through your fontshift settings and you'll see that the size6 font | does not get shifted upward. But size=5 does get shifted to size=6 | | | | ============================================================================ | | 320) Fri, Aug 26, 2005 - 6:28:57 pm | | Eric> | Ray, I just tried fontshift 2 with arachne 1.62 and it gave me | the same error message as with arachne 1.85. I also tried fontshift -2 | and it worked fine. The dial page lokked normal. | | ============================================================================ | | 321) Fri, Aug 26, 2005 - 8:33:50 pm | | ray> | Eric, | | Thanks. But Glenn says this isn't a problem, but I still don't buy | his reasoning. Glenn, that html page is a good start for figuring this | out, but how do you know the font file used is the one you show? | I realize that I can't use these fonts as system fonts but you miss the | point -- the font loading code just does what it's told, it has no | idea what a font is being used for, it just tries to load it, I could | use any name at all in that bachfile I showed you. | | | ============================================================================ | | 322) Fri, Aug 26, 2005 - 8:36:09 pm | | ray> | ... specificaly, now do you know those aren't the fixed fonts? | And if you are right, then explain why this doesn't work: | BRB | | ============================================================================ | | 323) Fri, Aug 26, 2005 - 8:39:35 pm | | ray> | in HTMLfont(): | | if (x_fnt_load(".\system\propnn6.fnt", 25, 1) != 1) | printf("Ray is right"); | else | printf("Glenn is right"); | | If those fonts are being used, how is it that the font loader can't | load them? | | | ============================================================================ | | 324) Fri, Aug 26, 2005 - 8:43:26 pm | | ray> | Make us a test page that let's us specify size, style and | fixed/proportional, and if that works with all the fonts, then | you win. In which case the above problem has me stumped. | | | | ============================================================================ | | 325) Fri, Aug 26, 2005 - 9:48:17 pm | URL | | glennmcc> | Done. | | | Now to check and see that the files I have indicated are in-fact those | being used. | | Rename probnn6.fnt as propnn6.fn_ | | copy fixedbi2.fnt to propnn6.fnt | | Now all of the places shown as using propnn6.fnt will now be displayed | in the fixed bold italic font. | | | | ============================================================================ | | 326) Fri, Aug 26, 2005 - 9:55:17 pm | URL | | glennmcc> | Like this. | | ============================================================================ | | 327) Sat, Aug 27, 2005 - 12:56:43 am | | Eric> | Boy, I'm having a hard time following this font stuff. I must be missing s | me. | Does fontshift just change the size of the lettering or does it change other | | | ============================================================================ | | 328) Sat, Aug 27, 2005 - 9:21:20 am | | ray> | Eric, | | It's suposta just change the size of the font, but font sizes must | be kept in range, so sometimes fontshift will be ignored. | | ============================================================================ | | 329) Sat, Aug 27, 2005 - 2:37:00 pm | | glennmcc> | A fontshift of -2 will use the fontfile which is 2 sizes smaller than | the current one. But of-course it can't go smaller than 1 | | So if the current fontfile being used is 2... a fonshift of -2 will use | fontfile 1 ... as will a current fontfile of 3 at -2 will use 1 | | We have 2 types in 4 styles in 6 sizes | | | types | | proportional | fixed | ________________ | | styles | | normal,normal | normal,italic | bold,normal | bold,italic | ______________ | | sizes | | 1 through 6 | | __________________ | | This page is using fixed normal,normal 2 which is .\system\fixednn2.fnt | | It is achieved in the HTML code via the use of 2 'tags' | | | | and | | | | | | ============================================================================ | | 330) Sat, Aug 27, 2005 - 2:42:08 pm | | glennmcc> | If we add the tag... fixedbn2.fnt will get used instead | | If we fontshift by +1 ...... fixedbn3.fnt | | If we add the tag ....... fixedbi3.fnt | | If we remove the tag but leave all of these in-place | | fontshift +1 | | | | | | ........ propbi3.fnt | | | ============================================================================ | | 331) Sat, Aug 27, 2005 - 2:45:21 pm | | glennmcc> | If we change to....... | | | | Then ther will be no change between these 2 settings | | fontshift 0 | fontshift +1 | | But fontshift -1 will use the size 5 fontfile and fontshift -2 will use | the size 4 fontfile. | | ============================================================================ | | 332) Sat, Aug 27, 2005 - 3:46:33 pm | | ray> | Glenn, | | I'm off for the weekend, but I'll look closely at your lattest html | latter. It seems to me that such a page should be kept around in | any case on general principals if someone want's to see all the | fonts available. | | | ============================================================================ | | 333) Sat, Aug 27, 2005 - 4:36:48 pm | | glennmcc> | There we go again... thinking alike. | | I was just thinking of including it in the next 'package' | and making a link to it on the F1 help page. | | | ============================================================================ | | 334) Sat, Aug 27, 2005 - 4:58:59 pm | | ray> | Damn, missed the boat ... litteraly ... and figuratively :-) | | OK, you're right, the steenking font works already. But only | sometimes! I changed my dialpage to fontsize.htm, set fontshift to 2 | and set my code to beep whenever propnn6.fnt is loading and the page | loads, and propnn6.fnt loads and the fontshift of 2 displays | exactly what you'd expect with no problems ... until you try to | load a page from cache, when she hangs. | | | ============================================================================ | | 335) Sat, Aug 27, 2005 - 5:05:33 pm | | ray> | So: why does your dialpage (pppglenn but you renamed it IIRC) | hang trying to load propnn6 but fontsize.htm has no problem with | the same font? Size of page in memory needed? Complexity? And why | can you load a page from cache with fontshift 1 but not 2? | Bugs! | | | ============================================================================ | | 336) Sat, Aug 27, 2005 - 5:45:27 pm | URL | | glennmcc> | There we go.... it now validates as "valid HTML" | | | ============================================================================ | | 337) Sat, Aug 27, 2005 - 5:51:40 pm | | glennmcc> | re:".... set fontshift to 2 ....." | | | There's the rub. | | Do not go above +1 ....... PERIOD | | Keep it a max of +1 ..... no problems........ EVER | | ============================================================================ | | 338) Sat, Aug 27, 2005 - 8:00:13 pm | URL | | glennmcc> | Some good info. | | ============================================================================ | | 339) Sun, Aug 28, 2005 - 12:11:14 am | | Eric> | Thanks Guys, I understand things a little better now. What started | out as a simple type size ajustment, like other things in life, has | turned into a complex witches brew. | | Ray, I'm sorry you missed your ferry! Is there another tonight? | | | ============================================================================ | | 340) Sun, Aug 28, 2005 - 3:41:23 pm | | ray> | Eric, | | They run every two hours, no problem :-) | | Glenn, | | Sure, we can avoid the problem by limiting fontshit but that's not | good enough for me. It's like saying "don't use 2nd gear in this | car, it's broken" I wanta know why you can't load a page from cache | with a higher shift, and I want to know why propnn6 works sometimes | and not others. Even if I can cope with these anomalies just fine, | for me, they remain bugs. | | | ============================================================================ | | 341) Sun, Aug 28, 2005 - 3:49:50 pm | | glennmcc> | Bad analogy. | | Here's a better one........ | | Use only 1st gear in in this car, since it it only has one gear.... | trying to JAM it into 2nd will break the fontshift lever. | | ============================================================================ | | 342) Sun, Aug 28, 2005 - 5:02:30 pm | | Eric> | Both good anaologies, I'm a firm believer in "Don't try to make your | car do something it can't do! At least untill you think it is fixed." | If your gas tank leaks when filled past half filled, then don't fill it | past half filled. It's when I don't know that something's broken, when | no one warns me, that I get into deep manure. | | | ============================================================================ | | 343) Sun, Aug 28, 2005 - 7:07:31 pm | | ray> | Sure, *until* you find the real fix, you guard against the problem, | no argument there, but Arachne *does* have a '2nd gear' it's just that | using it will cause a crash, and I'd like to know why. It's just the | sort of dude that I am, I gota know WHY. | | | ============================================================================ | | 344) Sun, Aug 28, 2005 - 8:05:59 pm | | glennmcc> | Eric, | | Nope... that's another bad analogy. | | It's more like.... | | Please don't try putting 20 gallons of gas into a 10 gallon tank. | | If you try.... it's gonna overflow all over the place and will most | likely explode when the overflowing gas fumes come into contact with the | cigarette the guy sitting in the chair in front of the 7-11 is smoking. | | | | ============================================================================ | | 345) Mon, Aug 29, 2005 - 2:28:44 am | | Eric> | Hey Glenn, That's not funny! A gas station I used to use almost every d | it kept seeping out gas until the spout was hungup. There was always a spot | nicely to the guy behind the 4 inch plexiglass 3-4 times over several weeks | service clerk said someone | had already started a fire at that pump a couple of days previously and the | fire. I hope your | aware, many aren't, that if you put you lawn mower gas can in the back of | your pick-up on one of those rubberized beds that it can build up a static | electrical charge which can ignite the gas when your go to take it out of th | with the gas pump spout. The same can happen, I'd guess you already know, i | don't touch a ground before you pump the gas. All these plasic & rubber car | | | ============================================================================ | | 346) Mon, Aug 29, 2005 - 3:25:37 am | | gregy> | Eric...you oughta see the explosion that can happen (has happened) from | static electricity buildup when filling the tank on a BIG military | fuel tanktruck.........that's why it's ALWAYS manditory to | ground things together. | ... | The lawnmower gascan explosion is so common, that the stations up here | mandate that you fill them on the ground, NOT in the back of your | vehicle... | | | ============================================================================ | | 347) Mon, Aug 29, 2005 - 3:30:39 am | | Eric> | Glennmcc, What happened to the indicator in the lower left corner | when one presses F4 to view/edit webpage source? It used to display | the name of the file in the cache, such as 12345678.tmp. It has disappeared | | ============================================================================ | | 348) Mon, Aug 29, 2005 - 3:50:09 am | | Eric> | Gregy, The way I understand it, the ignition can even occur just by | taking the gas can out of the truck if one is especially unlucky. I one | time came really close to blowing up my car in Ann Arbor, MI (U.of Mich) | In my hurry to get to class, someone pulled out in front of me and made me s | made electric | trunk opener, I found my trunk filled with gas from my emergency gas can | which had tipped over. I was lucky the wires that I just had twisted togethe | probably wouldn't be writing this right now. That 1960 Chevy had a trunk big | | | ============================================================================ | | 349) Mon, Aug 29, 2005 - 8:54:57 am | | gregy> | Eric...well, you know, the Good Lord watches over fools, drunks, and | college kids with gas cans in their trunks....... | ... | | | ============================================================================ | | 350) Mon, Aug 29, 2005 - 4:29:25 pm | | glennmcc> | Eric, | | re: F4 page info | | Just checked.... I got it on my screen. | | Which version of core.exe are you using that does not have it. | (version number, filesize, date/time ? ) | | ============================================================================ | | 351) Mon, Aug 29, 2005 - 5:48:25 pm | | Eric> | Glenmcc, I'm using v185;GPL,386+ which it says in lower right corner | right after the time. On the lower left it says 0096/0096:001 | when I F4'd this page. The core is from eric.zip which you gave | me so that email would work. The numbers I see appear to indicate | editor functions. | | | ============================================================================ | | 352) Mon, Aug 29, 2005 - 6:03:16 pm | | Eric> | Glennmcc, I found it! The whole page doesn't fit on the screen and | it is now in the middle of the page if I TAB out of the editor and Page | Down which one can't do if cursor is in the edit box/screen. Also, | it looks like something is written in lower left below the editor info | but only the very top of the letters are visible on the screen. | | ============================================================================ | | 353) Mon, Aug 29, 2005 - 6:11:26 pm | | Eric> | Glennmcc, Even in normal screen mode there appears to be a line of | text below the lower border of the screen which just barely shows the | top of the letters and which seems to be inaccessable for viewing. | | ============================================================================ | | 354) Mon, Aug 29, 2005 - 7:30:58 pm | URL | | glennmcc> | OK, it's there but not being displayed where it's supposed to be. | | Please try this one and see if it shows correctly. | http://www.cisnet.com/glennmcc/ar187.zip | (disregard and delete the 'hacked' copy of goph2htm.exe) | | Sam fixed the problem with gopher on his end... the hack is no longer needed | | | ============================================================================ | | 355) Mon, Aug 29, 2005 - 11:36:08 pm | | gregy> | okay...I'll dump that gopher2htm, also... | .. | Eric, yes...I looked after I just now read your post, and on the F4 | screen, itself, just below the "CANCEL" button, there's a box with the | cache/itemnumber.htm in it...telling us which object in the cache is | the one we're editing... | | | ============================================================================ | | 356) Tue, Aug 30, 2005 - 2:36:05 am | | Eric> | Glennmcc, With v1.87, I no longer see the tops of a line of print | at the lower border but my screen seems to be 1 line too short. | I only see the "cache\12345678.htm" if I TAB out of the editor | screen and page down and there is only one more line to page | down to. | | ============================================================================ | | 357) Tue, Aug 30, 2005 - 2:44:05 am | | Eric> | Glennmcc, I also only see the EXT in the external editor box | unless I page to right. I am using the DEFAULT video settings | from Setup wizard, this may have something to do with it. | | | ============================================================================ | | 358) Tue, Aug 30, 2005 - 3:15:04 am | | Eric> | Glennmcc, I just downloaded 49 emails with no difficulties | with the v1.87;GPL,386+ core and the new insight. The download | went as smooth as a hog's ear. | | ============================================================================ | | 359) Tue, Aug 30, 2005 - 4:55:18 pm | | ray> | This background download stuff: Glenn's change seems to work OK with | that that test page of his, but not with Google, it still shows | the page before it's finished downloading everything. Can anyone | confirm? | | | | ============================================================================ | | 360) Tue, Aug 30, 2005 - 6:37:58 pm | | ray> | testing | | | ============================================================================ | | 361) Tue, Aug 30, 2005 - 8:29:41 pm | | glennmcc> | At google.... | | Did you use the enter key on your keyboard ? | | Or did you use the left mouse button ? | | | | ============================================================================ | | 362) Tue, Aug 30, 2005 - 8:31:07 pm | | gregy> | Glenn...Ray...see my STATUS REPORT over on /cts/ | | | ============================================================================ | | 363) Tue, Aug 30, 2005 - 8:35:08 pm | | glennmcc> | Just now tested google. | | It works exactly as I said.... | | Enter downloads in the backgorund and displays after the D/L has | completed. | | The left mouse button works as it aways did.... downloads in foreground | and displays as it is downloading. | | | | ============================================================================ | | 364) Tue, Aug 30, 2005 - 8:40:17 pm | | gregy> | nope...maybe on your box...but what I reported in definitely what's | happening on both my boxes....multiple times. I don't report bugs | until I've reproed them several times, and check "operator malfunction" | variables to be sure it's the software, and not my fingers..... | | | ============================================================================ | | 365) Tue, Aug 30, 2005 - 9:00:37 pm | | glennmcc> | Did you clear your cache before begining the tests ? | | Cache cleared before each new attempt ? | | ============================================================================ | | 366) Tue, Aug 30, 2005 - 9:01:47 pm | | glennmcc> | 'Enter' will still instantly load the file from your cache if it exists ther | | | ============================================================================ | | 367) Tue, Aug 30, 2005 - 9:02:43 pm | | glennmcc> | 'R' has not been changed either. | | 'R' will still 'display as downloading'. | | ============================================================================ | | 368) Tue, Aug 30, 2005 - 9:05:08 pm | | glennmcc> | Del, G, Enter | | Will act the same as Enter on a link. | | ============================================================================ | | 369) Tue, Aug 30, 2005 - 9:22:44 pm | | gregy> | yeah...I always clear cache before testing. I mean, my homepage is | MAIL.HTM with a "LINKS" link, which loads a page of links. I F8, click | the "links" link, displays the page, and the yahoo link is there for | me to LMBclick, or put the cursor over and so...you decide... | ..am I clearing the cache...?? | | | | ============================================================================ | | 370) Tue, Aug 30, 2005 - 9:25:36 pm | | gregy> | In fact, I have a little two-step dance I go thru... | F8 | ALT-E, "h" (copies a "reset" HISTORY.LST over), | "x" (returns me to Arachne) | | | | ============================================================================ | | 371) Tue, Aug 30, 2005 - 9:31:30 pm | | gregy> | and...when I'm downloading...I'm always watching that status bar, and | I'd spot the "loading page from disk" message...so...nope..I'm getting | a REAL download..... | | | ============================================================================ | | 372) Tue, Aug 30, 2005 - 9:58:43 pm | | ray> | Glenn, | | I'm using the mouse or the Enter key (with and without CTRL) | from inside the history list. With both your 1.83 core, modified | as you showed, and with mine similarly changed, Google shows the | page while the progress idicator still shows about one or two | seconds of further downloading. Could this be due to some other | setting in the .cfg? | | | | ============================================================================ | | 373) Tue, Aug 30, 2005 - 9:59:30 pm | | ray> | Oh, and like Gregy, I always clear my cache between anything that | remotely a test. | | | ============================================================================ | | 374) Tue, Aug 30, 2005 - 10:00:52 pm | | ray> | But, like I said, your code works perfectly with your own test page | so it seems to be an improvement even if it's not perfect, or | maybe there's some other factor. | | | ============================================================================ | | 375) Tue, Aug 30, 2005 - 10:01:07 pm | | gregy> | no, Ray...what you're seeing is the default, old, Arachne performance. | ... | What you should see, is the history.lst continuing to display, which | you watch the download counter going up until pagedownload is complete, | THEN the page should display, all at once, complete. | | | | ============================================================================ | | 376) Tue, Aug 30, 2005 - 10:03:36 pm | | gregy> | yeah...this new functionality works slicker-n-snot...I actually saw it | on the news.yahoo page. Old page continuing to display, which d/l | counter was showing the download...until the end.......then.. | ..Arachne decided to display the yahoo page....:((((( Molly Hold THE | HORSES..!!!!..... | | | ============================================================================ | | 377) Tue, Aug 30, 2005 - 10:04:47 pm | | gregy> | oops...."...continuing to display, while you watch..." etc..... | | ============================================================================ | | 378) Tue, Aug 30, 2005 - 10:05:12 pm | | ray> | So you confirm it's ok with yahoo, hmmm. | | | | ============================================================================ | | 379) Tue, Aug 30, 2005 - 10:07:00 pm | | gregy> | no, I confirm that I could see the download going on, while my "links" | page (same as a history.lst) continued to display, until the d/l was | finished, then Arachne tried to (should have) display the downloaded | yahoo page...but...it all blew up at that point.... | | | ============================================================================ | | 380) Tue, Aug 30, 2005 - 10:08:27 pm | | gregy> | BUT...the "download completely first/display last" new functionality | was working up to the point of "display"..... | | | ============================================================================ | | 381) Tue, Aug 30, 2005 - 10:11:53 pm | | gregy> | Here...I'm not coherently describing it: | 1. clear cache, reset history.lst, exit shellout, return to homepage | 2. click "links" display my links page | 3. position cursor over the news.yahoo link and press | 4. linkspage continues to display, while download counter on the status | bar shows page download beginning. | 5. news.yahoo page download completes, and Arachne tries to display the | page. | 6. BLEWEE..!!!! | | | ============================================================================ | | 382) Tue, Aug 30, 2005 - 10:32:37 pm | | ray> | Hmmm, well this is Glenn's baby so I'll let him figure it out. Here, | I don't get that, I just don't get any change at all with Google. | This mightn't be some hidious thing that's dependant on some other | keyword or some ideosyncrasy with ISP's by any chance? Recall I never | have the sorts of problems that you have with these downloads, so | maybe my ISP is doing something different :( | | Hey it's so unfair: Yesterday it looked like N.O. was almost | spared, today on the evening news I see that it's one heck of a mess | down there. | | | ============================================================================ | | 383) Tue, Aug 30, 2005 - 11:04:22 pm | | gregy> | No, Ray...Google is a page that "downloads all" before displaying... | use the: | http://news.yahoo.com/ | | That will give you a clear indication of the differences. I could | actually see the new functionality working, until the download was | done and Arachne tried to display the page. | | | ============================================================================ | | 384) Tue, Aug 30, 2005 - 11:07:58 pm | | gregy> | No, Ray...NewOrleans WAS alright, until the levees, supposedly | maintained by the "local" authorities and our great Army Corp of | Engineers, on the Lake Poncatrain side sprung two leaks, and let the | lake flood in to the city. NOW they've got a mess....oh, well, what | else can we expect from that crap-head bunch of facistic "BM-gineers" | in the Corp plus the "local authorities" most of whom are traditional | dems....scumbag porkbarrelers and corruption-mongers. | | | | ============================================================================ | | 385) Tue, Aug 30, 2005 - 11:09:44 pm | | Eric> | Ray, Your not kidding. New O. L. really is a mess. Another | example of mankind trying to manipulate Nature. Makes a lot of sense | to build a city below sea level. Especially on the sea coast. | The last report I heard said 2 levies broke and the whole city's | under water. Those poor people! | | ============================================================================ | | 386) Tue, Aug 30, 2005 - 11:18:00 pm | | gregy> | Eric...the problem wasn't "mankind trying to manipulate Nature."..The | actual problem was "mankind failing to carry thru on his necessary | actions to manipulate nature"...to whit...they didn't maintain the | levees on the lake side. For goodness sake....!!! it wasn't even ON | the sea side. Just the rise of level in the lake and the shoddy, not | well maintained levee on the lake side breached in two places.... | ... | ...and...the news says 80 percent.....and recent pictures of the French | Quarter showed, presently, minimum flooding there.... | | | ============================================================================ | | 387) Tue, Aug 30, 2005 - 11:21:45 pm | | gregy> | Oh, BTW...it ain't "on the coast" it's many miles up the river at the | start of the delta. And...he he he...it was built there because that | was the best place to build it, if you were a frenchman and wanted to | control the traffic on the river (can't control all the little delta | branches, so built the blocking point upstream on the river)... | ... | And, of course, once you've got sumpin already started or built, it'll | take an act of god to close it down and move everyone out of there. | | | | ============================================================================ | | 388) Tue, Aug 30, 2005 - 11:25:03 pm | | gregy> | Ray, for an example of a page that does NOT try to display before it is | all downloaded, try: | http://www.opinionjournal.com/ | | | | ============================================================================ | | 389) Tue, Aug 30, 2005 - 11:26:14 pm | | Eric> | Gregy, Well, from what I've seen previously on public TV channel, | those levies are just heaps of dirt. Get a little leak and you | get a fast destruction of the levy by erosion. Kind of like | the Johns Town flood. From what I understand, all those water | ways around N.O.L. are artificially contained by levies etc. | Can't bury people there, You lay them to rest in little "boats". | | ============================================================================ | | 390) Tue, Aug 30, 2005 - 11:33:05 pm | | Eric> | Gregy, That Wall Street page seemed to download first for | me before it displayed. I don't necessarily like it that | way but that's the way it worked just now. I haven't changed | anything in any config files Just swapped cores and insight, etc. | | | ============================================================================ | | 391) Tue, Aug 30, 2005 - 11:35:51 pm | | Eric> | However, the little green gauge stalls at 41% and never fills | in the other 59% . | | ============================================================================ | | 392) Tue, Aug 30, 2005 - 11:46:29 pm | | Eric> | Gregy, Technically, your right, but I consider that delta a sea | coast. I understand why they had a military post there, that made | sense. But all the rest of construction there is foolish. Man claims | to be rational, but he is probably the most irrational of all the | animals. You and I and the rest of the country are paying for it. | the price of gas around here is now over 3.00 and still rising! | | ============================================================================ | | 393) Wed, Aug 31, 2005 - 12:02:33 am | | gregy> | Eric..the cost of gas it because of proggy-libs, like Ex-Dash-hole, who | have been buggering with the environ-bationists for the past umpteen | years, denying us the ability/right to drill in most areas under our | possession, blocking us from increasing our refinery capacity, and | totally stopping us from building/upgrading any nuclear reactors. | ... | so...we, the people, are indeed paying...paying for the fornicative | buggery between the proggy-lib dems and the environ-bationists. | ... | And I, and (it seems) most of the American people, have about had it. | | | | ============================================================================ | | 394) Wed, Aug 31, 2005 - 12:05:40 am | | gregy> | Eric, yes, the opinionjournal.com page is an example of a page that does | NOT seek to "partial-display" itself before it is completely downloaded. | ... | OnTheOtherHand, news.yahoo.com is a page that rudely "forces" partial | display by Arachne, before even 1/5th of the page is downloaded. | ... | The fix Glenn did is designed to FORCE Arachne to do a complete d/l | before it displays anything. | | | ============================================================================ | | 395) Wed, Aug 31, 2005 - 12:09:06 am | | glennmcc> | Here in NE Ohio... $2.50/gal | | I'll wager it rises to 3 bucks by Fri or Sat... perhaps sooner. | | ============================================================================ | | 396) Wed, Aug 31, 2005 - 12:11:22 am | | gregy> | Eric..."get a little leak"...yes..!!! You must not be a farmboy... | ...that's the way most of our resevoirs washout.......the stupid | farmer doesn't watch/maintain his dams, the gophers, muskrats and the | badgers burrow them out, they get a little leak....and....voila..!!! | ... | Just the way those NON-seaside levees (on the LAKE, not the sea side), | without any real storm surge or wave erosion, just a little extra | volume of rain and level in the lake....leaky, leaky...and ...voila..!!! | ..NewOrleans is 80percent flooded.... | ... | And you can bet the Corp of Engineers and the local pols will throw | frantic dirt in every direction possible to try to hide their own | complete and direct culpability.. | | | | ============================================================================ | | 397) Wed, Aug 31, 2005 - 12:12:44 am | | gregy> | I'll back Glenn's bet, and take any wagers he can't afford to cover... | ... | | | ============================================================================ | | 398) Wed, Aug 31, 2005 - 12:16:58 am | | gregy> | Oh, and BTW, Glenn...that rapid rise won't be due to the actual gas | situation...it'll be due to the gouging of the stinking oil companies. | .. | NOT your local stations....but the companies, themselves.. | | | ============================================================================ | | 399) Wed, Aug 31, 2005 - 3:07:21 am | | Eric> | The sad thing is your all right!!! | | | ============================================================================ | | 400) Wed, Aug 31, 2005 - 3:52:00 am | | glennmcc> | Oh yeah, got more wager to make. | | The 'legal price gouging' on building materials will soon happen as well. | | Sheets of plywood, dimensional lumber, nails, screws etc... etc | | Plywood will most likely double in price by next monday. | | ============================================================================ | | 401) Wed, Aug 31, 2005 - 7:59:18 am | | Eric> | Glennmcc, I just downloaded 99 emails from Juno with NO aborts. | Using core v1.87;GPL,386+ and the new insight that came with it. | | ============================================================================ | | 402) Wed, Aug 31, 2005 - 1:42:48 pm | | glennmcc> | Great !!! | | I'd say that's a pretty good test. | | Let me know when you have a chance to try that same type of test on this one | http://www.cisnet.com/glennmcc/ar187new.zip | | | ============================================================================ | | 403) Wed, Aug 31, 2005 - 3:40:30 pm | | ray> | Gregy, | | I dunno Sarge, I saw a complete documentary on the situation in N.O. | a couple of years ago that said, basicaly, that the city was | doomed -- that it was just a matter of time. What with the | city sinking due to groundwater use, the sea rising due to | (sorry) global warming and with storms becoming both more frequent | and stronger, the point had actually *passed* where the city | could be reliably protected at anything like an afordable cost. | | | ============================================================================ | | 404) Wed, Aug 31, 2005 - 3:44:09 pm | | ray> | As for the 'proggy-lib dems and the environ-bationists ... I | resemble that remark ... what *is* a proggy-lib dem anyway? < g g g > | | | ============================================================================ | | 405) Wed, Aug 31, 2005 - 3:46:33 pm | | ray> | BTW, I don't want to say much since my HTML is still almost | nill, but, since temparea.tmp stores all the info in the text area, | when this page is reloaded, wouldn't it be possible to have this | page itself call up textarea.tmp and thus restore the page to | it's full content after dialing? See what I mean? Instead of | getting Arachne to do it, maybe the page itself could do it. | Or not, I dunno. | | | | ============================================================================ | | 406) Wed, Aug 31, 2005 - 5:49:17 pm | | glennmcc> | No problem. | | | | | ============================================================================ | | 407) Wed, Aug 31, 2005 - 5:52:56 pm | | glennmcc> | re:"...since my HTML is still almost nill, ..." | | Do you mean that your knowledge of HTML almost does not exist ? | | | | ============================================================================ | | 408) Wed, Aug 31, 2005 - 8:52:33 pm | | gregy> | Ray, the modern liberals have made that a "dirty word"..even they are | uncomfortable being called that....so they've run for cover under their | "new" appelation "progressives"....but...they can't hide..!!!... | ...gregy is on to their schemes.......therefore...proggy-lib... | (rhymes with froggy-legs......), and, of course, dems...we all | know that means dem dems.... | | | | ============================================================================ | | 409) Wed, Aug 31, 2005 - 8:55:57 pm | | gregy> | Re: N.O....don't know about "the city is sinking" but I DO know about | "storms getting worse"....being charitible, I must give a resounding | BALDERDASH...!!!!!!!!!...Of course, you read that in what we call the | MainLeftStreamMedia....their all...ALL...environ-bationists...and can't | read history and data.... | ... | Let's just concede, to you and all other proggy-libs...that this is ALL | Bush/Rove/Cheney/etc's fault....then we can get on with the rest of our | lives, where the REAL reality resides.... | ... | | | ============================================================================ | | 410) Wed, Aug 31, 2005 - 8:57:50 pm | | gregy> | however...NO...I ain't gonna, EVER, EVER...vote for stupid-yoto, nor | let my congressmen/senators vote for it either... | | | ============================================================================ | | 411) Wed, Aug 31, 2005 - 8:59:12 pm | | gregy> | AND #2...this isn't the board for this kinda discussion, either... | ... | so...I'll post my next post over on my MISC board..... | ... | /brd/misc/mindex.php | | | | ============================================================================ | | 412) Wed, Aug 31, 2005 - 9:16:48 pm | | gregy> | sheesh..!!! It's already 31AUG...gotta archive these boards, tonight | ... | | ============================================================================ | | 413) Wed, Aug 31, 2005 - 11:02:42 pm | | ray> | | | | Gregy, | | Oh yeah, I guess I was a bit off topic ;-) | | Glenn, | | Yup I meant that my knowledge of ... | Anyway would fixing the problem be as simple as changing the board | like that? If so let's try it, just remembering to kill the | temparea.tmp when there's no need to retain the contents ... or | something like that. | | | ============================================================================ | | 414) Wed, Aug 31, 2005 - 11:05:27 pm | | ray> | Ha! Sly dog! Almost caught me there. But saying: | "The magnitude of my knowledge of HTML aproaches zero" would | sound a bit pedantic . | | | | ============================================================================