DOS Coders1 Message Board ______________________________________________________________________________ ,----------------------------------------------------------------------------- | | Arachne DOS Coders1 Message Board | | | | | | | | [ For the Operational discussions of the Arachne C Coders ] | | | | ------------------------------------------------------------+-------+---+ | | | ============================================================================ | | Sun, Oct 03, 2004 - 9:14:40 am | | phantom> | Hi Ray! | | Yeah, boards have advantages and disadvantages. The latter includes | the fact that they don't deliver to your mailbox, so if you don't | remember to visit regularly, the dialogue does tend to diminish ... | | | ============================================================================ | | Sun, Oct 03, 2004 - 3:16:45 pm | | ray> | Joe, | | Have you noticed how this tends to cycle? ... long periods of | no activity on the board, then 10 messages per day, then another | long dead spell. | | BTW, your the XSWAP guy, so how would a madman like me unload | arachne.cfg from memory? In theory, once I've got all it's information | safely stored in a .dat file, like I do now, there's no need to | keep it around unless the internal configuration routine is run. I | want to try unloading it to test my theory. | | Oh, my favorite niece just landed in Sydney to visit the relatives of | her prospective husband -- she's marying an Ausie bloke, the silly girl. | Anyway, treat her right ;-) | | | ============================================================================ | | Mon, Oct 04, 2004 - 8:33:13 pm | | ray> | Joe, | | Is there some simple test you can suggest to see whether I've got | full-length cookies working in B5? I have no real reason to | suspect not, but I'd like to be sure. | | | | | | ============================================================================ | | Mon, Oct 04, 2004 - 8:49:19 pm | | Joe> | Ray, | | 1. Eeek! I haven't looked at that XSWAP horror in months! | I see where you're going with this config info stuff ... I'll | see what's required and get back to you. You realize of course | this means you will then need to reload/modify/save this stuff | each time an internal configuration routine is run. | | 2. I'm sure your niece will have a great time, the people here are | very hospitable. | | 3. If you can log-in to Yahoo Mail, then full-length cookies are | probably working (they should be, as I imported the corresponding | code myself into B3, and I'm pretty meticulous with coding). | | | | ============================================================================ | | Tue, Oct 05, 2004 - 7:24:29 am | | ray> | Hi Joe, | | 1. Roger all that. It's looking like the internal config thing | won't be too much of a problem (touch wood) -- it works fine now | even with my using a pre-parsed binary file, to reload from. It's | quite a bit faster too. | | 2. Yes, so she is finding. | | 3. Will check that. I still can't upload a file, though -- I get a | message about not being alowed to upload a file > 512K even though | the file is much smaller than that. I'm not suspicious about B3c, | just about B5 since there was plenty of room for mistakes there. | ddc®! | Thanks, | | | ============================================================================ | | Tue, Oct 05, 2004 - 7:26:24 am | | ray> | Great balls of fire! | | It's the stinking IE bug risen from the dead :-( | | | | ============================================================================ | | Tue, Oct 05, 2004 - 1:31:23 pm | | ray> | OTOH my email via Yahoo worked, so I'll take that as | confirmation that I'm OK on the cookies front; thanks Joe. | | | | ============================================================================ | | Tue, Oct 05, 2004 - 4:39:55 pm | | glennmcc> | Looks to me as if you have created a new version | of the internal editor bug. | | Yours is now a registered trade mark ;-) | | ============================================================================ | | Tue, Oct 05, 2004 - 4:41:25 pm | | glennmcc> | It has not happened to me even a single time since compiling for 386+ | | | ============================================================================ | | Tue, Oct 05, 2004 - 8:06:57 pm | | ray> | It's the first time for me since J4, so there isn't much of | a statistical base to work from. If it starts cropping up | again routinely then I probably have fouled something up but that's | good, since it would give me a change to track the blasted thing | down. | | This message is Registered! (Don't even *think* of copying it :-) | | | | ============================================================================ | | Tue, Oct 05, 2004 - 8:28:22 pm | | glennmcc> | Copy it ???? | | Such a thing would never occur to me. ;-) | | ============================================================================ | | Wed, Oct 06, 2004 - 5:50:59 pm | | Joe> | Ray, | | I've had a look at unloading the Arachne config from memory. AFAIK, | you should be able to achieve this with : IEclearF(A,0) | | However, since the Arachne config resides in XMS, you won't gain | any (much) free conventional memory by doing so (yes, that should | have occurred to me earlier!;-). | | | ============================================================================ | | Wed, Oct 06, 2004 - 6:26:38 pm | | ray> | Thanks Joe, I'll try it but of course it ruins my party a wee bit | if its XMS -- I was hoping to save some conventional memory. Anyway, | live and learn. | | | | ============================================================================ | | Wed, Oct 06, 2004 - 6:27:56 pm | URL | | glennmcc> | But, it _would_ cure this problem. (would it not ?) | | http://www.cisnet.com/glennmcc/smiley.bug/ | | ============================================================================ | | Wed, Oct 06, 2004 - 6:30:20 pm | | glennmcc> | Framed pages will no longer be able to corrupt the copy that's in memory | if it ain't in memory no more. | | | ============================================================================ | | Wed, Oct 06, 2004 - 9:41:16 pm | | ray> | That's a good point. I've never had that problem, not even in the old | days, so it's not something I think about, and of course it's a | kludge, not a proper cure, to protect arachne.cfg by unloading it -- | as with the blasted IE bug, even if we can avoid it, we still haven't | really found out what was making it in the first place -- still, in | the interim, any port in a storm as the saying goes. | | At present, this | project is hampered by the stupid fact that if strings are pre-parsed | and stored in a structure image file, the pointers to those | strings become | invalid during many of Arachne's countless re-loads if the | absolute address of the program changes (as with dialing) -- Michal | can probably fix this -- he says the difficulty is a near/far pointer | thing. The second frustration is that some of the config options | are only used by miniterm.exe and others only by insight.exe, and I | won't be happy until pre-parsing works for everything ... once. | | | | ============================================================================ | | Thu, Oct 07, 2004 - 6:22:16 pm | | glennmcc> | The seems to me to be a nonsequitur. | | Why would would you have core.exe pre-parse variables | from arachne.cfg which are not used by core.exe ? | | ============================================================================ | | Thu, Oct 07, 2004 - 6:32:29 pm | | glennmcc> | The natural solution is to move those variables over into 2 new files. | | miniterm.cfg and insight.cfg | | Just as we now have our lsppp.exe options in lsppp.cfg | | This is currnet line in arachne.cfg ..... | | Dialer @lh lsppp.exe | | And this is lsppp.cfg | | /x | /i:4 | /b:bx800 | /d:330-423-4005 | /B:115200 | /V:60 | /U:glennmcc@cisnet.com | /P:*******>PPP.LOG | _______ | | As you see, all of these options that used to be on my dialer line in | arachne.cfg have now been moved into lsppp.cfg | ;/x /i:4 /b:bx800 /d:330-423-4005 /B:115200 /V:60 /U:glennmcc@cisnet.com /P: | | And last, my current ppp.log file......... | | PPPD status: IP address set to 4.224.207.142 | | Now all we need is for Michael to find and send us the SRC code for miniterm | so that we can change it to use miniterm.cfg instead of arachne.cfg | | | | ============================================================================ | | Thu, Oct 07, 2004 - 7:51:33 pm | | ray> | Glenn, | | Thoughtfull remarks. To clarify, I don't now pre-parse the | non-core options for exactly the reason you suggest, however | to avoid the absurdity of core.exe, insight.exe and miniterm.exe | all having to parse arachne.cfg independantly, it makes sense that | the day would come that all of these can access the information they | need from one pre-parsed file -- especialy in view of the fact that | insight.exe should almost certainly be folded into core.exe. | | In the mean time, as you suggest, it would be possible to | break up arachne.cfg as you discribe, but there would be obvious | implications to backwards compatablity. I think the best long | term solution is one pre-parsed file that serves all needs, but at | this stage, nothing is for sure. | | ============================================================================ | | Thu, Oct 07, 2004 - 7:54:54 pm | | ray> | BTW, as far as problem #1 above, it very much looks like the | reason for the need to re-synchronize the string pointers is that | after arachne has restarted, if any SET's have been made, this | apears to change the absolute address of the loading of core.exe and | consequently invalidate the existing pointer addresses. I thought | that environment space was inviolate so it seems very strange to me | that adding variables would have any effect on pointer addresses, | but that seems to be the case. | | | ============================================================================ | | Thu, Oct 07, 2004 - 7:58:13 pm | | ray> | OH, and as far as miniterm is concerned, why don't we just use | lsppp insted? -- I understand it has its own dialer. I know that | Joe has answered this question before, but I can't remember what he | said. | | | | ============================================================================ | | Thu, Oct 07, 2004 - 8:33:48 pm | | glennmcc> | Yes, | | Lsppp.exe is also a dialer in addition to packet driver. | | I have been using lsppp.exe to do both for over a year now. | | And as to needing the SRC code for miniterm.... | | Nevermind... I hacked her. | | She now takes her options from minterm.cfg | | --- my current miniterm.cfg --- | | ;This is MINITERM.CFG for Arachne version 1.79+ | | Port 0 | Irq 4 | Base 0xb800 | Mode 8N1 | Speed 115200 | InitString ATf50 | DialString ATDT | PhoneNumber 330-423-4005 | Autologin No | WaitFor1 | Response1 | WaitFor2 | Response2 | WaitFor3 | Response3 | PPPusername glennmcc@cisnet.com | PPPpassword ******** | | ;Port values: 0=non-standard serial (COM) port, 1-4 standard ports | ;Base values: 0x??? (eg. 0x3e8) - Base and Irq are used for "Port 0" | ;Mode values: databits/parity/stopbits (eg. 8N1) | _______________________________ | | ============================================================================ | | Thu, Oct 07, 2004 - 8:38:17 pm | URL | | glennmcc> | Grab a copy if you'de like. ;-) | | ============================================================================ | | Thu, Oct 07, 2004 - 9:49:34 pm | | ray> | Scary (for both of us :-) as it is, I was about to propose | hacking miniterm myself! | | Anyway, excelent work, I'll try it out manana. | | But, if we can use lsppp to do it all then why not just forget | miniterm? ... it's slowly coming back to me ... isn't it something | about only miniterm being able to ... what was it ... show the | 'terminal mode' screens?? | | | | ============================================================================ | | Fri, Oct 08, 2004 - 2:28:56 am | | Joe> | Hi all, | | Firstly, the reason why Arachne/Core wants to have all the config | stuff for Insight and Miniterm is that this allows the Arachne GUI | to make whatever changes are required by the user in a user-friendly | fashion. No need (in this case) for users to delve into the config | file and manually clobber (I mean change;-) stuff. We need more stuff | handled this way, not less (IMHO). | | For this reason, as well as the terminal window for diagnostics, as | well as the Miniterm dialler being more reliable with some modems (as | reported by some users), I use Miniterm for dialling with LSPPPCFG, | although I could just as easily have used LSPPP itself. | ( LSPPPCFG = No config editing required! :-) | | | ============================================================================ | | Fri, Oct 08, 2004 - 7:33:45 am | | ray> | Yup, separate config files might be of use in the short term, but in | the long haul we should be working towards integration, not | fragmentation. Still, it's miniterm.cfg for me latter on today :-) | | | ============================================================================ | | Fri, Oct 08, 2004 - 7:38:47 am | | ray> | Which reminds me, speaking of integration, I notice that | insight makes no effort to load arachne.cfg into memory nor to | perform any IE magic on it -- it just reads the stupid thing the | old fashioned way. I'm wondering, if all the config files were | loaded into ramdisk, how much slower would this be than all the | XSWAP stuff that core performs? You know what I'm thinking. | | | | | ============================================================================ | | Fri, Oct 08, 2004 - 4:39:05 pm | | glennmcc> | Yep, hacking of .EXEs and splitting the miniterm stuff over into | a manually created/manually edited/manually updated miniterm.cfg | is fine for us hackers. | | But 'for the masses'.......... it all needs to remain in arachne.cfg | | As for how core.exe reads arachne.cfg vs how insight.exe does it. | | I think you just might have hit on something there Ray. | | IMO, | The need for reading it all into memory and using that data from memory | instead of from disk as needed has now passed. | | Again IMO, | The only reason iot needed to be done that way in the past was... slower | machines/slower HDDs that were being used in the past. | | Now trhat we have decided to make the min req. machine a 386..... | We don't need to worry about that anymore. | | Reading any variable we need from arachne.cfg as needed/when needed | will now be just as fast with 386+/HDD as it was with 8086/memory. | | And let's face it... very few users are even going to be using 386s | anymore. Probably not very many 486s. | | Heck... I just had yet another P1-100MHZ given to me for free 2 weeks ago. | | Within the last month, I also had two P2-500MHZ machines given to me. | | It won't be much longer and the machines be given away for free | will P3s and P4s at 1GHZ to 2GHZ | | My P1-180MHZ machine is sitting right here next to me collecting dust. | | That's because this P4-1.7GHZ machine I'm using right now cost | just 238 bucks to put together 'piece by piece' from Ebay a year ago. | | ============================================================================ | | Fri, Oct 08, 2004 - 6:21:00 pm | | ray> | I've had similar thoughts since last night, but there could now be | an added reason to procced as you say that won't impact on speed | at all (well, it will impact but positively :-). It looks to me | like all the XSWAP stuff is basicaly XCH's way of creating a sort of | personal ramdisk for Arachne, but why bother when we have real | ramdisks? | | And if my method of reloading the structure values from an | image file is sucessful (it works now) then all the reloads | of Arachne suffer no delay due to the need to re-read ARACHNE.CFG, | so it hardly matters how long it takes to read the file the first | time since that need only happen once. But either way we're barking | up the same tree -- there's no need to load ARACHNE.CFG into | memory. BTW, I'm writing this with 183,184 bytes of free memory :-) | | | | | ============================================================================ | | Fri, Oct 08, 2004 - 9:24:42 pm | | ray> | Make that 183,216 free :-) | | | | ============================================================================ | | Fri, Oct 08, 2004 - 9:38:07 pm | | glennmcc> | Sheeeeittt... you got me beat by a full 7k :)) | | ============================================================================ | | Fri, Oct 08, 2004 - 9:43:03 pm | | glennmcc> | Dos memory(KB) 176[+] green | | ============================================================================ | | Fri, Oct 08, 2004 - 9:46:03 pm | | glennmcc> | No... wait a sec. | | 183,216 /1024 = 179k | | You only got me beat by 1k at most. | | 178k * 1024 = 182,272 | | You ain't even got 1,000 bytes on me. | | BBL | | | ============================================================================ | | Fri, Oct 08, 2004 - 10:00:20 pm | | glennmcc> | There, I just recompiled after removing the '>>10' | | She was doing some kind of strange rounding to KBs before :( | | I have 180,752[+] green | | So you got 2,463 bytes on me. | | ============================================================================ | | Fri, Oct 08, 2004 - 10:33:00 pm | | glennmcc> | Disregard the part about the strange rounding. | | | My old figure should have read........ | | 176 * 1024 = 180,224 | | The old one was actually 180752 /1024 = 176.5k | | So it was simply dropping the .5 to make it 176k | | Nuttin too strange about that ;-) | | ============================================================================ | | Fri, Oct 08, 2004 - 10:43:23 pm | | glennmcc> | Oh yeah, | | I also changed the next line in guitick.c to read <50000l instead of <50 | (tested it by loading my CD drivers low and then loading epppd low also) | | 85200[+] green before dialing | | 42368[!!!] red after dialing | | 36752[!!!] red while D/Ling this page | | Back up to 42368[!!!] red right now. | | Also changed the text it prints to read | | Dos mem (bytes) instead of Dos memory (KB) | | | ============================================================================ | | Sat, Oct 09, 2004 - 10:36:17 am | | ray> | Yup, why round the info. off when there's room for the full figure? | I have it indicated as "Free low memory" but Michal says that the | figure isn't actualy that, although it matches my 'mem /c' free low | memory almost exactly. Oh and I'm not bothering with the '[+]' ... | what does it realy tell us? | | | ============================================================================ | | Sat, Oct 09, 2004 - 3:04:00 pm | | glennmcc> | --- clipped from guitick.c --- | //DOS memory | value=farcoreleft();//>>10; | sprintf(str,"%4lu",value); | if(value<50000l || memory_model==0) | { | color=2; | strcat(str,"[!!!]"); | } | else | if(memory_model==2) | { | color=3; | strcat(str,"[+]"); | } | else | { | color=0; | strcat(str,"[?]"); | } | MemInfoLine("Dos mem (bytes)",str,color,y | ______________________________ | | <50kb of memory available ------------------------> [!!!] red | | >50kb of memory available and memory_model ==2 ---> [+] green | | >50kb of memory available but memory_model !=2 ---> [?] black | | =============================================================== | | And this is what memory_model is refering to. | | --- clipped from init.c --- | { | //2*2000: | memory_model=(char)((farcoreleft()-MIN_MEMORY)/IE_MAXSWAPLEN); | if(memory_model>2) | memory_model=2; | } | | if(memory_model) | { | BUF=4096; //!!JdS 2004/2/8: Increased from 4000 to 4096 | //BUF=4000; | x_fnt_alloc(SMALL_FONT_BUFFER); | //we will temporarily need some more 21000 kB... | } | else | { | BUF=2048; //!!JdS 2004/2/8: Increased from 2000 to 2048 | //BUF=2000; | x_fnt_alloc(BIG_FONT_BUFFER); | } | | if(memory_model==2) | { | if(ie_optimize()!=1) | memory_model=1; | } | | ___________________________ | | ============================================================================ | | Sat, Oct 09, 2004 - 4:24:14 pm | | ray> | I don't get your point, unless it's that we only get the '[+]' over | a given amount of memory, but I think the change in color is | enough to indicate trouble when we're below 50K ... besides I don't | think Arachne should even try to run with that little memory ... I'd | have her quit with a message to the effect of making more memory | available. | | BTW, Michal did solve my problem with pointer strings, so I can | now run Arachne with no ARACHNE.CFG at all, like I'm doing right now | (got it renamed to ARACHNE.SAV and I'm running off the structure image | saved in CONFIG.DAT). I'm thinking this project is going to | be a keeper. | | | | ============================================================================ | | Sat, Oct 09, 2004 - 5:50:18 pm | | glennmcc> | Yes, The color change is enough .... unless you are in grayscale mode. ;-) | | That was the reason Michael also used [!!!] [+] and [?] | | As for at what point we don't run at-all..... | | IIRC, MIN_MEMORY determines that. | | | ============================================================================ | | Sat, Oct 09, 2004 - 6:01:29 pm | | ray> | Yup, one must always beware the unconcious asumption that one's | environment is THE environment, but I did think of that, which is | why I've kept the '!!!' for insuficient memory. Come to think of | it, a 'beep' might be an added notice. | | ... 183,312 free ... we ARE going to get to 200K free by Xmas. | | | | ============================================================================ | | Sat, Oct 09, 2004 - 7:27:09 pm | | glennmcc> | How much we have free while in Arachne is of course dependant upon how | much we have free before starting her. | | IMO, | The more important figure is 'how much she is using'. | | BRB | | | | ============================================================================ | | Sat, Oct 09, 2004 - 7:29:57 pm | | glennmcc> | Conventional_memory: | _Name__Size_in_Decimal__Size_in_Hex | ______DOS_________91,552_(________89K_)__________165A0 | _QEMM386$____________768_(_________0K_)____________300 | __COMMAND__________1,936_(_________1K_)____________790 | _____CORE________463,232_(_______452K_)__________71180 | __command__________7,200_(_________7K_)___________1C20 | _____FREE________172,224_(_______168K_)__________2A0C0 | _Memory_Type__Total_Bytes_(_Kbytes__)__Available_For_Programs | _Extended______________67,043,328_(_65,472K_)_______________0_(______0K_) | _EMS__________________267,517,952_(_261,248K_)_____157,745,152_(_154,048K_) | _Largest_executable_program:__172,208_(_168K_) | _Total_Free_DOS_memory:_______250,160_(_244K_) | | ============================================================================ | | Sat, Oct 09, 2004 - 7:31:52 pm | | glennmcc> | So.... forget about getting that free memory figure 'up to' 200k | | Can we get core.exe usage 'down to' 400k ? | | ============================================================================ | | Sat, Oct 09, 2004 - 8:01:29 pm | | ray> | Quite correct! The free memory figure is only realy useful for | 'internal comparisions' i.e. to guage one's progress given an | unchanging personal computer configuration -- the 'real' figure is | as you quote it. | | Of course geting that number is a wee pain for me, being a 4dos guy | 'cause I've gotta restart under that kludge COMMAND.COM and | tweek MIME.CFG to get the number, which is ... | | | ============================================================================ | | Sat, Oct 09, 2004 - 8:07:42 pm | | ray> | ... 456,448 | | ... and I'll wager we can get it down to 400k, but not before | Xmas. | | | ============================================================================ | | Sun, Oct 10, 2004 - 8:19:56 am | | ray> | Joe, | | I noticed that you have IE_MAXWAPLEN tweeked right tight. What is | in there, and how can I take a peek at it? The reason I ask is | that right now I'm looking at "FreexSwap (KB) 1727" and | "UsedxSwap(KB) 255" ... it looks to me like these numbers have | nothing to do with your value of '30700u' because they don't add up | and becuase it seems that I've got xSwap to burn whereas I know that | if I reduce IE_MAXWAPLEN even a little I get a crash. Are these | numbers infact unrelated perhaps? | | | ============================================================================ | | Sun, Oct 10, 2004 - 5:49:32 pm | | glennmcc> | command.com _can_ handle this. | | file/memc.dgi >HTM|[10]mem /C|$esystem\\dgi\\fixmem.com /C>$2 | | 4dos.com _can't_. | | Now, which of them is thae 'kludge' ???? | | | ============================================================================ | | Mon, Oct 11, 2004 - 3:28:50 am | | Joe> | Hi folks, | | A few things covered here, but I'll be brief ... | | xSwap is not really a disk-cache, rather it is an interface into | the XMS world (or EMS or (eeek!) even Disk), thereby breaking the | 640K barrier without going POSIX. In other words, it's sort of a | "DOS extender". | | Arachhne is too disk-intensive NOW, that's one of the reasons it's | relatively slow. We should not make changes that will make it even | more disk-intensive. Try running Arachne without a disk cache and | you'll see what I mean, it's one of the most disk-intensive programs | I have ever used. | | The black/red/green stuff is very significant when comparing free | memory. With green, Arachne allocates two xSwap buffers, otherwise | it only allocates one. Therefore, the same number in green or in | black represents two different situations. | | Arachne reports how much memory is free. This may not be the best | number for developers, but it IS the best number for users, who | need to know if they have adequate memory and don't really care | how much Arachne takes, as long as they have enough for it. | | Yes, I did tighten some xSwap numbers - Arachne has become a bit | bloated over the years, so this was the start of reversing that | trend, since inexperienced users have difficulty achieving sufficient | free memory. The total xSwap space in XMS is (in theory) 64*30700, | so this should correspond to your FreexSwap+UsedxSwap numbers. | | ============================================================================ | | Mon, Oct 11, 2004 - 9:12:33 am | | ray> | Glenn, | | So far, that's the one and only defect I've found in 4DOS and it | remains to be seen if even *is* in 4DOS -- maybe there's something | in the Arachne code that makes some unwarandted assumption about | something -- we'll see. Regardless, apart from this problem, 4DOS | is as superior to C.COM as an AK47 is to a flintlock (for it's | indended purpose). | | | ============================================================================ | | Mon, Oct 11, 2004 - 9:19:52 am | | ray> | Joe, | | Thanks for that -- informative enough to be printed off and saved :-) | This sort of overall understanding of how the program works is | esential to any constructive systemic changes. | | Of course you are right about Arachne's disk use and that is part of | the motivation behind my little project. Even on my '800 the | improvement in speed bypassing ARACHNE.CFG is noticable and nice. | | | | ============================================================================ | | Mon, Oct 11, 2004 - 5:06:41 pm | URL | | glennmcc> | Just left this post over on the Yehaw arachnebugs group site........ | | | | Houston, We have a problem. | | With & amp; | | http://www.cisnet.com/glennmcc/testing_&_symbol/&-amp-bug.htm | | | ============================================================================ | | Mon, Oct 11, 2004 - 5:28:10 pm | | glennmcc> | I am almost 100% sure that the bug and the fix for it are somewhere in URLST | | I looked for it for about 1hr last evening but did not find it. | | | ============================================================================ | | Mon, Oct 11, 2004 - 5:48:52 pm | | ray> | That's your department Glenn :-) | | | | ============================================================================ | | Mon, Oct 11, 2004 - 7:10:05 pm | | glennmcc> | No problem.... just call me "bug hunter". ;-) | | ============================================================================ | | Mon, Oct 11, 2004 - 8:05:50 pm | | ray> | What? have you found it already? | | | ============================================================================ | | Tue, Oct 12, 2004 - 6:57:54 am | | phantom> | Well ... finding a bug is half-way towards squashing it, right? | | | | ============================================================================ | | Tue, Oct 12, 2004 - 7:37:50 am | | ray> | Na, 3/4 of the way! | | Joe, ruminating on your Xplanation of xswap -- so there is nothing | 'in there' as such, IE_MAXSWAPLEN is just a buffer for swapping | stuff in and out of XMS (or EMS or disk), right? If that's the case | then what does xswap do that a disk cache doesn't do? It looks to | me like xswap is Michael's re-invention of something like EMM. I | understand that there is more overhead in theory in storing stuff in | a ramdisk or in a disk cache, OTOH that functionality comes 'free' with | the existing software, whereas xswap might be more efficient in | theory but it all comes at the cost of the code to do it which is | over and above the ramdisks and (God save us) Smartdrive that almost | all of us use anyway. I'm wondering if Arachne stored all it's | swap stuff in a ramdisk, and if xswap was eliminated, would things | really be slower? I'm inclined to doubt it, given the noticable | improvement in speed I'm already noticing with CONFIG.DAT. | All speculation at this point. | | | ============================================================================ | | Tue, Oct 12, 2004 - 4:21:39 pm | | glennmcc> | Nope, did not find it yet. | | That's why I'm still hunting. ;-) | | ============================================================================ | | Tue, Oct 12, 2004 - 5:39:46 pm | | Joe> | It's rabbit season! ;-) | | Ray, | | I wonder what proportion of Arachne users use a RAM disk - I know I | don't, my upper memory is pretty busy as is. Anyway, you can check | your performance theory quite easily ... remember xSwap supports | using Disk in lieu of XMS, just use "arachne -s" to configure for | this mode and see how it goes ... | | | ============================================================================ | | Tue, Oct 12, 2004 - 6:59:39 pm | | glennmcc> | IMO, that's not a good test. | | Xswap will still get used.... it's just that it will be swapping to | and from a RamDrive instead of to and from XMS or EMS | | If I understand correctly what Ray was getting at, he wants to eliminate | Xswap all-together. | | | ============================================================================ | | Tue, Oct 12, 2004 - 7:02:13 pm | | glennmcc> | As for what proportion use a RamDrive. | | I'd bet on it that it's the same proportion who've read this file. | | arachne_main_dir\doc\speed.htm | | | | ============================================================================ | | Tue, Oct 12, 2004 - 7:04:33 pm | | glennmcc> | Joe, | | I posted a reply to....... | | http://groups.yahoo.com/group/ArachneBugs/message/7 | | That reply _should_ show-up as........ | | http://groups.yahoo.com/group/ArachneBugs/message/8 | | But......... | | Message 8 does not exist in ArachneBugs | | | ============================================================================ | | Tue, Oct 12, 2004 - 8:04:50 pm | | ray> | Joe, Glenn, | | Thanks for the input. Joe has a point, but Glenn has a sharper one | since, as he says, the goal would be to use a ramdisk to *replace* | xSwap, and we can't test that directly. Besides, if eliminating xswap | were to turn out to be a worthwhile goal I can't see any reason why | we couldn't say that Arachne requires a ramdisk just as she requires | a dialer and a packet driver. Why not use existing resources rather | than reinvent the wheel? The question that should be asked is: | What does xswap do that couldn't be done just as well by other means | and with one hell of a code simplification? | | | | ============================================================================ | | Wed, Oct 13, 2004 - 1:03:44 am | | Joe> | Glenn, | | Because ArachneBugs can accept messages from non-members, the | "Reply To" field for messages is not ArachneBugs, but rather | the originator of the message. In other words, the ArachneBugs | list does not behave as lists normally do in this regard, to | send a reply back to the list, you have to manually edit the | "Reply To" field to 'ArachneBugs'. | | I hope that made sense, and certainly via e-mail, that's how | it works. I'm not 100% sure what happens when replying via | the web interface, since the message originator's e-mail | address might be suppressed. | | | ============================================================================ | | Wed, Oct 13, 2004 - 1:13:48 am | | Joe> | Ray, Glenn. | | My first reaction to reading Ray's idea of replacing xSwap with | RAM Disk was - Do you have any idea of the amount of work that | would be involved?! There are a couple of thousand different data | items to keep track of, which would be unworkable if each had its | own file, I doubt any RAM Disk and/or Disk Cache and possibly DOS | would be able to handle it, let alone provide decent performance. | | Then I realized that xSwap already has the "smarts" to consolidate | all the data items into a smaller number of blocks, and further, | has the option to use Disk to handle them. In other words, this | IS a valid test, since I don't think the alternative is feasible | anyway. | | | | ============================================================================ | | Wed, Oct 13, 2004 - 7:21:43 am | | ray> | Joe, | | "A couple of thousand different data items", you say ... well, what | is in there? At present the only things I know are in xswap are | the three .cfg files and ARACHNE.PCK which makes four. I've already | almost eliminated the need to keep ARACHNE.CFG in memory and it | wasn't that hard. I'd love to do the same with TOOLBAR.CFG since | I'm quite sure that's the biggest speed bottleneck, so maybe it | wouldn't be that tough. Anyway, I won't contemplate actualy *doing* | anything to xswap for a long time, I'm just curious as to it's | nature. | | | ============================================================================ | | Wed, Oct 13, 2004 - 4:00:56 pm | | glennmcc> | Joe, | | Yep, that is what I would have expected IF I had replied via email | | Since I read the message and replied to it via the Yahoo web interface. | | I expected that it would have done exactly as it had for my previous | posts and replies via that same web interface. | | And that is, that within a few seconds there was one more message | in there. | | BRB | | ============================================================================ | | Wed, Oct 13, 2004 - 4:08:39 pm | | glennmcc> | Just now posted this peply........ | | | 2nd try to post a reply. | | About 10 min ago. ;-) | | | ---------- | | My 1st reply consisted only of that 2nd line. ;-) | | Which was of-coarse in answer to this question from you in post#7........ | | ..... So (you can see where this is going, right?;-), do you have | an ETA yet on releasing the 1.77 sources? | | | ----------- | | I just checked for the 4th time and it still di not 'take'. | | Message 8 does not exist in ArachneBugs | | | Even tho this is what it said after hitting the 'send' button. | | Your message has been posted. | | | | ============================================================================ | | Wed, Oct 13, 2004 - 4:12:41 pm | URL | | glennmcc> | See ? | | ============================================================================ | | Thu, Oct 14, 2004 - 3:48:13 am | | Joe> | Glenn, | | Just tried the "Reply" option at ArachneBugs, selecting ArachneBugs | as the "To" address - and it worked (as you can see). | | Also tried selecting myself as the "To" address (being the message | originator) but it didn't seem to work. Lastly, tried the same with | you as the "(Reply) To" address (you being the message originator in | this case), so you'll have to tell me it that worked ... | | ============================================================================ | | Thu, Oct 14, 2004 - 3:53:31 am | | Joe> | Ray, | | Arachne stores just about everything in xSwap (or so it seems to me;-). | If you look at the "Free xSwap" and "Used xSwap" in the Arachne GUI, | you'll see there's a lot more KB in use than just the CFG stuff, even | taking into account quantization/wastage effects. You can also use | this to see how much xSwap you save in B5 by not loading 'arachne.cfg' | into xSwap. | | | ============================================================================ | | Thu, Oct 14, 2004 - 7:28:03 am | | ray> | Hmmmm, yes. And if I take your suggestion about using a ramdisk as | the xSwap area, then I can spy on the contents easily enough. | | ============================================================================ | | Thu, Oct 14, 2004 - 9:36:09 am | | glennmcc> | Joe, | | There are now 8 messages. | | Your reply to message 7 is message 8 | | None of the others are there. | | ============================================================================ | | Thu, Oct 14, 2004 - 9:38:43 am | | glennmcc> | Can't say what happened to the email version of them because I have my | subscription setup to not receive any emails. | | I do all of my reading and replying via the web interface. | | | ============================================================================ | | Thu, Oct 14, 2004 - 1:53:51 pm | | glennmcc> | Ok, I got your message in my Yahoo inbox. | | Replied to you and sent a CC to Arachnebugs. | | That CC is now message 9 | | So, it _is_ taking them when sent via email..... but not from the web | interface in the group itself. | (even tho it says that the massage was posted.... it wasn't) | | ============================================================================ | | Thu, Oct 14, 2004 - 4:43:42 pm | | ray> | Joe, | | No luck finding the xWsap file on disk. It's supposed to end with | ._$B but there seems to be no such file created anywhere on the | machine. Oh and it is much slower. | | ============================================================================ | | Thu, Oct 14, 2004 - 5:35:36 pm | | glennmcc> | Yes, the same result here. | | Much much slower. | | But... my Dos mem went from 180672 to 185800 | | To find the files you'll need a file undelete utility. | | They will be found in %temp%\arachne.tmp | | Here's what I was able to undelete. | | | Volume in drive L is MS-RAMDRIVE | Directory of L:\TEMP\ARACHNE.TMP | | . 10-14-04 7:22p | .. 10-14-04 7:22p | DONOTPAN IC! 36 10-14-04 7:22p | 627 _$B 19659 10-14-04 7:22p | 633 _$B 32280 10-14-04 7:22p | 2 _$B 32065 10-14-04 7:22p | 635 _$B 32354 10-14-04 7:22p | 4 _$B 32065 10-14-04 7:22p | 43 _$B 32065 10-14-04 7:22p | 10 _$B 32065 10-14-04 7:22p | 543 _$B 32065 10-14-04 7:22p | 47 _$B 32065 10-14-04 7:22p | 634 _$B 32065 10-14-04 7:22p | 547 _$B 32065 10-14-04 7:22p | 550 _$B 28950 10-14-04 7:22p | 551 _$B 18433 10-14-04 7:22p | 552 _$B 18433 10-14-04 7:22p | 553 _$B 18433 10-14-04 7:22p | 554 _$B 18433 10-14-04 7:22p | 104 _$B 18433 10-14-04 7:22p | 105 _$B 18433 10-14-04 7:22p | 151 _$B 3033 10-14-04 7:22p | 153 _$B 18433 10-14-04 7:22p | 154 _$B 18433 10-14-04 7:22p | 155 _$B 18433 10-14-04 7:22p | 220 _$B 18049 10-14-04 7:22p | $ROURA$ BAT 79 10-14-04 7:22p | 27 File(s) 97736704 bytes free | | | | ============================================================================ | | Thu, Oct 14, 2004 - 5:37:54 pm | | glennmcc> | That was after just 30 seconds of operation using disk instead of XMS | | I'll try another test for several minutes to see what I get. | | ============================================================================ | | Thu, Oct 14, 2004 - 6:03:44 pm | | ray> | Glenn, | | Thanks for that. I had the same level of increase in my free mem, too. | It looks like figuring out what's in these files won't be | straightforward. As to the drop in performance, as you pointed out, | using disk involves double jeapordy, since the disk opperations and | the xSwap opperations are both being performed which would not be | the case if xSwap were bypassed. | | | | ============================================================================ | | Thu, Oct 14, 2004 - 6:04:36 pm | | glennmcc> | Very strange. | | When I first start Arachne, 152800 | | Dial/connect Arachne restarts.... 152800 | | Visit several pages.... 152800 | | Check cache index ... 152800 | | Check mem stats ....... she just jumped to 184976 | | Check more details with memc.dgi ...... 184976 | | Still at 184976 right now | | Conventional_memory: | | ____ Name________ Size in Decimal_________ Size in Hex | _____ DOS________ 91,552_(_______ 89K )_________ 165A0 | _QEMM386$___________ 768_(________ 0K )___________ 300 | _ COMMAND_________ 1,936_(________ 1K )___________ 790 | ____ CORE_______ 459,136_(______ 448K )_________ 70180 | _ command_________ 7,200_(________ 7K )__________ 1C20 | ____ FREE_______ 176,320_(______ 172K )_________ 2B0C0 | this section----> +7,200_<-- command.com | typed manually__ 183,520 | _______________ -184,976 | discrepency of -> -1,456 | | 1,456 bytes additional usage by core.exe during this opperation ??? | | | | ============================================================================ | | Thu, Oct 14, 2004 - 6:04:55 pm | | ray> | ... | | Hey! you're beating me in the free memory departement again :-( | | | ============================================================================ | | Thu, Oct 14, 2004 - 6:05:56 pm | | glennmcc> | OK, I have now been using RamDisk instead of XMS for 26min | | BRB | | | ============================================================================ | | Thu, Oct 14, 2004 - 6:10:59 pm | | glennmcc> | Damn it, I'll have to try again. | | I just had a 100% lock-up. Everything was completely frozen. No key | combo would unlock it. Had to do a 'cold reboot'. :(( | | | | | ============================================================================ | | Thu, Oct 14, 2004 - 6:11:15 pm | | ray> | Ah, no I'm still ahead: 456,448 ... catch me if you can :-) | | | ============================================================================ | | Thu, Oct 14, 2004 - 6:29:20 pm | | glennmcc> | OK, got it this time. | | I hit 'Q' while on my main hotlist. | | Allowed her to laod 100 different pages from all over the world. | | Simple text only pages, pages loaded with tables, framed pages, FTP | listings.... etc ....etc ....etc.... | | Here's what I undeleted....... | | | Volume in drive L is MS-RAMDRIVE | Directory of L:\TEMP\ARACHNE.TMP | | . 10-14-04 8:12p | .. 10-14-04 8:12p | DONOTPAN IC! 36 10-14-04 8:13p | 10820 _$B 32749 10-14-04 8:23p | 10640 _$B 32748 10-14-04 8:22p | 3 _$B 32065 10-14-04 8:13p | 4 _$B 18433 10-14-04 8:13p | 5 _$B 18433 10-14-04 8:13p | 6 _$B 18433 10-14-04 8:13p | 1961 _$B 32354 10-14-04 8:15p | 9 _$B 32065 10-14-04 8:13p | 51 _$B 32065 10-14-04 8:13p | 15 _$B 32065 10-14-04 8:13p | 122 _$B 32065 10-14-04 8:13p | 81 _$B 32065 10-14-04 8:13p | 153 _$B 32065 10-14-04 8:13p | 1084 _$B 32065 10-14-04 8:14p | 412 _$B 32065 10-14-04 8:13p | 490 _$B 32065 10-14-04 8:13p | 178 _$B 18433 10-14-04 8:13p | 1101 _$B 32065 10-14-04 8:14p | 1121 _$B 32065 10-14-04 8:14p | $TMP$ ASF 5918 10-14-04 8:13p | 1366 _$B 32065 10-14-04 8:14p | 1206 _$B 32065 10-14-04 8:14p | 7390 _$B 32749 10-14-04 8:19p | $ROURA$ BAT 176 10-14-04 8:12p | 10639 _$B 32179 10-14-04 8:22p | 1955 _$B 18133 10-14-04 8:15p | 10876 _$B 8717 10-14-04 8:23p | 1958 _$B 18133 10-14-04 8:15p | 10823 _$B 32747 10-14-04 8:23p | 10824 _$B 5003 10-14-04 8:23p | 10825 _$B 6283 10-14-04 8:23p | 10341 _$B 5531 10-14-04 8:22p | 8159 _$B 32732 10-14-04 8:20p | 8172 _$B 32748 10-14-04 8:20p | 8175 _$B 31826 10-14-04 8:20p | 8174 _$B 3596 10-14-04 8:20p | 7157 _$B 29109 10-14-04 8:19p | 7198 _$B 32583 10-14-04 8:19p | 7199 _$B 18793 10-14-04 8:19p | 7180 _$B 9873 10-14-04 8:19p | 43 File(s) 96200704 bytes free | | | ============================================================================ | | Thu, Oct 14, 2004 - 9:56:06 pm | | ray> | ... so it's cache stuff?! | | | | | ============================================================================ | | Fri, Oct 15, 2004 - 7:23:07 am | | phantom> | Well, | | The (default) Arachne cache is actually c:\arachne\cache | | As I see it, the xSwap stuff is simply all the data structures, | especially strings, that could no longer fit in conventional | memory. I'll see if I can dig up MP's description of xSwap | from when he introduced it, if that'll help any. | | As for using Disk services to substitute for xSwap, sure, that's | "lateral thinking", but I think the DOS overhead for such services | is excessive. In other words, the xSwap functions will be more | efficient anyhow, even if a RAM disk is involved. Just MHO, FWIW. | | However, xSwap is complicated and perhaps unstable, so if you're | really interested in avoiding it, start looking seriously at | making a POSIX port of Arachne (with DJGPP and GCC). | | | ============================================================================ | | Fri, Oct 15, 2004 - 7:34:08 am | | ray> | Thanks Joe, if you could find more info on this question, that would | be just ducky ... but I'm not sure that xSwap is data structures, | for example, my friend (now 'sUI' formerly 'UserInterface') is not | stored anywhere fancy and if anything was to be given this sort of | treatment, sUI would be as well, I'd say. I think all structures | are in _DS and the .cfg files are in xSwap along with God know what | else. Anyway the whole thrust of my current experiments revolves | around the fact that the whole of ARACHNE.CFG can be stored in | an image file that's only 1K long and this seems like a nice improvement | expecialy if the same trick will work with the other .cfg files. | | | ============================================================================ | | Fri, Oct 15, 2004 - 7:35:32 am | | ray> | ... Oh, and as for POSIX, Glenn ignores all questions about it, so | what does that tell us? | | | | | ============================================================================ | | Fri, Oct 15, 2004 - 4:35:05 pm | | glennmcc> | Oh.... but I do not 'ignore' them. | | I _do_ listen to them and pass all of them right along to "Mr. X' as you | decided to refer to the person who's working of a DJGPP port of core.exe | | As you will recall, this was my response some months ago when asked what | my feelings were on POSIX. | --- now repeating the response --- | http://www.cisnet.com/glennmcc/images/posix.jpg | Nuf said ? | ----------------- | | | Also..... if you will recall. | | I asked specifically for everything concerning posix to be left as-is | in the src code files. | | I only changed that position when Bernie said that it most likley did | not work and would not be needed for our particular purposes. | | As Joe has pointed-out.... the stuff you are now starting to work with | just might need that posix stuff. | | So looking back, was I correct in my original request that it left alone ? | | Or am I still full of **it ? | | ============================================================================ | | Fri, Oct 15, 2004 - 5:17:05 pm | URL | | glennmcc> | Ah ha, found it.... There's where it was. | | ============================================================================ | | Fri, Oct 15, 2004 - 6:23:16 pm | | ray> | Glenn, | | You continue to see these questions as some sort of contest. | Life would be simpler if you'd stop. My *guess* is that almost | all of that old code will end up scrapped, but I could be wrong. | If I am right that doesn't mean you were full of **it, and if | I'm wrong that doesn't mean I am. I've gone to great lengths to | preserve the POSIX stuff up to B3c precicely because I recognise | that it's preservation is prudent. It would be polite of you to | keep us up to date on what X is up to since we are all very interested | in that project and hope it suceeds. | | | | ============================================================================ | | Fri, Oct 15, 2004 - 7:30:42 pm | | glennmcc> | Sorry. I should have used capital Gs | | So that you would have known for sure that I just joking around. | | | | ;-) | | ============================================================================ | | Fri, Oct 15, 2004 - 7:49:08 pm | | ray> | Ok we're cool then, glad to hear it. | I am in an ironic situation with this POSIX stuff and the DJGPP port: | I *want* to be wrong :-/ | | | | ============================================================================ | | Fri, Oct 15, 2004 - 7:51:52 pm | | ray> | OH, but now that the matter is raised, since you *aren't* ignoring | our qestions about project X, how is he doing? -- we haven't had | a progess report in some time. | | Given that there is a 32 bit port of WatTCP, it looks to me like | the only obscacle is the graphics stuff. | | | | ============================================================================ | | Fri, Oct 15, 2004 - 8:26:42 pm | | glennmcc> | Yep, you are right on that assumption. | | I have not heard an update on the progress either since about 6weeks or | so ago. | | The last I heard.... hixlov.lib _was_ the 'stumbling block'. | (or more precisely... converting the SRCs for it to DJGPP 32bit) | | | ============================================================================ | | Sat, Oct 16, 2004 - 8:13:22 am | | phantom> | I might be wrong, but my guess is that "Mr X" is the other Glenn. | (Actually, I don't know why the "cloak and dagger" mystery with this | stuff, it's hardly the most productive of sorting out this stuff, but | no matter, I for one have plenty to keep me occupied anyway ...) | | Ray, I haven't located MP's announcement about xSwap as yet, however | if you look in your Arachne directory, you should find a file called | 'xswap.txt' that gives a pretty good overview of xSwap. | | | ============================================================================ | | Sat, Oct 16, 2004 - 3:26:56 pm | | glennmcc> | Look who's talking about "cloak and dagger" | | Who is this "phantom" guy anyways ??? | | But seriously now. It's not Wizzy. | | In-fact, this email is my first-ever contact from this person. | | -------------- | Date: Mon, 10 May 2004 15:28:19 -0700 (PDT) | From: "Mr. X" | To: Glenn McCorkle | Subject: Re: 6 months ago.......... | | Glenn, your new Arachne build works very well! I am planning to put | some effort into a DJGPP port, but I don't know how far I'll get. | Can you make your source code available to me? | ------------------ | | "Mr. X" (as Ray has dubbed this person), visits here and reads | the posts via web interface which we have made to the Yahoo group. | | If they want to reveal their identity, they will. | | If not.... that is their choice. | | | ============================================================================ | | Sun, Oct 17, 2004 - 9:29:36 am | | ray> | Joe, thanks for 'xwap.txt' ... never noticed it. | | Glenn, | | Sure, X has the right to anonyminity, but what purpose does it | serve? ... of course that's for X to decide ... anyway, I hope | he/she drops the mask, since I know we'd all like to thank that | person directly for their efforts and maybe even help. Right now | Michal and I are wrestling with converting some far pointers to | near pointers and you can't believe how much trouble it is -- typecasts | till the Devil himself would give you a break. Did you know that | (char near *) NULL != (char far *) NULL, and that | | 'if (somepointer)' tests that work with far pointers are always true | with near pointers? And on and on ... give us flat memory X! | | | | ============================================================================ | | Sun, Oct 17, 2004 - 1:43:01 pm | | glennmcc> | That gives us 'flat memory' ??????? | | You're pulling my leg .... aren't you ? | | If you're not.......... that is some of the most FANTASIC news I have | ever heard in my entire "computing life". :))) | | ============================================================================ | | Sun, Oct 17, 2004 - 3:59:09 pm | | ray> | Hmmm? | | I ain't pulling your leg, all that $hit is true, Michal is helping | make sense of it; this isn't flat memory, it's the opposite (segmented | 16 bit architecture), which is why I'm hoping that X can | *give* us flat memory (DOS32) where NULL == NULL == NULL. | | | | | ============================================================================ | | Sun, Oct 17, 2004 - 5:19:08 pm | | glennmcc> | OK, now I see what you mean. | | Not 'flat memory', but the next best thing to it. | | So I'll repeat........... FANTASTIC !!!!!!!!!! | | ============================================================================ | | Sun, Oct 17, 2004 - 7:23:56 pm | | ray> | Got 'memc.dgi' reporting under 4DOS :-) | I"m hot on the trail of the bug. | | | ============================================================================ | | Sun, Oct 17, 2004 - 9:05:00 pm | | mht> | DOS32 does not necessarily mean flat memory model. Flat is just | something like 32-bit Tiny, but without its size limitations, and | has advantages and disadvantages. You can use segmented 32-bit | memory models as well (probably not with DJGPP, but this is | a portability issue). | | | | ============================================================================ | | Mon, Oct 18, 2004 - 3:37:26 am | | phantom> | | glennmcc> | Look who's talking about "cloak and dagger" | | Who is this "phantom" guy anyways ??? | | | Dunno ... Ask "Little Birdie" ... | | | ============================================================================ | | Mon, Oct 18, 2004 - 7:41:41 am | | ray> | Touche | | | | ============================================================================ | | Mon, Oct 18, 2004 - 4:56:27 pm | | LittleBirdie> | Hahahahahahahaha !!!!! | | Touche is right. ;-) | | | ============================================================================ | | Mon, Oct 18, 2004 - 6:35:59 pm | | ray> | Looks like I've got the 4dos bug beat ... well, controled, anyway. | | | ============================================================================ | | Mon, Oct 18, 2004 - 7:20:44 pm | | glennmcc> | What was causing it ? | | Something to do with 'memory reporting' differences between | command.com and 4dos ??? | | ============================================================================ | | Mon, Oct 18, 2004 - 8:39:32 pm | | ray> | Nope, it turns out that there is no bug at all, it's just that | 4dos is to big to allow a sucessful 'system()' call. I'm not clear | on the details, but it seems 'system()' has to re-load the command | processor even though there isn't an obvious shell, and 4dos is too | big, which puzzles me, since it loads itself into XMS anyway. At any | rate 'system()' reports 'not enough memory' so maybe it tries to | load the whole thing low, with 4dos then moving itself into XMS. | In any case the solution is some vile cheating ;-) | | | | ============================================================================ | | Sat, Nov 06, 2004 - 5:30:04 pm | | ray> | Ah, the first sub 407' core: 406,992. We'll crack 400,000 by | Christmas. | | | ============================================================================ | | Sun, Nov 07, 2004 - 7:33:32 am | | phantom> | Well done Ray, | | I notice that the B3c+ project file has the "Supress Redundant Loads" | option set. Apart from that spot that Michal identified in 'putikonw' | (IIRC), where else in the code was this option causing trouble? | | Also, what is the "vile cheating" you've effected on 4DOS? | | | ============================================================================ | | Sun, Nov 07, 2004 - 8:34:22 am | | ray> | Joe, | | SRL seems to provide little improvement in the code, and Michal | doesn't like it, since it looks like it might cause trouble | in the future, so I've since turned it off. IIRC turning it on only | saved about 40 bytes. | | As to 4DOS, I've just created this shell game, where I run Arachne | under command.com transparently, i.e. I never see it, but Arachne | uses command.com for all her external program calls. If your interested | I'll send you the whole recipe. There are some interesting | anomalies in what's needed to make it work and I am again beginning | to think that there is some bug, not in Arachne or 4DOS, but in | 'system()', since it manages to foul up even command.com in | some situations. | | | | ============================================================================ | | Sun, Nov 07, 2004 - 8:39:55 am | | ray> | BTW, Joe, all: | | Does anyone know what's going on with the AltTab() function? | Why do we need a keyboard interupt capture? AFAICT there is no | TSR feature in Arachne, so why not just handle Alt-Tab like | any other keystroke. Also, this is a configuration option; why? | Would there ever be a reason not to have the Alt-Tab feature | enabled? The only thing that comes to mind is running under | Window$ ;-( | | | | | ============================================================================ | | Sun, Nov 07, 2004 - 12:47:48 pm | | ray> | Holy cow, I'm on line right now with 188,752 bytes of free memory! | But there is a bug. Nevermind, I shall slay it! | | | ============================================================================ | | Sun, Nov 07, 2004 - 12:58:47 pm | | ray> | The God's shall surely slay me for my hubris -- I am on line right | now with 197,168 bytes of free memory. I fly too close to the | sun and shall be cast into the sea. | | | | ============================================================================ | | Sun, Nov 07, 2004 - 1:27:25 pm | | ray> | Bug found. Core.exe loaded size = 443,136. | I need a drink. | | | ============================================================================ | | Sun, Nov 07, 2004 - 1:43:49 pm | | glennmcc> | My current CORE.EXE is using 463,232 bytes | | Yours is using 443,136 | | So, you gonna' share the code that gives you a savings of 20,096 bytes ? | | | ============================================================================ | | Sun, Nov 07, 2004 - 1:46:43 pm | | glennmcc> | BTW, | | This core.exe (v1.79), _does_ include the fixes that needed to be done | visa-vi Joe's Cookie code. | | v1.77 was screwed-up in that it did not include all of his code. | | | ============================================================================ | | Sun, Nov 07, 2004 - 1:51:54 pm | | glennmcc> | Have just now swiched-back to v1.77 | | This is core.exe v1.77 usage..... 463,232 | | As you see..... the remainder of Joe's code does not change usage by | even a single byte. | | Now _that_ is some good work Joe. :)) | | ============================================================================ | | Sun, Nov 07, 2004 - 5:35:58 pm | | ray> | Glenn, | | Naturaly, but there's better yet before it's released. | BTW glad to here you've got | all of Joe's stuff in your next version, but what about all | Michal's stuff -- the re-write of (h/m)drawbmp.c ? | | | ============================================================================ | | Sun, Nov 07, 2004 - 7:06:57 pm | | glennmcc> | Is the SRC for it available for D/L somewhere ? | | | ============================================================================ | | Sun, Nov 07, 2004 - 7:18:08 pm | | glennmcc> | Just back from looking through all of our file dirs at Yahoo. | | Nope.... no updated hdrawbmp.c to be found in any of the dirs | | ============================================================================ | | Sun, Nov 07, 2004 - 10:46:23 pm | | ray> | I'll have a look around my machine and send it to you, but you'll have | to mess it up, because it's in 3c format < g g g > | | | ============================================================================ | | Mon, Nov 08, 2004 - 12:17:48 am | | Joe> | Ray, | | Interesting about SRL, when I originally tried this option, it saved | something like 20K of memory, but made the code unstable, that's why | I was surprised to see it in the B3c project file ... | | I take it that "Arachne uses command.com for all her external program | calls" means you're using the normal DOS command processor rather than | the 4DOS command processor for calling external programs? What about | JPEG images in HTML pages (which use an external conversion program), | did this stuff also fail under 4DOS? | | Regarding your AltTab question, it's not something I've looked at, but | your guess about Windoze may be correct. If Windoze normally intercepts | this key sequence, then Arachne will never see it. So it might install | its own interrupt capture routine, to ensure that it does see it. Just | off-the-cuff speculation on my part here ... | | | | ============================================================================ | | Mon, Nov 08, 2004 - 4:26:45 pm | | ray> | Glenn, | | Come to think of it, I don't need to send you Michal's work, just | grab base3c from Yahoo, and look at 'mdrawbmp.c' and I expect | that you'll need 'compiler.h' as well IIRC. | | Joe, | | Hmmm, I'd better look at SRL again! | | As to 4DOS, that's right, I set 'comspec' to 'command.com' even | though I'm realy running under 4DOS (disgusting eh?) ... but it works. | 4DOS didn't fail to do anything, but it's so large that the | "[xxx]" tests in 'mime.cfg' had to be removed, else Arachne | tried (and failed) to execute the external commands without | quiting. I've fixed that too, so now if Arachne can't run a program | internaly, it will quit to DOS without missing a beat. | | | ============================================================================ | | Mon, Nov 08, 2004 - 7:19:28 pm | | glennmcc> | OK, | | THAT was a lesson in futility :(( | | Have been trying for 2hrs to compile a mixture of 3c and my current SRC. | | Forget it... not possible. :( | | With all of the changes you made (UPPERCASE, UPPERlower, split some stuff | off into seperate files, added some new files, etc, etc, etc)...... | | Your SRC is no longer compatible with my SRC. | | Our code has forked and we are f***ed. :( | | So, I compiled yours instead. | | 2 small points. | | 1) 3c still uses toolbar.tb instead of toolbar.cfg | 2) something is messaed-up in CSS | (pppglenn.htm shows the topbar correctly | but the midbar is gray instead of silver) | | ------- skin.css has both set at silver ----------- | .topbar | { | background-color:silver; | } | | .midbar | { | background-color:silver; | } | ___________________________________________________ | | screenshot....... http://www.cisnet.com/glennmcc/images/css.gif | | ============================================================================ | | Mon, Nov 08, 2004 - 7:26:32 pm | | glennmcc> | Forget that screenshot............ | | The BMP is FUBAR :(( | But, you already knew about that bug. | BMP header info in the wrong order... Right ? | | And IIRC, you already fixed it. | | Anyways.... I'll wait till your new SRC code is ready and try again. ;-) | | | ============================================================================ | | Mon, Nov 08, 2004 - 7:53:50 pm | | ray> | Glenn, | | Na, we're not f***ed, in fact, if I take it that you're ready tp | start working in the cleaned code, that's a cause for celebration, | since now I won't have to import your code across the verion gap | and we'll all be singing from the same page, more or less. | | If I have this right, I very much look forward to your critique of | '3c'. Roger the PrintScreen() foul up -- a fixed file is uploaded | to my folder at Yahoo. If there is a CSS problem, I await your | confirmation. | | Joe, | | Sorry man, I had my wire's crossed vis. SRL, that's been active for | some time and is OK. What I was thinking of was "assume no pointer | aliasing" -- that's off with good reason. | | | | I'll say we're just starting to move forward; if | I take it that you're going to start working in the cleaned code. | | | ============================================================================ | | Mon, Nov 08, 2004 - 7:54:47 pm | | glennmcc> | 1 more point. | | 3c has the same bug as mine. | | http://www.cisnet.com/glennmcc/testing_&_symbol/ | (try accessing the subdirs under it) | | Also try accessing the links inside of.... | http://www.cisnet.com/glennmcc/testing_&_symbol/&-amp-bug.htm | | ============================================================================ | | Mon, Nov 08, 2004 - 8:08:56 pm | | glennmcc> | Yep, you 'caught the drift'. | | You won't need to worry about doing any more importing. | | I'll import the changes I've made since v1.75 into 3c and see what happens. | | | | ============================================================================ | | Mon, Nov 08, 2004 - 8:16:21 pm | | glennmcc> | And yes, I look forward to seeing your current stuff. | | Especially whatever it is that saved 20,096 bytes of memory usage. :)) | | Since my current core.exe is showing | | Dos mem (bytes) 180,256 | | Yours would show 200,352 | | ============================================================================ | | Mon, Nov 08, 2004 - 9:58:31 pm | | ray> | Well, that is cause for a party, glad to have you on board Glenn, | now we can do some serious debuging. Bedtime now, but I'll get | to work on the above tommorow. For now, I'd say don't bother with | importing 1.77 stuff, that's already done in B5 which should be | ready for upload in provisional status in a day or two (I'm waiting | for something cool from Michal). Rodger the 'toolbar.tb' that's fixed | in B5 too. | | As for the 20K memory savings, I should make you guess how I did it | (or at least the 14K savings with one mornings work). I'll give | you a hint: Question -- what determines the size of the loaded .exe??? | | | ============================================================================ | | Tue, Nov 09, 2004 - 5:49:01 pm | | glennmcc> | Kewl, I'll wait for B5 to be ready. | | Then all that I'll need to import is the few little | things I've done since Oct 1st | | This is clipped from my current changes.txt | | ----- | 1.79;GPL,386+ stable for DOS | ---- | Release date: ??? ??, 2004 | | Oct 19, 2004 | Re-enabled HTTPSTUB which allows Arachne to use the Windows DUN | instead of a DOS packet driver. This makes it possible to run | Arachne on machines which have WinModems which simply refuse to | work with a DOS packet driver. | Thanks to Glenn Gilbreath (aka Wizzy), for making this suggestion. | | Oct. 26, 2004 | Fixed my mistakes from v1.77 in which I had not included all of | the new cookie code written by Joe de Silva. | | Joe, | Thank you very much for all of the help you gave me in fixing this. | (and for not biting my head off) | | ??? ??, 2004 | | ??? ??, 2004 | | ??? ??, 2004 | | ??? ??, 2004 | _________________________________________________________________ | | | ============================================================================ | | Tue, Nov 09, 2004 - 6:32:23 pm | | ray> | Glenn, | | Aw c'mon *guess* ;-) | | As to those latest changes, if you want I can import them for you -- | I'm quite good at that given all my practice :-) I've got everything | but the STUB stuff, looks like 10 minutes work. | | BTW, some good news and some bad: | | I'm on line right now with 201,264 bytes of free memory! | That's under command.com itself running under 4dos. Running | under 4dos alone, I get 209,648 bytes free, which is supposed to | be enough for a full 4dos shell with room to spare, but even so, | I get the not enough memory error from 'system()' which is the most | increadible | | BUMMMER #$%^&*@#$%% !!!!! | | So: either 4dos is reporting the wrong size to system() or | system() is reading it wrong, in any case the problem isn't with | Arachne, that's for sure. | | Damnation, I've worked bloody hard to get enough memory to sove the | 4dos problem and now this :-(( .. Oh well, it's still nice to have | the memory :-\ | | | | ============================================================================ | | Tue, Nov 09, 2004 - 7:17:25 pm | | glennmcc> | Oh yeah, I forgot... you already have the file showing that very small | change I made to re-enable HTTPSTUB | | And of-coarse, sisnce you did not make that same screw-up I made | | Your source code files already include _all_ of Joe's cookie code. | | So re-enabling HTTPSTUB is all that I've done since Oct 1st | | Just remembered one other change I made. | | But then of-coarse, you already have that change because you gave it to me. | | Dos memory(KB) is now Dos mem(bytes) | | | ============================================================================ | | Tue, Nov 09, 2004 - 8:00:58 pm | | glennmcc> | Now working on the wording of the new top-paragraph of changes.txt | | What do think of it so-far ? | | ______________________________________________________________________ | | While I have continued to contributed a few 'tweaks' and 'fixes' here & ther | | This version of Arachne is in-fact the result of work done by our new | team of programmers. | | These three men deserve all of the credit for making Arachne v1.79;GPL a rea | | Ray Andrews | Joe de Silva | Michal H. Tyc | ________________________________________________________________________ | | | | ============================================================================ | | Tue, Nov 09, 2004 - 9:58:56 pm | | ray> | Well, shucks :-) | | ... | | anyway, back to work ... now running with 211,056 free memory, | 219,184 under 4dos alone and system() still won't cooperate :-( so | I'm thinking the problem is with system(). Hmmmm maybe Michal | will be able to fix it with his ASM magic tricks. | | BTW Glenn, it's "da Silva" | | So you stil won't risk a guess as to how I'm getting these numbers? | Hint # 2: look at the sizes of all the .OBJ's ... which one is | the biggest? | | | | ============================================================================ | | Tue, Nov 09, 2004 - 10:02:31 pm | | ray> | Oh, forgot to report: loaded core.exe is now 420,464. | | | ============================================================================ | | Tue, Nov 09, 2004 - 10:09:29 pm | | glennmcc> | Dang-it, I thought I fixed that. :( | | _NOW_ it's fixed. :) | (in all places in the file) | | Changes since release 1.71;UE01 | -------------------------- | | 1.79;GPL,386+ stable for DOS | ---- | Release date: ??? ??, 2004 | | ____________________________________________________________________________ | While I have continued to contribute a few 'tweaks' and 'fixes' here & there | | This version of Arachne is in-fact the result of work done by our new | team of programmers. | | These 3 men deserve all of the credit for making Arachne v1.79;GPL a reality | | Ray Andrews | Joe da Silva | Michal H. Tyc | ____________________________________________________________________________ | | | | ============================================================================ | | Tue, Nov 09, 2004 - 10:32:03 pm | | glennmcc> | Well I'll be a ......... | | Turned-off overlaying of HTML.C and recompiled. | | Dos mem(bytes) 190624[+] green | | CORE.EXE is now using only 452,992 bytes | | How many files did you do that to ? | | ============================================================================ | | Wed, Nov 10, 2004 - 7:37:33 am | | ray> | Holy COW. | | | That's not what I did, but I sure like it, gotta try that out! | Maybe you've stumbled onto something. | | | ============================================================================ | | Wed, Nov 10, 2004 - 10:19:08 am | | glennmcc> | Yep, sometimes stumbling-along gets us further than stepping carefully. | | | | OK, | But if that's not what you did...... what the heck DID you do ? | | | ============================================================================ | | Wed, Nov 10, 2004 - 4:14:23 pm | | Joe> | Hi folks, | | Glenn, don't worry about the imperfect importation of the cookies code | in 1.77, I'm not fussed, these things happen. | | Oh yeah, you've reminded me again, I need to supply you a proper | list of changes for the 'changes.txt' file. One thing to consider, | should this be exclusively newest-first or latest-first? At the | moment, it's a hybrid, which is somewhat odd. | | Ray, was B5 derived from B3 before or after I ported my cookies code | into B3? I'll only accept responsibility if the latter. | | BTW, I have a few more little changes "in the pipeline", which I am | currently testing. Nothing major, but quite worthwhile, IMHO. | | | ============================================================================ | | Wed, Nov 10, 2004 - 5:24:28 pm | | ray> | Joe, | | No worries, it's all there in B5. If I've fouled anything up | during importation, of course the rap is mine. Also, I have an exhaustive | list of all the changes imported which should be sufficient, either | to stand on it's own, or Glenn can work from it if you don't feel | like making up your own list of changes. OTOH, my documentation lacks | much comment -- just line numbers, etc. | | Glenn, | | What I did was disect HTML.C into three pieces, since the size of | the program in memory is: startup code + static & global data + | staticaly linked code + the size of the largest overlay -- by | chopping up the largest overlay we reduce the loaded size. And since | HTML.C was three times bigger than any other module .... | | Now, what I can't figure out is how you got an improvement in | load size by making HTML.C static -- that defies my understanding | of what's going on vis. the explanation above; perhaps I missunderstand | something. Anyway, I'm going to see if I can duplicate your | results just now. | | | ============================================================================ | | Wed, Nov 10, 2004 - 5:34:54 pm | | glennmcc> | Yep, | | All I did was to select the project window from the window pull-down | | Scrolled down to HTML.C and hit Cntl+O for its options. | | Unchecked the box next to 'overlay this mod' | | Hit enter to save the change and recompiled. | | That's all it took to save more than 10,000 bytes of memory usage. | | | ============================================================================ | | Wed, Nov 10, 2004 - 5:48:14 pm | | ray> | Well, practice trumps theory every time but I'd still like to know | why that works. Perhaps static modules are somehow more efficient | than overlayed ones so even though HTML.C was added to the static | size (which should cost memory) the improvement in efficiency more | than compensated. My method is still much tighter (.exe = 420,464), | but what you've done should be *much* faster since a static module | is always in memory. Mail all done so back to the trenches :-) | | | ============================================================================ | | Wed, Nov 10, 2004 - 7:26:12 pm | | glennmcc> | I'm going to fire-up BCC right now and try combination of both. | | Split html.c into 3 parts AND staticly likn all 3 parts. | | BBL | | ============================================================================ | | Wed, Nov 10, 2004 - 7:32:38 pm | | ray> | Hey, cutting html.c isn't that simple! It took me several days | to figure out where to apply the knife, but you may do it faster. | Anyway this aproach has almost reached its limit: | | SENDMAIL.OBJ 8231 11-07-04 10:57a | ONMOUSE.OBJ 9211 11-07-04 10:57a | CGIQUERY.OBJ 9238 11-07-04 10:57a | PUTIKONW.OBJ 9472 11-07-04 10:57a | URLSTAT.OBJ 9646 11-07-04 10:57a | MISC.OBJ 10023 11-07-04 10:57a | HTMLSTAT.OBJ 10276 11-07-04 10:57a | GLOBALS.OBJ 10307 11-07-04 10:57a | HTMLUTIL.OBJ 11349 11-07-04 10:57a | HDRAWGIF.OBJ 12861 11-07-04 10:57a | HTTP.OBJ 13182 11-07-04 10:57a | GUIDRAW.OBJ 13298 11-07-04 10:57a | CONFIG.OBJ 13770 11-07-04 10:57a | HTMLDRAW.OBJ 14362 11-07-04 10:57a | MAIN.OBJ 14363 11-07-04 10:57a | ATOMS.OBJ 14399 11-07-04 10:57a | HTML.OBJ 14463 11-09-04 8:12p | HTMLTAG2.OBJ 20610 11-10-04 6:40p | GUIEVENT.OBJ 20784 11-07-04 10:57a | HTMLTAG.OBJ 21553 11-10-04 6:40p | 412,804 bytes in 57 files and 0 dirs 983,040 bytes allocated | 20,944,699,392 bytes free | | Without fooling around with GUIEVENT.C, the three largest overlayed | modules are now about the same size, so that's about as far as we | can go with this aproach (cutting) but who knows what might | turn up with experimenting with overlay >> static ?? | | | | ============================================================================ | | Wed, Nov 10, 2004 - 7:33:52 pm | | glennmcc> | Nevermind. | | I had forgotten just what kind of a nightmare that file is. | | Since yours is already split.... | | I'll let you try statically linking the 3 parts. | | ============================================================================ | | Wed, Nov 10, 2004 - 7:42:05 pm | | glennmcc> | re:"Well, practice trumps theory every time...." | | Yep... there is 'theory' and there is 'reality'. ;-) | | The reality of this particular situation is that the theory of the | 'classic formula' for calculating loaded size sure as heck don't | seem to prove-out in actual practice. ;-) | | ============================================================================ | | Wed, Nov 10, 2004 - 8:12:07 pm | | ray> | Ah, now that you mention it, let me wax personal: when I cleaned that | file it was the toughest thing I had ever done in my life, that file | alone took me almost a month. 4000 lines, and try to find a bug | entered by one of my automatic formating code-grinders :-(( | | I vowed at the time I would kill that file, or at least chop it | into managable pieces which I've now done and, boy, do I feel | good about it :-> Revenge is sweet | | I think I'll let you fool around with this, if I don't hear from Michal | by Friday (he is very busy at work right now) I'll send what I've | got, the code he has for me isn't vital, just a way of forcing | some strings into the data segment (only Michal can do this sort | of black magic, it's part of the strategy of my changes to 'config.c') | | | | BTW, odds and ends: I'm just remembering about that BCC stuff -- I | didn't pay much attention at the time, since up till now I've always | waited for your releases before updating my own code, but that looks | like an obsolete strategy. | | | ============================================================================ | | Wed, Nov 10, 2004 - 8:15:28 pm | | ray> | !@#$%^# | | My on line clock has stopped working. Same .exe moved to | by 'regular' directory and it works fine. | | Joe, is there some other file needed for the online clock to | work? (Man, there is always something to bug you). | | | | ============================================================================ | | Wed, Nov 10, 2004 - 9:23:03 pm | | glennmcc> | Just checked it. | | The online timer works with both | | IP_Address PPP | | and with | | IP_Address BOOTP | | But it does not work with | | IP_Address %MYIP% | | Nor does it work with static IP | | IP_Address 4.224.207.158 | | | So, does your other dir in which the online timer is not working | contain something other than PPP or BOOTP in arachne.cfg ? | | ============================================================================ | | Wed, Nov 10, 2004 - 11:04:29 pm | | ray> | Na, it's not that. Blasted bugs, I'm fried. | | | | ============================================================================ | | Thu, Nov 11, 2004 - 1:58:45 am | | Joe> | Hi all, | | The online timer uses 'ppp.log' as its time reference. | If you change to a different directory during an online | session, don't forget to copy over your 'ppp.log' file. | When you terminate the session, use "arachne -u" from | both directories to clean up. As you can probably tell, | Arachne was never intended to change directories during | a session (I presume this is what you've done ...) | | | ============================================================================ | | Thu, Nov 11, 2004 - 9:56:20 am | | ray> | Joe, | | Thanks, but that wasn't it. Somehow I managed to mangle your PPPlog() | code importing it into B5. God, did I mangle it; I feel my soul | is in peril I mangled it so badly. I pride myself on the precision | of my work but I must have been having one hell of a bad day. Or ... | did you ever release more than one version of PPPlog()? Maybe I | got versions mixed up or something (I need an excuse in the worst | way ;-) Anyway I've fixed it now, but this is going to give me | nightmares -- sloppy work just isn't an option, not never. | | | ============================================================================ | | Thu, Nov 11, 2004 - 9:58:39 am | | ray> | Oh one more thing: | | Joe, I notice that PPPlog() is linked overlay. Given that the timer | is accesses so often, would it be good to link this static? ... or | is the actual drawing of the timer controled elsewhere such that | PPPlog() iself can remain overlayed? | | | | | ============================================================================ | | Fri, Nov 12, 2004 - 4:17:44 am | | phantom> | Ray, | | No, I don't recall more than one version of PPPlog (senility | notwithstanding;-). | | As for overlays, that's something I've never looked into. | Someday I may have the time and inclination, but there's | so much to investigate in the code itself, I'm not sure | I'll get around to this any time soon. | | | ============================================================================ | | Fri, Nov 12, 2004 - 4:18:55 pm | | ray> | Joe, | | Well, so much for my excuse ;-) ... maybe I got hit on the head | or something or someone put LSD in the coffee. | | As to overlays vs. static modules, I have a feeling Glenn might want | to look into this, and I'm quite sure there are useful changes to | be made. I've had the feeling that some of the choices that have | been made are poor. For example 'mime.c' itself tells us that it should | be linked static, but in fact it has always been linked overlay. | Also, it seems hard to believe that 'main.c' should not be static | for obvious reasons, but it is. | | | | | ============================================================================ | | Fri, Nov 12, 2004 - 6:17:29 pm | | glennmcc> | Do I misunderstand ? | | In all of my PRJ files main.c is overlayed. | | | | ============================================================================ | | Fri, Nov 12, 2004 - 6:22:28 pm | | glennmcc> | Just tested it 3 new compiles after changeing each of these | from overlayed to static. | | main.c | mime.c | init.c | | Each one causes a 4k increase in mem usage when overlay is turned off. | | ============================================================================ | | Fri, Nov 12, 2004 - 7:43:09 pm | | ray> | No, that's right, main.c is overlayed, but it seems very strange that | it is, no? Given that it contains the main timer loop it looks to | me like it should be static (costs 8K but we can afford it now :-) | | ... hey, it only cost you 4K? ... cost me 8 (I think, better double | check) | | | | ============================================================================ | | Fri, Nov 12, 2004 - 7:45:35 pm | | ray> | Glenn, I've got B~5 more or less ready, I'd like to wait for | Gregy to finish XLOPIF.H but if you want I'll upload what I've got | now -- it's just provisional code anwyway, so it hardly matters that | it isn't finished. Let me know. | | | | | ============================================================================ | | Fri, Nov 12, 2004 - 10:34:00 pm | | glennmcc> | IIRC, Gregy mentioned that he'll be done in few days. | | I can wait if you can. ;-) | | ============================================================================ | | Sat, Nov 13, 2004 - 8:37:38 am | | ray> | OK | | I'll take the weekend off and see what Gregy hath wrot on Sunday. | | | ============================================================================ | | Sun, Nov 14, 2004 - 10:15:07 am | | ray> | Glenn, | | While waiting for Sgt. Fieg, I may as well import that htmlstub stuff | but I can't remember where it came from, please advise. | | | | ============================================================================ | | Sun, Nov 14, 2004 - 10:15:56 am | | ray> | Oopps, make that "Sgt. Fieg" | | | ============================================================================ | | Sun, Nov 14, 2004 - 3:07:45 pm | | glennmcc> | 1) it's not htmlstub .... it's httpstub | | 2) if I were looking for something concerning httpstub........ | I'd look in http.c (specifically, for this section) | | | | | //!!glennmcc: begin Oct 17, 2004 -- re-enable HTTPSTUB | //(this entire section had been commented-out) | ///* | else //httpstub = non TCP/IP stuf =========================== | { | int f,l; | char reqname[80]; | struct ffblk ff; | | | ============================================================================ | | Sun, Nov 14, 2004 - 3:10:09 pm | | glennmcc> | The whole thing is of-coarse in this file........ | | http://f1.grp.yahoofs.com/v1/......../glennmcc/sr175177.zip | | | | ============================================================================ | | Sun, Nov 14, 2004 - 4:05:35 pm | | ray> | Ok, just that one block then I take it. | Sure, I could have looked for it myself but it's faster and more | acurate to ask. | | Then there's the file name change, got that, then anything else to | bring me up to date? | | | ============================================================================ | | Sun, Nov 14, 2004 - 4:06:26 pm | | ray> | Make that "file numbering system". | | | ============================================================================ | | Sun, Nov 14, 2004 - 5:17:21 pm | | ray> | Glenn, it looks like that block has been changed, I get two errors | compiling it within B0. I tried the url above with no luck, | can't get near it. | | | | ============================================================================ | | Sun, Nov 14, 2004 - 5:43:32 pm | | glennmcc> | The URL above was incomplete (/......../) was what looks just like a | yahoo cookie. | | The file is in my dir on the Yahoo group. | | Yep, same 2 errors here till I commented-out | one small section within that block. | | ============================================================================ | | Sun, Nov 14, 2004 - 5:47:22 pm | | glennmcc> | The file names code in popget.c consists of............ | | maketime: | //!!glennmcc: Nov 10, 2004 --- HEX names instead of decimal | //sprintf(fname,"%ld.CNM",starttime+process); | //!!glennmcc: Nov 11, 2004 -- begin at Annnnnnn.CNM | //it is 4nnnnnnn.CNM right now so we need to add | //60000000 hex which is 1610612736 dec | //sprintf(fname,"%lx.CNM",starttime+process); | sprintf(fname,"%lx.CNM",starttime+process+1610612736l); | | makename: | strcpy(str,configvariable(A,"MailPath",NULL)); | if(strlen(fname)>12) | strcat(str,fen(fname)-12]); | else | strcat(str,fname); | | if(!findfirst(str, | { | fname[8]++; | //!!glennmcc: Nov 10, 2004 --- HEX names instead of decimal | //if(fname[8]>'9') | if(fname[8]>'F') | { | starttime=time(NULL); | goto maketime; | } | else | goto makename; | } | | | | ============================================================================ | | Sun, Nov 14, 2004 - 8:57:40 pm | | ray> | OK I'll grab the file from Yahoo. But are you saying that | the code above from popget needs to be changed as well? | (This is your project, I don't think, I just clean ;-)) | | | | ============================================================================ | | Sun, Nov 14, 2004 - 9:01:18 pm | | ray> | Hey, what file? I don't see anything new enough. | | ============================================================================ | | Sun, Nov 14, 2004 - 9:02:58 pm | | ray> | Or just show me the commented block, I've got everything else | imported Ok so once the error is commented I'm good to go. | | | ============================================================================ | | Mon, Nov 15, 2004 - 12:07:14 am | | Joe> | Me confused! :-( | | 1. What's happening with/to 'xlopif.h'? | | 2. What's the "oops" about? | | 3. What are you talking about? | | 4. Again, | what's this sub-thread about? | | Maybe that senility is catching up with me, but I can't make sense | of most of the above, it looks like snippets of dialogue ... maybe | I'm having a bad day, perhaps it will make sense to me tomorrow. :-/ | | PS. Good to hear Sgt. Fieg is still alive & kicking. | | | | | ============================================================================ | | Mon, Nov 15, 2004 - 4:04:01 am | | Not Fieg> | Ah, I've figured out #2 ... what confused me was that the "Oopps" | posting had the same spelling as the previous one, so I guess it's | a case of double-oops! | | Anyway, it's good to hear Sgt. Feig isn't AWOL ... ;-) | | | ============================================================================ | | Mon, Nov 15, 2004 - 7:25:33 am | | ray> | Sgt Feig may not be AWOL, but he'll be busted down to private if | he doesn't come up with that translated file < G g G > | | | ============================================================================ | | Mon, Nov 15, 2004 - 7:28:22 am | | ray> | Anyway, Joe, the bits of confusion above don't amount to much; I was | reactivating that code for HTTPSTUB and there are two bugs, but | Glenn says he has a few lines to comment out to fix it. 'File numbering' | is just the conversion of the automatic mail names to Hex from | Decimal to end the rollover problem. | | | | ============================================================================ | | Mon, Nov 15, 2004 - 5:42:36 pm | | ray> | Joe, | | Looks like B5 will be uploaded in a few days, do you have anything | to throw into the stew? | | | ============================================================================ | | Mon, Nov 15, 2004 - 7:38:44 pm | | glennmcc> | Sorry Ray.... my bad. | | The file on Yahoo does not have the re-enabled HTTPSTUB. | | Here's the modified http.c | | http://www.cisnet.com/glennmcc/my-stuff/http.c | | And this is just the modified 'block'. | | http://www.cisnet.com/glennmcc/my-stuff/httpstub.txt | | | ============================================================================ | | Mon, Nov 15, 2004 - 7:40:45 pm | | glennmcc> | This is the section that needed to be commented-out to stop those errors. | | /* | if(*poststring) | { | querystring=ie_getswap(GLOBAL.postdataptr); | if(!querystring) | MALLOCERR(); | write(f, querystring,strlen(querystring)); | } | */ | | | | ============================================================================ | | Mon, Nov 15, 2004 - 10:41:27 pm | | ray> | Ah, much better. But there is one more change besides the small | block above, which is shown in your full http.c file and that is | the commenting of the line: | | p->buf[httplen] = '0' | | ... which produced my 2nd error. That line is interesting because | it *has* to be a bug since 'httplen' is a structure member not | an independant variable so the line above can't compile as written. | I'm curious as to what it's intent was. ... anyway, like I said | this is your project so I jus' mind my own business ;-) | | | ============================================================================ | | Mon, Nov 15, 2004 - 10:45:54 pm | | ray> | Oh, but that 2nd change should be commented no? I've yet to see you | change something structural without flagging it, which is exactly | as it should be. | | BTW, Gregy is in that hell on earth X_LOPIF.H if he makes it out | alive I'm recomending a decoration. | | | ============================================================================ | | Mon, Nov 15, 2004 - 11:11:53 pm | | glennmcc> | Yep, quite correct. | | Since HTTPSTUB re-enablement (is that a word), is still experimental. | | I have not finished my testing/debugging. | | When I'm done... I'll have everything properly commented. | | | | ============================================================================ | | Mon, Nov 15, 2004 - 11:15:38 pm | | glennmcc> | Oh my gawd, he's in THERE ???? | | He deserves that decoration right now | for going in there in the first place ;-) | | ============================================================================ | | Tue, Nov 16, 2004 - 5:32:22 pm | | ray> | re HTTPSTUB: OK | | re Gregy: | | Yes, it's true, and I sent him ... I did all the other files myself | but I just didn't have the grit for X_LOPIF.H so what to do but | turn to one's sargeant? | | Purple Heart for sure, and what? Silver Star or should it be | the CM of H? | | | ============================================================================ | | Tue, Nov 16, 2004 - 5:34:10 pm | | ray> | Too bad you Yanks don't have the VC | | | ============================================================================ | | Tue, Nov 16, 2004 - 8:01:05 pm | | glennmcc> | You sent him in there ??? | | I must have been mistaken, I thought you liked Gregy. | | You must really HATE him to have sent him in there. ;-) | | ============================================================================ | | Tue, Nov 16, 2004 - 8:02:14 pm | | glennmcc> | What's the VC ? | | ============================================================================ | | Tue, Nov 16, 2004 - 9:05:20 pm | | ray> | The VC is the Victoria Cross, the most respected military medal | in the world in any country at any time. All the rest are | baubles by comparison. Anyway, I hope Gregy will execute a | strategic withdrawl if it get's too hot in there < g g g > | | | | ============================================================================ | | Thu, Nov 18, 2004 - 4:42:18 pm | | glennmcc> | No comments on my 'speedometer' ? | | ============================================================================ | | Thu, Nov 18, 2004 - 6:38:50 pm | | ray> | ... just playing with it now, and I love it :-). | | ("Nuts", the apprentice said to himself ... "and I thought it would | be difficult to code") | | BTW, Sgt. Feig is fully engaged and is advancing on all fronts; | we expect victory within a few days. | | | ============================================================================ | | Thu, Nov 18, 2004 - 7:03:53 pm | | glennmcc> | That's what he said a few days ago. ;-) | | But then again... we must always expect that | there will be set-backs during any battle. | | | ============================================================================ | | Thu, Nov 18, 2004 - 7:05:45 pm | | glennmcc> | As for the byte rate indicator. | | Which do you prefer ? | | The current code which shows bytes/sec ? | | Or would you prefer KB/sec ? | | ============================================================================ | | Thu, Nov 18, 2004 - 8:35:56 pm | | ray> | I've just been farting around with that myself. I'm thinking that | bytes/sec is better, but that the display should show a | 5 sec average, just so that it's easier to read. Also, I've been | playing with the idea of minimum width fields for the display so that | the line isn't always acordioning in and out but shows each | field in the same place all the time, it should make everything | more readable, what do you think? | | | ============================================================================ | | Thu, Nov 18, 2004 - 10:04:51 pm | | ray> | Hmmm, this field width thing might be good, it looks a little | strange, but it sure is easier to follow what's happening. And it's | good to know that Arachne can achieve download speeds of up to 7KB/S. | On my machine, that's as good as anything I've ever had in | Window$ | | | | ============================================================================ | | Thu, Nov 18, 2004 - 10:05:38 pm | | glennmcc> | Just now tried it.......... Don't like it. :( | | I'll keep mine the way it is. :) | | ============================================================================ | | Fri, Nov 19, 2004 - 7:14:15 pm | URL | | glennmcc> | Is this 'too much of a good thing' ? | | ============================================================================ | | Sat, Nov 20, 2004 - 8:03:55 pm | | Joe> | Hmmm ... | | I'd scrap the ET and just keep the ETA (partly because the ET is | more clutter than it's worth, partly because this ET doesn't stand | for the same two letters in ETA, so it kinda conflicts with it). | | Conversion to hh:mm:ss format would be nice (else expect continual | requests for this from the user base;-). | | | | ============================================================================ | | Sun, Nov 21, 2004 - 2:54:12 pm | | glennmcc> | Yeah, they are a demanding lot..... aren't they ? | | ============================================================================ | | Sun, Nov 21, 2004 - 7:21:38 pm | URL | | glennmcc> | How's this look ? ;-) | | ============================================================================ | | Sun, Nov 21, 2004 - 9:14:41 pm | | ray> | The cosmetics of formating the string can be fooled around with, | I'd say, but the extra info is very nice. Hand over your code, | stranger, I've gotta play with this. | | BTW I haven't given up on | minimum field widths, I'm now thinking that the trick is to | pad to 5 or 6 spaces for download file size, that way the display | isn't jerking around too much with the smaller files, but with the | bigger files the display can still expand rightwards as needed. | And since you are on such a roll with this, how 'bout the 5 second | average? Not important of course, but might be nice. | | | | | | | ============================================================================ | | Sun, Nov 21, 2004 - 10:52:59 pm | | glennmcc> | re: 5sec average | | See....... Thu, Nov 18, 2004 - 10:05:38 pm | | ============================================================================ | | Sun, Nov 21, 2004 - 10:59:41 pm | | glennmcc> | --- in messages.h --- | //!!glennmcc: Nov 17, 2004 -- include bytes/sec rate | //!!glennmcc: Nov 21, 2004 -- added Elapsed time and ETA | #define MSG_X_OF_Y "%s%ld of %ld bytes -- @ %ld bytes/sec. )(time remaining | //#define MSG_X_OF_Y "%s%ld of %ld bytes -- @ %ld bytes/sec. ) ..." | //#define MSG_X_OF_Y "%s%ld of %ld bytes ) ..." | //!!glennmcc: end | _____________________ | | --- in http.c --- | void Download(struct HTTPrecord *cache) | { | long fpos=0; | char dl[80],str[80]; | int rd=1, prc=0; | | //!!glennmcc: Nov 17, 2004 -- include bytes/sec rate | //!!glennmcc: Nov 21, 2004 -- added time remaining | long starttime=time(NULL), elapsedtime, heta, meta, seta; | //!!glennmcc: end | | if(cache->handle!=-1) | fpos=a_filelength(cache->handle); | | while(rd>0) | { | #ifdef POSIX | rd=tickhttp(cache,p->buf,socknum); | #else | rd=tickhttp(cache,p->buf,socket); | #endif | fpos+=rd; | | strcpy(dl,MSG_DOWNLD); | | #ifndef POSIX | if (!sock_dataready( socket ) || rd==0) | { | #endif | if (cache->knowsize && cache->size>0) | { | prc=(int)(100*fpos/cache->size); | | //!!glennmcc: Nov 17, 2004 -- include bytes/sec rate | //!!glennmcc: Nov 21, 2004 -- added time remaining | elapsedtime=(time(NULL)-starttime); | if(elapsedtime>1 && prc<99) | { | //if(prc>5) //uncomment to start only after 5% | //{ //uncomment to start only after 5% | heta=(cache->size-fpos)/(fpos/(int)(elapsedtime))/3600;//hours | meta=((cache->size-fpos)/(fpos/(int)(elapsedtime))-(heta*3600))/60;//min | seta=(cache->size-fpos)/(fpos/(int)(elapsedtime))-(heta*3600)-(meta*60);//se | //} //uncomment to start only after 5% | //else{heta=0;meta=0;seta=0;} //uncomment to start only after 5% | sprintf(str,MSG_X_OF_Y,dl,fpos,cache->size,fpos/(int)(elapsedtime),heta,meta | } | else | { | heta=0;meta=0;seta=0; | sprintf(str,MSG_X_OF_Y,dl,fpos,cache->size,fpos/1,heta,meta,seta); | } | //sprintf(str,MSG_X_OF_Y,dl,fpos,cache->size); | //!!glennmcc: end | _________________ | | | ============================================================================ | | Mon, Nov 22, 2004 - 12:29:27 am | | Joe> | Hi Glenn, et. al. | | Re: Yeah, they are a demanding lot..... aren't they ? | Sure, but they become very quiet when you ask for volunteers! | | Re: How's this look ? | I'd do something like this : | (250632 of 194543616 bytes @ 5221 bytes/sec. - time remaining 10:20:13) | Something fairly neat, hence no double parenthesis. Double spacing around | the "@" and possibly the "-" to visually break up the fields). Just my | aesthetics here ... | | | | | ============================================================================ | | Mon, Nov 22, 2004 - 9:14:14 am | | glennmcc> | No problem. | | We just change this ..... | | #define MSG_X_OF_Y "%s%ld of %ld bytes -- @ %ld bytes/sec. )(time remaining | | To this........ | | #define MSG_X_OF_Y "%s%ld of %ld bytes @ %ld bytes/sec. - time remaining %ld | | | ============================================================================ | | Mon, Nov 22, 2004 - 5:35:36 pm | | ray> | Yeah, I like Joe's aesthetics, but these are trivial things. | Glenn, I'll wait for the code above in the cleaned sources -- it gives | me nightmares to look at the old code ;-) ... now if Gregy would | just report for duty ... | | ============================================================================ | | Mon, Nov 22, 2004 - 6:53:40 pm | | glennmcc> | Strange.... Greggy keeps showing up from time-to-time on the email list. | | But he is AWOL from his own message board. | | He can't even seem to find the time to fix the incorrect commenting | that causes that '-->' to be down at the bottom of this page. | | | --> Copyright(c)2002? maybe? maybe not? | | | ============================================================================ | | Mon, Nov 22, 2004 - 9:50:44 pm | | ray> | Yeah. Nuts, don't know what to do. He said he'd have the translations | done by this weekend and it seems like it would be rotten of me to | upload without waiting for him, but he hasn't said anything ... | But we don't pull out and leave one of our guys abandoned in X_LOPIF.H | do we? Seems like a question of honor. | | | ============================================================================ | | Mon, Nov 22, 2004 - 11:57:08 pm | | Joe> | Silly question Ray, but have you tried e-mailing Greggy | about progress on the translation? Message boards are | fine, but you have to remember to visit, whereas e-mail | gives notification of anything new. | | This is a big file, so to be "running a bit late" is not | unexpected. BTW, when did Greggy learn Czech? | | | ============================================================================ | | Tue, Nov 23, 2004 - 7:34:35 am | | ray> | Joe, | | Of course I've e-mailed him directly. And he's just importing | Christof's translations into B5, not making the translations | himself. Mind in x_lopif.h even that work is a stiff job. | | | | ============================================================================ | | Tue, Nov 23, 2004 - 5:53:00 pm | | ray> | Gregy just checked in -- seems he'll be a while yet so I'll upload | B~5 as it is now, and we'll just ammend X_LOPIF.H latter. | Besides, it's provisional code anyway and there are bound to be other | changes to make -- it is so very very too big a jump. | | | ============================================================================ | | Tue, Nov 23, 2004 - 6:26:16 pm | | glennmcc> | Give us a yell when it's uploaded. | | I am very anxious to compile it. | | ============================================================================ | | Tue, Nov 23, 2004 - 8:18:36 pm | | ray> | It otta be there within the hour, or tommorow at the latest. | | | | ============================================================================ | | Wed, Nov 24, 2004 - 4:26:40 pm | | Joe> | Hey, x_lopif.h wasn't among the code Christof translated back | whenever (February, IIRC)! If there's new stuff translated, | the sooner we can study it, the sooner we can benefit from it. | It may have have helped me some when I was doing my changes | to lopif :-(still no volunteers to see if it works on other | video cards, before I release it on an unsuspecting world). | | So Ray, can you please upload any new stuff from Christof to | AD, in the 'Christof' directory. Thanks. :-) | | ============================================================================ | | Wed, Nov 24, 2004 - 5:27:51 pm | | ray> | Joe, | | Hmmmm, to tell the truth, I've just poached stuff hear and there, | as I could grab it -- I don't have a definitive version of | Christof's work and ... (where did I get it) ... it must have | come from his directory on Yahoo, since I haven't been it touch with | him directly. IIRC he released two or maybe three cumulative | editions of his work and the x_lopif.h translation was in the last | one. Check the above, and if you can't find what you want, I'll | scour my system, but probably won't be able to add anything. | | | ============================================================================ | | Wed, Nov 24, 2004 - 8:35:18 pm | | ray> | Joe, | | Yup, the above is correct, you can grab Christof's latest work in | his directory. | | BTW, started work on insight today and it's an even worse mess | than the core sources, if such a thing were possible -- the first | build shows 49 warnings :-( Dead variables, faulty argument lists, | improper typecasting ... this is unbelievably bad code. | Glenn you're a tougher guy than me. | | | | | ============================================================================ | | Wed, Nov 24, 2004 - 10:09:54 pm | | glennmcc> | | You must be doing something wrong. | | Insight compiles perfectly for me. | | No warnings , no errors...... no problems of any kind. | | All you need to do is CD to the insight SRC dir. | Run BC and hit ALT+C and choose 'build all'. | | Everything works perfectly if you do it that way. | | ============================================================================ | | Thu, Nov 25, 2004 - 12:31:29 am | | Joe> | Hi, | | Something I've noticed - BC often doesn't present warning | messages when you "build all". Errors are fine, but I've | found sometimes that I need to compile the individual | module to see all its warnings. Perhaps that's the case | here? | | I don't (can't) use Insight for e-mail, so I haven't taken | much (any) interest in it, but if Wwwman was any indication, | the ancillary applications have much worse source code than | Core itself! :-/ | | I've just checked Christof's web site (I hadn't done so for | quite a while, since there hadn't been anything new for ages) | and the 'aratrans.zip' file is updated as of a month ago. I | guess this is what you've picked up "along the way", Ray (the | version at AD is from February, IIRC). Well, I'll have to | check what's new in there when I have a chance ... | | ============================================================================ | | Thu, Nov 25, 2004 - 10:03:46 am | | ray> | JOe, rodger all that. | | Glenn, do you have all your warnings activated in the IDE? | | | | ============================================================================ | | Thu, Nov 25, 2004 - 10:05:17 am | | ray> | Hey, I just heard from Michal, he's being worked to death I think, but | he's still alive ;-) | | | ============================================================================ | | Thu, Nov 25, 2004 - 6:16:44 pm | | glennmcc> | Yep, all warnings are active and it's set to stop on errors. | | No warnings and no errors have ever happened when I compile insight. | | ============================================================================ | | Thu, Nov 25, 2004 - 10:05:43 pm | | ray> | All warnings ON you say, now that's strange. The code as I have | it from you is ... | | | ============================================================================ | | Thu, Nov 25, 2004 - 10:07:26 pm | | ray> | ... 8.57 and there are definitely warning level defects in the code, | like unused variables. Let me go get some ... | | ============================================================================ | | Thu, Nov 25, 2004 - 10:13:51 pm | | ray> | Ok, insight.c: | ² | //========================================================================== | //Insight main function ² | //========================================================================== | void main(int argc, char *argv[]) ² | { //Tmsghdr msghdr; //Variable not used, get warning ² | int t, i; ² | //int t1, t2, t3; //Ditto ² | char s[256], ² | //s1[256], //Ditto ² | querystr[256], *sp; ² | char fname[256]; ² | Tcfgblk cfg1; ² | | Anyway, these variables aren't used so the warning I get from the | compiler is correct. You get no warning? Am I using the right code? | Your sure all your warnings are on? Huh, don't know what to think. | ² | | ============================================================================ | | Thu, Nov 25, 2004 - 10:15:29 pm | | ray> | What happens when you comment those variables? | | | | ============================================================================ | | Fri, Nov 26, 2004 - 6:34:18 pm | URL | | glennmcc> | Looks just fine to me. | | ============================================================================ | | Fri, Nov 26, 2004 - 6:42:52 pm | | glennmcc> | --- t1 is used in all of these places --- | | ---------- e:convmsg.c | [88] unsigned int t,t1,t2,t3; | [92] for(t=0;t<8;t++){t1=(s[t]-'0')&0x0f; | [93] for(t2=0;t2<4;t2++){if(t1&0x01)barr[t3]=1;t1>>=1;t3++;}} | [96] t1=0;t3=0; | [97] | for(t=0;t<8;t++){t1=0;for(t2=0;t2<5;t2++){t1<<=1;t1|=barr[t3];t3++;}t1&=0x1f | | ---------- e:idx.c | [151]{ int t,t1; | [214] int t,t1; | [215] do{t1=0; | [219] {t1=1; | [223] }while(t1); | [228] int t,t1,t2,t3; | [232] for(t1=0;t10x3f)t=0x3f;else if(t<=0)t=1; | [236] for(t1=0;t1maxrowlength){printf("\n");rowlen=0;} | [338] if(s[t]==' '){sp=strchr(s+t+1,' ');if(sp){t1=(int)(sp-(s+t+1));}else { | [339] if(wrap)if(rowlen+t1>maxrowlength){fprintf(f,"\n");rowlen=0;putquotehd | [586] int t1,t2,t3; | | ---------- e:string1.c | [24]{ int t,t1;char *sp; | [26] t1=strlen(sp); | [27] for(t=t1-1;t>0;t--)if(sp[t]<=' ')sp[t]=0;else break; | [33]{ int t,t1; | [34] t1=strlen(s); | [35] for(t=t1-1;t>=0;t--)if(s[t]<=' ')s[t]=0;else break; | [66]{ int t,t1; | [67] t=0;t1=0; | [68] while(s[t]!=0){s[t1]=s[t];if((t1==0)||(s[t1-1]!=' ')||(s[t1]!=' '))t1++ | [69] s[t1]=0; | [74]{ int t,t1; | [75] t=0;t1=0; | [76] while(s[t]!=0){s[t1]=s[t];if(s[t1]!=' ')t1++;t++;} | [77] s[t1]=0; | [97] int t,t1; | [99] t1=sp-s; | [100] for(t=t1;t>3;t-=3) | | __________ | | --- t2 is used in all of these places --- | | ---------- e:convmsg.c | [88] unsigned int t,t1,t2,t3; | [93] for(t2=0;t2<4;t2++){if(t1&0x01)barr[t3]=1;t1>>=1;t3++;}} | [97] | for(t=0;t<8;t++){t1=0;for(t2=0;t2<5;t2++){t1<<=1;t1|=barr[t3];t3++;}t1&=0x1f | [398] ext2mime(s,sn); | | ---------- e:idx.c | [228] int t,t1,t2,t3; | [234] for(t2=0;t2<6;t2++){if(t&0x01)s[t3]=1;t>>=1;t3++;}} | [237] {t=0;for(t2=0;t2<8;t2++){t<<=1;if(s[t3])t|=0x01;t3++;}st[t1]=t;} | | ---------- e:insight.c | [305]{ int t,t1,t2; | [328] if(testhttp&cddy(s+t,tpy(sh,s+t,t2);sh[t2]=0; | [586] int t1,t2,t3; | | _________________ | | | --- t3 is used in all of these places --- | | ---------- e:convmsg.c | [88] unsigned int t,t1,t2,t3; | [90] t3=0; | [93] for(t2=0;t2<4;t2++){if(t1&0x01)barr[t3]=1;t1>>=1;t3++;}} | [94] for(t=0;t<8;t++){if(attachnmbr&0x80)barr[t3]=1;attachnmbr<<=1;t3++;} | [96] t1=0;t3=0; | [97] | for(t=0;t<8;t++){t1=0;for(t2=0;t2<5;t2++){t1<<=1;t1|=barr[t3];t3++;}t1&=0x1f | [100] s[1]=t3; | | ---------- e:idx.c | [228] int t,t1,t2,t3; | [230] memset(s,0,255);t3=0; | [234] for(t2=0;t2<6;t2++){if(t&0x01)s[t3]=1;t>>=1;t3++;}} | [235] t3=0; | [237] {t=0;for(t2=0;t2<8;t2++){t<<=1;if(s[t3])t|=0x01;t3++;}st[t1]=t;} | | ___________________ | | | Need I go on ? | | ============================================================================ | | Fri, Nov 26, 2004 - 6:47:51 pm | | glennmcc> | Do you have all of the correct files ? | | _6,210__11-05-00___6:02p__e:\arachne\src\insight\attach.c | _2,212___6-26-04___8:15p__e:\arachne\src\insight\base64.c | ___298___1-22-99___5:56p__e:\arachne\src\insight\cfg.h | 21,989___8-03-04___8:03p__e:\arachne\src\insight\convmsg.c | ____49___1-03-03__11:45p__e:\arachne\src\insight\css.txt | _1,159___4-11-04__11:18p__e:\arachne\src\insight\define.h | ___165___7-21-00__11:02a__e:\arachne\src\insight\detidx.c | _9,326__11-02-95__12:00a__e:\arachne\src\insight\dirport.h | _2,172__11-02-95__12:00a__e:\arachne\src\insight\dosfiles.h | _1,561__11-02-95__12:00a__e:\arachne\src\insight\extkword.h | _4,930___9-10-04___2:53p__e:\arachne\src\insight\folders.c | _3,156__12-10-99___4:00a__e:\arachne\src\insight\getico.c | 15,377___4-27-04__10:23p__e:\arachne\src\insight\hdrout.c | _1,057___4-11-04__11:24p__e:\arachne\src\insight\help.c | 14,672___7-21-00__11:01a__e:\arachne\src\insight\idx.c | _1,925__12-10-99___4:04a__e:\arachne\src\insight\inifile.c | 24,870___9-10-04___2:55p__e:\arachne\src\insight\insight.c | ___627__11-26-04___7:22p__e:\arachne\src\insight\insight.dsk | 79,360___9-10-04___3:22p__e:\arachne\src\insight\insight.exe | _8,893__11-26-04___7:22p__e:\arachne\src\insight\insight.prj | _3,333__12-11-00___2:34a__e:\arachne\src\insight\procmime.c | ___278__10-02-99___3:32p__e:\arachne\src\insight\savefold.c | ___638__11-02-95__12:00a__e:\arachne\src\insight\sniptype.h | _1,144___1-15-00__12:37a__e:\arachne\src\insight\str.c | _1,050__12-11-00___2:46a__e:\arachne\src\insight\str.h | _3,117__12-10-99___3:55a__e:\arachne\src\insight\string1.c | ___666__12-10-99___3:46a__e:\arachne\src\insight\string1.h | ___464___1-22-99___2:18p__e:\arachne\src\insight\tidx.h | _2,026___4-26-97___4:59a__e:\arachne\src\insight\truname.c | ____69__12-10-99___3:55a__e:\arachne\src\insight\truname.h | | | | | ============================================================================ | | Fri, Nov 26, 2004 - 7:07:32 pm | | glennmcc> | Just now checked the ZIP file on my site to be sure that it is intact. | | No problem, 30 files with a total size of 212,794 bytes | | ============================================================================ | | Fri, Nov 26, 2004 - 8:22:28 pm | | ray> | Hmmm, this is strange. I've been using insight with those variables | commented with no problem. Thanks for the detailed info, I'll check the | quotations above and see what gives. It's hard to believe I'd get | a working .exe with needed variables commented, still ... | | | ============================================================================ | | Fri, Nov 26, 2004 - 9:17:16 pm | | ray> | Glenn, those variables you quote above are not the same ones! | They are all local automatics which might be needed in their | respective functions, but that says nothing about their | status in the function I quoted. | | But, it seems I don't have the latest files which is a pain in the | arse since I've already cleaned a few, but it shouldn't be too much | trouble to update. Off to your site ... but before I go, let me | be prudent enough to ask for the file name, just to be sure I don't | grab the wrong thing. | | | | | ============================================================================ | | Fri, Nov 26, 2004 - 11:45:31 pm | URL | | glennmcc> | http://www.cisnet.com/glennmcc/ar177gpl/insight.zip [77,458 bytes] | | ============================================================================ | | Sat, Nov 27, 2004 - 12:42:11 am | | glennmcc> | Well sir...... You are right, and I was wrong. | | I did not have the option checked 'ident declared but never used' | | Once I enabled that one... I got the same warnings you did. | | All of those unused variables are now commented-out and a fresh compile done | | Guess what. | | It does not make bit of difference in the finished EXE | | Comparing files in... | E:\ARACHNE\SRC\INSIGHT | with files in... | I:\ARACHNE | | Comparing INSIGHT.EXE with INSIGHT.EXE - Compare failure. | Offset 8b14h Source = 1bh Destination = ah | Offset 8b15h Source = bh Destination = 9h | 2 Mismatches - No more mismatches were present. | | As we see.... the only differnce is the date of complle. | | | Volume in drive E is MCCORKLE_03 | Directory of E:\ARACHNE\SRC\INSIGHT | | INSIGHT EXE 79360 11-27-04 2:30a | 1 File(s) 19472384 bytes free | | Volume in drive I is MCCORKLE_07 | Directory of I:\ARACHNE | | INSIGHT EXE 79360 9-10-04 3:22p | 1 File(s)1095172096 bytes free | | ============================================================================ | | Sat, Nov 27, 2004 - 1:00:14 am | | glennmcc> | There, just did the same with wwwman and got ridd of the 3 warnings. | | | ============================================================================ | | Sat, Nov 27, 2004 - 8:22:58 am | | ray> | Yup, there's a compiler option (can't remember the exact name) that | alows unused code to be ignored, so commenting those variables | won't change the .exe. However, when I'm cleaning I tolerate no | errors or warnings, so they had to go. And, that option isn't perfect -- | it seems to only involve things internal to the module. I've already | found about half a dozen unused functions and commenting them out | has saved c. 1K in the .exe. | | | ============================================================================ | | Sat, Nov 27, 2004 - 10:39:40 am | | ray> | Glenn, | | downloading that file, I end up in a loop -- it downloads then just | starts downloading again. Hitting the big 'X' and then 'F2' | will save the file OK, but what's with the loop? I get a | 'bad command' msg. too, but that's some of my additional error | code in B5. | | | ============================================================================ | | Sat, Nov 27, 2004 - 10:56:23 am | | ray> | Nuts, what I got was a truncated file; also, I can't download anything | else from your site, same using 1.77/3c. | | | ============================================================================ | | Sat, Nov 27, 2004 - 4:17:33 pm | | glennmcc> | If you have these lines enabled in mime.cfg | | application/zip ZIP>TXT|[300]UNZIP -v $1>$2 | application/x-zip ZIP>TXT|[300]UNZIP -v $1>$2 | application/x-zipped ZIP>TXT|[300]UNZIP -v $1>$2 | | And you are still using that cludge 4dos | | You'll need to remove the [300] from each of those lines. | | Alternately, comment-out those 3 lines and this will happen....... | | Save or view downloaded file | Arachne doesn't know how to handle following MIME type: | application/zip | | _______ | | Now you can save it to disk. | | ============================================================================ | | Sat, Nov 27, 2004 - 5:45:42 pm | | ray> | Much better! But it isn't 4dos, I've solved that little problem, | it was that mime couldn't find the unzip program. Don't you | hate it that you get a message "unknown file or command" without the | damned thing telling you WHAT the unknown file or command is? | I wonder if there's a way of making those blasted messages a wee bit | more helpful. Maybe it would be practical to echo the line from | mime.cfg in case of any error -- dunno. | | | ============================================================================ | | Sat, Nov 27, 2004 - 6:30:11 pm | | glennmcc> | Hows come it couldn't find it ? | | Didn't I include both unzip.exe and zip.exe | right in the arachne_main dir of the new install package ??? | | BRB | | | ============================================================================ | | Sat, Nov 27, 2004 - 6:38:23 pm | | glennmcc> | Yep, right there they are in both trhe 'full' and 'lite' packages. | | --a--- 49,286 5-27-04 1:14a h:\a177gplf\unzip.exe | --a--- 86,868 12-24-99 3:40a h:\a177gplf\zip.exe | --a--- 3,095 5-23-04 2:59a h:\a177gplf\zipunzip.txt | | --a--- 49,286 5-27-04 1:14a h:\a177gpll\unzip.exe | --a--- 86,868 12-24-99 3:40a h:\a177gpll\zip.exe | --a--- 3,095 5-23-04 2:59a h:\a177gpll\zipunzip.txt | | | | ============================================================================ | | Sat, Nov 27, 2004 - 7:01:46 pm | | glennmcc> | GFot a little 'fix' for not knowing WHAT the command was. | | application/zip ZIP>TXT|[300]echo "command line = UNZIP.EXE -l -v $1>$2">$2\ | | ============================================================================ | | Sat, Nov 27, 2004 - 7:13:27 pm | | glennmcc> | Let's modify it slightly | | That 1st line results in..... | | "command line = unzip.exe -l -v \arachne\cache\01607744.zip>\arachne\cache\0 | | This is the modified line. | | application/zip ZIP>TXT|[300]echo "command line = UNZIP.EXE -l -v 1>2">$2\nU | | It results in simply ........ | | "command line = unzip.exe -l -v 1>2" | | | That's a little more 'terse' and IMO more readable. | | | ============================================================================ | | Sat, Nov 27, 2004 - 11:09:26 pm | | ray> | My mime lines refer to PKUNZIP so that would be the problem, but | that's probably my fault due to importing a customized mime.cfg | across versions. This is a good, though petty, example of why | I'm concerned for improving our compatablity handling across | versions. | | I like the verbosity above, but how can we make it universal? | Some way of making it routine for mime lines to always echo themselves, | or at least to echo themselves, and then pause, if there's trouble? | | If you check in B5 in main.c you'll see a little bit of checking | that I added, trouble is that 'no such file or command' is the | only error that seems to be bogus, so I've disabled it alone even though | in the case above, that error is pertinent. | | | ============================================================================ | | Sun, Nov 28, 2004 - 10:56:18 am | | ray> | Taking a look at the blasted ".\cache" bug, it seems not to be | a problem in B~5. Can anyone confirm? Without looking at older | code, my guess is that the more organized code in B~5 vis. configuration | options has eliminated the bug, probably because of the fact that | defaults need only be set once. Bugs hate clean code. | | | ============================================================================ | | Sun, Nov 28, 2004 - 2:02:27 pm | | glennmcc> | Confirmed. | | I am running your core.exe right now with 'CachePath .\cache\' | | FTP access is working perfectly. :))) | | Thank you, Ray........... GREAT WORK !!!! | | BTW, | | Free low memory 221024 (green) | | I say again....... GREAT WORK !!!! | | | | | ============================================================================ | | Sun, Nov 28, 2004 - 2:13:28 pm | | glennmcc> | Houston, we have a problem. :( | | F4 no longer works. | | Can you confirm ? | | ============================================================================ | | Sun, Nov 28, 2004 - 3:46:57 pm | | ray> | $#|t, I am getting like Gregy, I did have that problem a while back | and I fixed it without documenting it and I can't remember what | the hell it was. I can't even remember if it was in core or somewhere | else <:-( anyway, I don't have it anymore. | | ============================================================================ | | Sun, Nov 28, 2004 - 5:52:52 pm | | glennmcc> | SHIT ! :( | | | The only thing that occurs to me as being different between my complie | of your SRC code and your compile of the same SRC is that I compiled from | inside the IDE using your core.prj file and IIRC you compiled from the | command line using a make file. | | We did find some differences in the past between the two methods. | | It looks like we have now run into the same thing again. | | BRB after trying a fresh compile from the command line. | ______ | | No go. :( | | This is what happens. | | MAKE Version 3.6 Copyright (c) 1992 Borland International | | Available memory 66836960 bytes | | copy MAKE0001.$$$ core.cfg | 1 File(s) copied | bcc +CORE.CFG -c main.c protocol.c popget.c sendmail.c ftp.c fingophr.c init | mime.c config.c http.c | Borland C++ Version 3.1 Copyright (c) 1992 Borland International | main.c: | Loaded pre-compiled headers. | protocol.c: | Loaded pre-compiled headers. | popget.c: | Loaded pre-compiled headers. | sendmail.c: | Loaded pre-compiled headers. | ftp.c: | Loaded pre-compiled headers. | fingophr.c: | Loaded pre-compiled headers. | init.c: | Loaded pre-compiled headers. | atcpip.c: | Loaded pre-compiled headers. | urlovrl.c: | Loaded pre-compiled headers. | mime.c: | Loaded pre-compiled headers. | config.c: | Loaded pre-compiled headers. | http.c: | Loaded pre-compiled headers. | Error http.c 561: Undefined symbol 'f' in function AuthenticatedHTTP | Error http.c 562: Undefined label 'abort' in function AuthenticatedHTTP | Error http.c 562: Undefined label 'sock_err' in function AuthenticatedHTTP | Warning http.c 562: Function should return a value in function Authenticated | Warning http.c 562: 'willkeepalive' is assigned a value that is never used i | AuthenticatedHTTP | Warning http.c 562: 'line' is declared but never used in function Authentica | Warning http.c 562: 'header_done' is assigned a value that is never used in | AuthenticatedHTTP | Warning http.c 562: 'count' is assigned a value that is never used in functi | Error http.c 563: ) expected | Error http.c 565: Type name expected | Error http.c 565: Too many error or warning messages | *** 6 errors in Compile *** | | Available memory 3619324 | | | | ============================================================================ | | Sun, Nov 28, 2004 - 5:58:45 pm | | glennmcc> | Could you please send me a copy of your core.exe | | This is what my compile of your SRC looks like. | | Volume in drive E is MCCORKLE_03 | Directory of E:\ARACHNE\TEAM\RAY_A\B5 | | CORE.EXE 408992 11-24-04 8:34p | | ============================================================================ | | Sun, Nov 28, 2004 - 8:48:04 pm | URL | | glennmcc> | Houston, we have a major problem at Yahoo. :(( | | Before visiting any other page at Yahoo to login. | | Go to http://my.yahoo.com/to do your login there. | | What happens ? | | In my case....... Complete CRASH to a 'warm reboot' of my computer. | | ============================================================================ | | Mon, Nov 29, 2004 - 5:07:21 am | | Joe> | Ouch!!! | | Yep, I get the same thing, loads (IIRC) about 45K, then just | sits there! :-(( | | BTW, if you use DR-DOS (IIRC Glenn, you use 7.03) beware that | there's a bug in COMP - IIRC, if differences are found in the | first 64K, differences beyond 64K are missed. FC /B seems OK. | | | | ============================================================================ | | Mon, Nov 29, 2004 - 3:36:58 pm | | glennmcc> | Close..... OpenDos v7.01 | | No such problem in this version of comp | | Just tested it by changing 1 byte near the start of the file | and a 2nd byte near the end of the file. | | | | Comparing files in... | L:\ARACHNE | with files in... | L:\ARACHNE | | Comparing ZIP.EXE with ZIP.EXB - Compare failure. | Offset 6bh | The STUC.EXE stub loader is Copyright (C) 1993-1995 DJ Delorie. | The STUB.EXE stub loader is Copyright (C) 1993-1995 DJ Delorie. | -------^ | Offset 14636h | $Id: DJGPP libd built Dec 13 1998\ :14:51 by gcc 2.8.1 $@(#) äl4 !proxy | $Id: DJGPP libc built Dec 13 1998\ :14:51 by gcc 2.8.1 $@(#) äl4 !proxy | ---------------^ | Missing End Of File character | 2 Mismatches - No more mismatches were present. | | ============================================================================ | | Mon, Nov 29, 2004 - 3:40:06 pm | | glennmcc> | Ray, | | You got any thoughts on what might be causing those errors | when I try to compile from command line ? | | ============================================================================ | | Mon, Nov 29, 2004 - 4:28:33 pm | | ray> | Nuts, it looks from the above like something is fouled up in the | uploaded files. For a start, I'll upload the core.exe I'm using now, | with no F4 problems (here anyway). Give that a try, and if it works | OK, then my uploaded zip will need to be replaced. BTW, I don't think | its an IDE vs. MAKE problem since I get identical .exe's with either | one; yes, there was a problem before, but we fixed it, IIRC it was | that 'global optimization' was the default with MAKE but not with | the IDE, so when that's ticked in the IDE, PRJ2MAK will produce a | make file that gives an identical .exe. | | Standby for .exe upload. | | | | | | ============================================================================ | | Mon, Nov 29, 2004 - 4:59:27 pm | | ray> | core.exe zipped, uploaded, downloaded unzipped and tested OK. | Now I'll download the uploaded sources and see what gives there. | | | | ============================================================================ | | Mon, Nov 29, 2004 - 5:06:48 pm | | glennmcc> | On my way to get core.exe righ now. | | BBS | | ============================================================================ | | Mon, Nov 29, 2004 - 5:14:22 pm | | glennmcc> | No good. | | F4 still does the same thing. | | My screen goes black for about 1 seond and then I'm right back viewing | the same page again. | | | | | ============================================================================ | | Mon, Nov 29, 2004 - 5:46:34 pm | | ray> | Nuts, could it be something external to core? Something in mime.cfg | or arachne.cfg? I have a horror of having introduced some bug | that is specific to some environment or other. | | ============================================================================ | | Mon, Nov 29, 2004 - 5:52:56 pm | | ray> | just logged in to yahoo no problem. | Hey, did you delete your old cache.idx and your cookies.log? | IIRC ... nuts was that it with F4 ????? ... think think think ... | Oh, and you do know that you have to delete config.dat if you've made | a change in arachne.cfg? In my experimental directory I just | put a delete line in arachne.bat, in my 'stable' directory I don't | change arachne.cfg hardly ever so I just leave config.dat alone. | | | ============================================================================ | | Mon, Nov 29, 2004 - 6:02:19 pm | | ray> | just unipped and rebuilt from my uploaded source files ... .exe is | 408,992 and works fine, but you got build errors, so ... so ... | bad download? Borland not using the right project file? | | | | ============================================================================ | | Mon, Nov 29, 2004 - 6:19:08 pm | | glennmcc> | Download is fine. (not corrupt, not incoplete) | | It's using your core.prj included in the download. | | My command line attempt to compile used your core.mak | | ______ | | | I tried loading-up and using 4dos.com ..... same problem. | | Copied everything over to my W98 partition and rebooted into MsDos v7.x | (directly to DOS ... not into Win98 and then restarted in DOS mode) | | Same problem. | | ___ | | | I have diagnosed what is happening. | | With all of these in arachne.cfg | | Editor NUL | Editor c:\pedit\pedit.exe %1 | | And with the Editor line deleted from Arachne.cfg | | In all cases F4 shels to DOS and creates a new copy of | c:\temp\arachne.tmp\$roura$.bat | | In all cases this is what's inside that file. | | | NUL C:\HTML\MYPAGES\HOTLIST1.HTM | arachne -r | | | _________ | | | OK, found the 'fix' | | I just now deleted both config.dat and config.log and restarted. | | F4 is now working. | | It looks like you need to re-do the section of code for config.dat | So that it will get overwritten each time we restart. | | The reason I sugfgest this ias that if we go to opt_pers.ah and change | the stuff there (specifically the external editor setting) | | We need to have the correct info placed into config.dat without the need | to exit completely so that arachne.bat will finish running and encounter | the new line which deletes config.dat | | ============================================================================ | | Mon, Nov 29, 2004 - 6:23:59 pm | | glennmcc> | re: "just logged in to yahoo no problem." | | Did you go to http://my.yahoo.com/to do the login ? | | It works fine at groups.yahoo.com and at mail.yahoo.com | | | The crash or lock-up happens if I try going to my.yahoo.com _before_ | I have logged-in at mail or groups. | | DosLynx does not crash or lock-up, but it _does_ show the message | "Invalid HTML" at my.yahoo.com | | | | ============================================================================ | | Mon, Nov 29, 2004 - 6:25:33 pm | | glennmcc> | I'm going to do a Cntrl+Enter on my.yahoo.com and then diagnose exactly | what's in that "Invalid HTML" that's crashing Arachne. | | BBL with the answer. | | | ============================================================================ | | Mon, Nov 29, 2004 - 6:38:45 pm | | glennmcc> | To put it simply....... | | Yahoo employes a bunch of freaking ASSHOLES as 'web page designers' !!!! | | 42 'nested tables' is what's doing it. :(( | | | | ============================================================================ | | Mon, Nov 29, 2004 - 8:25:54 pm | | ray> | re F4: Phew, that's good. | re config.dat: This is very much a questionable 'feature' of the code | as I have it now, the idea was to speed up re-starts, and it does that | nicely (about 5 - 10%) since the entire reading of arachne.cfg can be | bypassed since the structure information is scooped directly from the | image if config.dat. But is 5 - 10% worth the trouble? It's easy | enought to put "del config.dat" in your arachne.bat, but ... | | | | ============================================================================ | | Mon, Nov 29, 2004 - 8:28:14 pm | | glennmcc> | IMO, | | config.dat == bad idea ;-) | | | ============================================================================ | | Mon, Nov 29, 2004 - 8:28:47 pm | | ray> | another option is to put some code in that deletes config.dat | intelegently when needed based on the age of arachne.cfg vs. config.dat | (I do that myself using 4dos commands not available in command.com | batch files). For now, consider it experimental. Oh another | possibly neat thing about it is that one can save a known good | config.dat and restore it in case of trouble. | | | ============================================================================ | | Mon, Nov 29, 2004 - 8:32:47 pm | | ray> | re: http://my.yahoo.com: yup, crash and burn, but did previous versions | do any better? If so, that's a problem. | | re:config.dat: just seeing your comment above, you might be right, | I don't have any great love of this, but I'd like to kick the idea | around a bit more, me I've gotten quite used to it and I've even come | to like the extra startup speed, but I won't kick and scream if | the idea dosen't win any converts -- out the window it goes. | | | ============================================================================ | | Mon, Nov 29, 2004 - 8:35:33 pm | | ray> | Oh, speaking of giving things time, what do you think of config.log? | I can see how this could be of great utility in your help desk work. | Consider: no comments, no double values, all values shown as parsed, | and always in the same order ... once you get used to it, I'd say | that you'd be able to find config bugs in less time than having | folks mail over their arachne.cfg's no? | | | | ============================================================================ | | Mon, Nov 29, 2004 - 8:40:08 pm | | ray> | Oh one more thing about config.dat, it would be very easy to make it's | use optional in the code with a config option UseDat (Y/N) | | ============================================================================ | | Mon, Nov 29, 2004 - 9:11:23 pm | | glennmcc> | re:"one can save a known good config.dat and restore it in case of trouble." | | In-order to this. | | We must | 1) exit Arachne | 2) copy config.dat to 041129.dat | 3) restart Arachne | | In case of trouble | | 1) exit Arachne | 2) copy 041129.dat to config.dat | 3) restart Arachne | | ------ | | We already have the ability to save a known good state and to restore | it when ever we need to. (without exiting) | | 1) go to the dial page and click on 'profile' | 2) save the current arachne.cfg as 041129.acf | | In case of trouble | 1) go to the dial page and click on "ACF profiles" | 2) link to 041129.acf | | Done. | | We are now back to that known good state without ever exiting. | | | ============================================================================ | | Mon, Nov 29, 2004 - 9:15:54 pm | | glennmcc> | In addition... since chanhges to arachne.cfg will not get used untill we | exit Arachne, delete config.dat and restart Arachne with the altered | Arachne.cfg | | We can no longer make config chages from with in the various setup pages | and have them take effect by hitting the "use new settings' button. | | We also can not go to the dial page, click "ACF profiles' and link to | one of our saved AXCF files and have those settings get used. | | Not till we exit Arachne, delete config.dat and restart Arachne. | | So I'll stick to my above opinion. | | config.dat == bad idea. | | | ============================================================================ | | Mon, Nov 29, 2004 - 9:58:42 pm | | ray> | First point: quite right, the saving/restoring would be logicaly | identical. | | Second point, not quite so bad as that since 'use new settings' does | a complete resart which would delete config.dat automaticaly if a line | to do that is contained in arachne.bat or if time stamps were compared | in the .exe itself and config.dat deleted in any case where it is | older than arachne.cfg -- anyway, you may well be right, the | idea might just wither on the vine due to the simple fact that the | savings in startup time simply don't justify the incompatablity | or the possible complication in use. I just want to let the | idea simmer for a while. To tell the truth, I was expecting a much | better improvement in start time, if I'd known that it would be so | little, I would never have bothered with it. | | | ============================================================================ | | Mon, Nov 29, 2004 - 10:00:14 pm | | ray> | Beddie buy time. | | | ============================================================================ | | Mon, Nov 29, 2004 - 10:34:49 pm | | glennmcc> | G'nite Ray | | CU tomorow | | | ============================================================================ | | Wed, Dec 01, 2004 - 2:33:45 am | | Joe> | Rise and shine! (actually, it's getting late for me here!;-) | | Glenn, the details of the DR-DOS COMP bug I found are a bit hazy | now, since I no longer trust COMP and simply don't use it. However, | brief testing suggests the following : If there are more than 20 | differences (ie. 21+) in the first 64K of a file, then differences | beyond 64K are missed. This is even with the "/m:0" option! :-( | | | ============================================================================ | | Wed, Dec 01, 2004 - 8:05:29 am | | glennmcc> | Tested and confirmed. :( | | Where do I get FC ? ;-) | | | ============================================================================ | | Wed, Dec 01, 2004 - 8:11:53 am | | glennmcc> | Duh, nevermind.... it's in c:\opendos | | Thanks for the heads-up. | | Just tried FC and while it does not have that bug. It does not tell us | the total number of differences like comp does. | | | | ============================================================================ | | Wed, Dec 01, 2004 - 6:36:44 pm | | ray> | In case you missed my post to Vladimir, note the bug in | guidraw.c -- I improperly commented out a justification line. | Interesting that it has no effect on the funcion in which it occurs, | but does foul up another function. | | | ============================================================================ | | Thu, Dec 02, 2004 - 7:23:34 pm | | ray> | Glenn, | | I've got a question about insight that's fundamental. OK, so the | program's job basicaly is to create an HTML page for display under | core. But it seems to do the whole thing using 'printf's and I don't | understand how this creates anything like a file that can be | read as HTML -- it should just print the stuff to the screen, no? | Obviously 'no' but I don't see how it works. Can you explain? | | | ============================================================================ | | Thu, Dec 02, 2004 - 8:34:18 pm | | glennmcc> | | Volume in drive L is MS-RAMDRIVE | Directory of L:\ARACHNE\CACHE | | . 12-02-04 9:42p | .. 12-02-04 9:42p | HEADERS 12-02-04 9:42p | 02041775 HTM 163 12-02-04 9:42p | 02041809 HTM 4686 12-02-04 9:42p | $IDX$CNM IDX 0 12-02-04 10:24p | $SRT$CNM IDX 0 12-02-04 10:24p | 02041824 HTM 13032 12-02-04 9:43p | 02041870 HTM 11745 12-02-04 9:43p | 02041907 HTM 29436 12-02-04 9:44p | 02041955 JPG 4295 12-02-04 9:44p | 02041955 BMP 81414 12-02-04 9:44p | 02044242 HTM 6240 12-02-04 10:23p | 02042030 HTM 4595 12-02-04 10:01p | 02044363 HTM 6406 12-02-04 10:25p | 02044382 HTM 12166 12-02-04 10:26p | 02044399 HTM 14302 12-02-04 10:26p | 02044429 HTM 9556 12-02-04 10:26p | 18 File(s) 98103296 bytes free | | | Most of those are the reulting .HTMs created by this line in mime.cfg | | file/.cnm >HTM|[115]$einsight.exe $1 -cache=$t>$2 | | If we simply wrote the data to the screen, none of thse files would exist. | | And now that I'm thinking about it. | Perhaps that actually would be better. | | We don't save any of the .HTMs so why create them in the first place ? | | Things that make you go... hmmmmmmmmmmmmm | | | ============================================================================ | | Thu, Dec 02, 2004 - 9:40:35 pm | | ray> | Careful Glenn, we're thinking the same thing again ;-) | | But this again begs the question: how are these files created? | Where and how are the printf() calls redirected from the screen | into files? I know that the '$1' is the name of the file to | be written to, but don't know how the input gets written there. | | | ============================================================================ | | Fri, Dec 03, 2004 - 4:36:30 pm | | glennmcc> | I knopw... mime.cfg is quite confusing. | | file/.cnm >HTM|[115]$einsight.exe $1 -cache=$t>$2 | | >HTM == redirect output into a file with .HTM as extension. | | Therfore the line 'pipes' $1 through insight.exe and | redirects the output to >$2 in our cache dir with .HTM as extension. | | Now... is that as clear as mud ? | | ============================================================================ | | Fri, Dec 03, 2004 - 5:18:23 pm | | ray> | Ah yes, I see it now, the piping is controled at the batch level. | Hmmmm, the whole contents of each mail page created, piped, and | then read ... this seems ponderous and slow. And as I understand it, | the whole thing is redone for every new mail page. It's too early to | say yet, but my first impression is that there's got to be a better | way. Why can't there be a permanently existing mail page display | file, that's just modified as needed with updated contents? If | I can integrate insight into core then the whole thing could be | done internaly which should be nice, but we'll see. | | | | ============================================================================ | | Sat, Dec 04, 2004 - 12:18:59 pm | | ray> | Nuts, trying to set my mother up with Arachne to ween her off | Outlook: I get as far as a connection (PPP.LOG shows "Success : IP | address set to" ...). the online timer ticks away but I get | "Cannot find" for every thing I try to access. Any thoughts? | | | ============================================================================ | | Sat, Dec 04, 2004 - 2:48:36 pm | | glennmcc> | Just now replied to your email on the list. | | Sounds like a problem with DNSs | | ============================================================================ | | Sat, Dec 04, 2004 - 4:43:21 pm | | ray> | Yup! On line now on Ma's computer using her ISP. | Thanks Glenn, your the greatest. | | | | ============================================================================ | | Sun, Dec 05, 2004 - 9:03:04 pm | | ray> | Ma's on her way to being another happy Arachne user. I just had to | wipe her hardisk and re-install W98 becuase it had become infested | with spyware, viruses, etc. I told her that there was basicaly only | one solution to her need to connect to the internet without getting | infected ... | | | ============================================================================ | | Sun, Dec 05, 2004 - 9:18:09 pm | | ray> | Hey Glenn, you know on the title bar how there has been concern that | if the 'Arachne side' is too long it can interfere with the title? | I'll bet that that could be solved by simply writing the Arachne | first, then the Title, so that it could overwrite the former if needed, | no? When ever the screen is repainted, the 'Arachne GPL' (or whatever) | would be redrawn, so it would only be temporarily gone. Wadaya think? | | | ============================================================================ | | Mon, Dec 06, 2004 - 2:00:59 am | | Joe> | Hi folks, | | FWIW, after running the "download speed" and "ETA" text as | ... (1234 of 67890 @ 1234 bytes/sec - time remaining 0:1:2), | I have changed the '@' and '-' both to '---', this seems | to be the most aesthetic version of this text so far (well, | at least for me). | | BTW, I am sitting at a friend's W98 machine (using IE) and | trying to sort out a variety of strange gremlins. | | | ============================================================================ | | Mon, Dec 06, 2004 - 2:05:11 am | | Joe> | Ray, | | Does that mean your Mom is able to forward an e-mail back | to me using Outlook, so I can get that Bas64 problem looked | at again? I wonder, if she uses a different ISP, if the same | problem occurs? | | Also, why aren't you using LSPPP with LSPPPCFG for your Mom, | such that you don't need to configure the DNS settings? | | Finally, your "Arachne string" idea is interesting ... | | ============================================================================ | | Mon, Dec 06, 2004 - 5:25:37 pm | | ray> | Joe, | | Strange you should mention that -- I was thinking the same thing | myself, since now that I have actualy used Outlook for the first | time, I suppose I have no more excuse for not giving you a hand with | that. Send me a file and I'll send it back via Outlook. | | As for LSPPPCFG, I did use it, but since DNS numbers were required, | but I specified 0.0.0.0, there was little hope of it working, or is | that not correct? BTW | speaking of LSPCFG, (or rather PPP-CFG?) there was one anoyance, | it kept overwritting arachne.cfg with arachne.lsp(?) until I deleted | the latter. | | Re 'Arachne string' yeah, I haven't looked into it, but from what I've | learned about 'guidraw.c' it's just a question of which string | is drawn first. | | | ============================================================================ | | Mon, Dec 06, 2004 - 5:30:49 pm | | ray> | Joe, | | But, but, but ... if you send me an email and I send it back as | an attachment, doesn't that guarantee that the contents will be | unaltered regardless of which program sends it? I can see that | if it was returned as an email (like 'reply to sender') then the | contents are re-worked, but surely an attachment is just a binary | file that won't be changed, no? ... Unless the whole point of the | exercise is to see how the program decodes it ... but then I'd have | to import it into Outlook, as what HTML? and then send it, no? | Please advise. | | | | ============================================================================ | | Mon, Dec 06, 2004 - 5:33:55 pm | | ray> | OH, one more thing, Joe, Glenn don't like it, but I am very happy | giving the download line minimum field widths of 5 or 6 -- this | keeps the line from jumping about too much with smaller files, but | still alows the line to grow with big files, what's your oppinion? | | | | ============================================================================ | | Mon, Dec 06, 2004 - 5:59:02 pm | | Joe> | Hi Ray, et. al. | | WRT my e-mail problem - remember, I don't actually send the message | as a binary attachment, this just seems to be one manifestation of | how it gets decoded at the other end (ie. your end) sometimes. | | A "forward" is the least likely method to modify what you've received | and perhaps give me/us here an insight (pun intended;-) of how things | arrive at your end. Unlike "reply", I don't think HTML encoding by | your e-mail client and similar nonsense will come into play. | | Regarding LSPPP/LSPPPCFG and DNS ... the 0.0.0.0 setting with BOOTP | enables automatic DNS negotiation with your ISP (works in 99% of | cases), so even if you need definitive settings, these are handled | automagically for you. | | Regarding 'arachne.lsc', you should not have deleted that. Simply, | the CFG-PPP utility is used only once, to install LSPPPCFG, or once | again to uninstall. So, if it was overwriting 'arachne.cfg', you | must have been uninstalling - but why? Surely once installed and | working happily, what do you gain by unistalling? Unless you want | to go back to EPPPD or try some other packet driver (eg. LAN), | there should be no reason to, in which case 'arachne.cfg' will | not be overwritten by 'arachne.lsc'. (I hope that makes sense.) | | | ============================================================================ | | Mon, Dec 06, 2004 - 6:09:40 pm | | Joe> | Oh, about the minimum width thing - for small files, typically | images on a web page, these often just "fly past", particularly | when "multitasking" is enabled, with Arachne downloaded several | of these in "parallel". So it is impossible to follow the old | "downloaded X of Y bytes" stuff, let alone the extra "bytes/sec" | and "ETA" stuff, regardless of minimum width settings. | | In other words, I don't see much point trying to achieve fixed | formatting to benefit the small stuff, it's only the big stuff | that's really relevant anyway, so _if_ minimum widths compromise | the aesthetics (I haven't tried minimum widths as yet), then they | aren't worth it. | | | | ============================================================================ | | Mon, Dec 06, 2004 - 9:59:20 pm | | ray> | Joe, | | OK, I'll fire up Outlook, copy the .CNM to whatever extension Outlook | uses so that it finds the message and forward it back to you, if | you don't get it pronto, it didn't work. | | Re: cfg-ppp. It was one of those situations where things weren't | working and I'm trying this and that and God knows what I did. | | If running cfg-ppp a second time uninstalls, then that's probably what | happened, but I'm quite sure that things didn't work with 0.0.0.0 -- | but I wouldn't bet my life on it, I'll try it again next time I see | her machine. | | Re: formating -- I like my minimum field widths, but | to each his own. I see your point about small files, its the ones of | about 10K - 100K where I'd say the minimum field width is nice. But | these issues don't mean much to me either way. | | Outlook, here I come ;( | | | | | ============================================================================ | | Mon, Dec 06, 2004 - 10:29:38 pm | | ray> | Outlook, there I go :( ... it seems my copy is very upset with me | because I've never used it before -- it want's to be reinsalled and | then it will need to phone Bill probably. I'll get to it manyana. | Damned thing. | | | ============================================================================ | | Tue, Dec 07, 2004 - 2:42:10 pm | | ray> | Once more into the breach. | | | | ============================================================================ | | Tue, Dec 07, 2004 - 3:46:25 pm | | ray> | Joe, | | OK, I imported the message as text and sent it back to you. If that's | not right, let me know and I'll send it as an attachment. BTW, my | windows text editor didn't like the file -- reported that it | contained null characters! Maybe that's where the problem is coming | from. I opened a whole fistful of other emails as text, and none of | them had that comment. | | ============================================================================ | | Tue, Dec 07, 2004 - 5:28:06 pm | | little_birdie> | FYI, | | It is 'unto' ... not 'into' | | http://www.liebreich.com/LDC/HTML/Opinion/Iraq/Breach.html | | ============================================================================ | | Tue, Dec 07, 2004 - 6:09:58 pm | | ray> | Touche, ... thou rump red runion. | | | | ============================================================================ | | Tue, Dec 07, 2004 - 6:50:36 pm | | little_birdie> | Just thou waitheth there, thou rump fed runion, | and I'll skewer thee and thy mates as the great Poofs that ye bee. | | ============================================================================ | | Tue, Dec 07, 2004 - 7:01:16 pm | | little_birdie> | Or perhaps you meant to quote this line from Macbeth. | | Aroint thee, witch!’ the rump-fed ronyon cries. | | | ============================================================================ | | Tue, Dec 07, 2004 - 8:45:59 pm | | ray> | Oh worthy thane, hence forth, be Cawdor! | | Yes, it was Macbeth ... Hoarse is the raven that croaks the entrance | of Duncan beneth my batlements ... but in truth I only have | two soliloquies from that play actualy memorized, the rest is | recalled with some imperfection, as you noted -- but well done, | it's good to see someone taking his Shakespeare seriously. | | Is this a dagger I see before me, the handle to my hand? | Come, let me clutch thee. I have the not, but see thee still ... | | God, I love it. | | Worthy Joe, | | Recievest thou that returned e-mail? | | Alas, poor Ray, I knew him, Horatio, a ... well, I'll probably | get that wrong too. | | | ============================================================================ | | Tue, Dec 07, 2004 - 8:49:14 pm | | ray> | Nuts, I am loosing it, that should be: | | Is this a dagger I see before me, the handle toward my hand? | | | | ============================================================================ | | Tue, Dec 07, 2004 - 9:38:31 pm | | glennmcc> | Still my favorite........ "Cry havoc, and let slip the dogs of war !!!" | | ============================================================================ | | Tue, Dec 07, 2004 - 10:10:31 pm | | ray> | A worthy choice, and from my favorite play, too. | | | | ============================================================================ | | Wed, Dec 08, 2004 - 5:29:27 am | | phantom> | Thanks for the e-mail Ray, I've passed it on to our IT people. | If you can forward it as an attachment too, that may also help | sort out this stupid problem. | | Regarding 0.0.0.0, don't forget you need to use the BOOTP method, | although that's the default configuration that CFG-PPP gets you | anyway, so unless you change that, it should work nicely. | | Re. Shakespeare, sorry but ... | God, I hate it. | Elizabethan melodrama, IMHO, no better than present day soapies, | except the language is archaean in the extreme. | | My English teacher was French (with a horrible accent that I | wouldn't say resembled English) and she just LOVED Shakespeare; | we had so much of it forced down our throats it made us sick! | Not that I liked Shakespeare before then, just to a different | degree. (It still gives me nightmares!;-) | | | ============================================================================ | | Wed, Dec 08, 2004 - 7:52:10 am | | phantom> | PS. Tip: Ray, the CFG-PPP "update" option will often fix a broken | LSPPPCFG installation, so may be worth trying if appropriate ... | | | ============================================================================ | | Wed, Dec 08, 2004 - 5:39:53 pm | | ray> | Joe, | | OK, I'll do it again as an attachment. When I see my mum again | I'll retry the CFG-PPP thing again. Now that I have a known | good instalation it should be easy enough to figure out what went | wrong last time. | | Too bad about the Skakespeare -- force feeding is no way to develop | a taste for anything, and IMHO school kids are too young to | apreciate the Bard anyway. But soap it aint! | | | | | ============================================================================ | | Wed, Dec 08, 2004 - 10:49:41 pm | | glennmcc> | Ahhh.... the play's the thing ! | | | | ============================================================================ | | Thu, Dec 09, 2004 - 4:06:29 pm | | ray> | Yes, and the sonnet's the thing as well -- the best poetry ever | written. | | | ============================================================================ | | Thu, Dec 09, 2004 - 8:28:36 pm | | ray> | Glenn, | | How do we access the insight help? | | | ============================================================================ | | Thu, Dec 09, 2004 - 9:17:49 pm | | glennmcc> | As with most DOS programs. | | At the DOS command prompt......... | | insight | | Same applies to wwwman | | wwwman | | ============================================================================ | | Thu, Dec 09, 2004 - 9:51:07 pm | | ray> | Sure, I just thought there would be a DGI to do this internaly, since | neither insight nor wwwman are really meant to be run as stand alone | programs. Infact, given that both of them are only accessed via. | DGI, the hard coded help information seems a bit redundant, and should | IMO be simply moved into a help file. | | | ============================================================================ | | Fri, Dec 10, 2004 - 12:13:43 am | | Joe> | Hi all, | | Ray, I got your re-forwarded e-mail for the "Base64 problem" test, | but there was no attachment. :-/ | | | ============================================================================ | | Fri, Dec 10, 2004 - 12:15:59 am | | Joe> | Re "help", | | I'm not sure what sort of help would be required with wwwman, | but does the F1 key do anything useful in insight ... ? | | ============================================================================ | | Fri, Dec 10, 2004 - 4:13:38 pm | | ray> | Joe, | | No attachment, eh? Well, I couldn't figure out how to do it -- Outlook | only lets you make an attachment from certain folders, except that | you don't know the 'real' names of the folders (only Micro$soft | could think of something that dumb) so I couldn't find the right | way of moving the file into the folder from which to make the | attachment -- and I was ready to kill after trying. But when I have | the stomach for it, I'll try again. | | Re insight help, no F1 gives no help on insight or at least it | doesn't access the coded-in help material, but that information | is only of use to mime programmers anyway, so I'd say there's no | point in building it into insight.exe, it should be in a file | by itself, or it should be part of the main Arachne help system | or it should be build into mime.cfg's internal documentation. | | | ============================================================================ | | Fri, Dec 10, 2004 - 4:48:31 pm | | glennmcc> | DGI... no problem. | | Add this line to the top of mime.cfg | | file/inshelp.dgi >TXT|$einsight.exe>$2 | | Now type file://inshelp.dgi into the URL line and press enter. | | | ============================================================================ | | Fri, Dec 10, 2004 - 5:29:50 pm | | ray> | Glenn, | | Yeah, that's about what I was fishing for ;-) ... but I'd still | say that that info. belongs in the documentation section of 'mime.cfg' | itself, since insight is *only* ever called via MIME, no? | | | Joe, | I just now got around to including your and Michal's reworking | of ctype.h thanks to Vladimir having trouble with his Russian | toolbar. I must say you've done an impecable job in the file. | BTW, as far as I'm concerned, the file with be included in the | distribution, and will be called via: | | #include "ctype!.h" so that whether or not anyone has this file | in their Borland copy, the right file will always be used. | | | ============================================================================ | | Fri, Dec 10, 2004 - 6:51:38 pm | | glennmcc> | Ya mean like this ? | | | | ; \n ... end of line (note: this won't work for "|[nn]xx" style commands!) | ; %% ... % character | ; $$ ... $ character | | ; command line options of Insight.exe | | ;Insight 3.85.8, Arachne mail maintenance tool | ;GNUpyright (G)1997-2000 Arachne Labs.(now being maintained by glennmcc) | ;INSIGHT converts mailmsg to HTML | ;INSIGHT -a insert attachments to outgoing mailmsg | ;INSIGHT -o converts mailmsg to HTML, editable | ;INSIGHT -i converts maildirectory to HTML | ;INSIGHT -foldes list folders in mail directory | ;INSIGHT -r replies to msg | ;INSIGHT -d deletes msg | ;INSIGHT -1 move maili phase 1 | ;INSIGHT -2 move maili phase 1 | ;INSIGHT -e resends msg | ;INSIGHT -f forward | ;Program writes to STDOUT; syntax is INSIGHT [options] > file | | | ============================================================================ | | Fri, Dec 10, 2004 - 8:06:34 pm | | ray> | Yep, exactly like that ;-) | With that info where it belongs, insight.exe can be trimmed of | all that. And now, of course, the information can be edited or | anotated by users at will without anyone having to re compile an .exe. | | | | | ============================================================================ | | Fri, Dec 10, 2004 - 8:13:26 pm | | ray> | BTW Glenn, are you still hacking around with B~5? I'll feel better | about it when I know that you have given it a vigorous shakedown | for bugs, etc. Vladimir has already found two bugs. OH and for | what it's worth I don't expect to ever sell you on 'config.dat' | but I've made it transparent -- the file is automaticaly updated | if it's timestamp is older than arachne.cfg, which should be | quite robust. If there is still any trouble, I'll kill the idea | myself. | | | ============================================================================ | | Fri, Dec 10, 2004 - 8:34:31 pm | | glennmcc> | How's that old saying go ? | | Give him enough rope and he'll hang himself. | | Keep checking it out and I think you'll find that my assessment was correct. | | config.dat == bad idea | | ============================================================================ | | Sat, Dec 11, 2004 - 10:37:51 am | | ray> | Nope, no danger of a hanging here, the config.dat idea is | a trivial appendage to the code, that can be disabled with about | two or three commentings ... hmmmm, it just occured to me, you do | know how to disable it don't ya? Just find the lines in 'init.c' | that boil down to: | | if exist conifig.dat used config.dat | else ConfigureUserInterface(); | | and at the very bottom of config.c find the code that writes the | file. Disable or remove as the spirit moves you. But as for me, | if it is a bad idea as you think, I'd be more convinced by reasons | than by assertions. Conseptualy I see it as being anologous to | 'arachne.pck' -- a file where Arachne stores stuff for it's own | use on restarting. If it can be made trouble free (if!) then it has | the small advantage of speeding up reloads, while having no real | negative. Or that's how it looks to me. But, as I said, cut it out | if you don't like it. | | | | ============================================================================ | | Sat, Dec 11, 2004 - 10:00:58 pm | | glennmcc> | snip, snip, cut, cut ...... tossed into trash can. | | ============================================================================ | | Sun, Dec 12, 2004 - 9:11:38 pm | | ray> | ... and you'll save about 400 bytes, too! Right now I'm trying to convince | Vladimir that config.dat is a good thing; if that fails, then | I'm right behind you -- out the winder goes the code, out the winder | goes the code (to the tune of a song in 'Paint Your Wagon'). But | still, it was interesting to learn just how fast and easy you can | save and restore the contents of a structure, I suspect that | this kind of thing will be useful somewhere. | | | | ============================================================================ | | Mon, Dec 13, 2004 - 12:56:29 am | | Joe> | Ray, | | I've incorporated some code in 1.77J1 and 1.77C1 that checks for | an updated 'ctype.h' and terminates with an error message if it | doesn't detect the updated version. Have a look in 'posix.h' ... | | | ============================================================================ | | Mon, Dec 13, 2004 - 6:33:07 pm | | ray> | Joe, | | Cool, I'll check that out. Hey that makes me think: is there | some file there that documents any code changes over and above | what I've had from you up to 3C (which was up to J4 IIRC)? I'd | like to keep current with whatever you've been up to. | | | ============================================================================ | | Tue, Dec 14, 2004 - 12:49:31 am | | Joe> | Hi Ray, | | Ummm ... can't remember (with certainty) :-/ | | There might be some minor stuff, I've been promising for ages to | review what I've done and do a proper 'changes.txt' update, but | in the little time I have available, I tend to focus more on | hacking the code. You will find some interesting (and hopefully | useful) stuff in 1.77J2/1.77C2, which is progressing nicely ... | | | ============================================================================ | | Tue, Dec 14, 2004 - 3:45:49 pm | | ray> | Joe, | | Ok, when I've finished with insight and am ready to get back into | core, I'll download your latest and see what you've got. BTW I have | a very broad conseptual difficulty with insight and I'd like your | opinion of it: it seems rather strange to me that the program | is (or seems) designed to rewrite an HTML page from scratch for every | new insight page shown. Wouldn't it make more sense to have permanent | 'specimins' of each type of page and just 'fill in the blanks' as it | were to show the current information? Not only is the current method | apparently needlessly complex, but each new insight page becomes | a cache item even though the previous pages are, obviously then | useless. Any thoughts? | | | | ============================================================================ | | Tue, Dec 14, 2004 - 11:22:09 pm | | Joe> | Hi Ray, | | Firstly, I don't use Insight. Secondly, I'm no HTML expert. | | Bearing that in mind, I've only ever seen web pages importing graphical | objects (for displaying on the page), not text objects, which is what I | think you're proposing, right? So I'm not sure what you're suggesting | is actually possible. | | Perhaps the way to avoid writing HTML pages and then invoking Arachne | to display them, would be to add a GUI (based on Hixlov/Lopif, I'd | guess) to Insight, and have this take care of that stuff. Sounds like | a fairly major task ... is Insight too slow with the way it presently | works? | | ============================================================================ | | Wed, Dec 15, 2004 - 5:56:07 pm | | ray> | Joe, | | No, you missunderstand me, I realize that Arachne will display | the mail pages via HTML (though I like your thoughts on using | Lopif directly but suspect that you are right -- it would be alot | of work) but what seems strange to me is that the pages are | completely re-written from scratch every time. What I'd expect | is that only the 'contents' would need to be reworked each time. | Given that insight exists as an appendage to Arachne, maybe I'm right, | it's just that durring the development phase of the .exe, total | re-writes made things easier to debug -- I dunno. | | | ============================================================================ | | Wed, Dec 15, 2004 - 10:53:59 pm | | Joe> | Hi Ray, | | Perhaps I have misunderstood here (given the caveats), however | I'm not so sure. As I understand it, you want a standard web page | template that imports dynamic text content, which is the e-mail | message, I expect. But I've never seen a web page "import" text | content, only image content, so I'm not sure HTML can actually do | this (again, I'm no expert on HTML). | | | | ============================================================================ | | Thu, Dec 16, 2004 - 6:20:29 pm | | ray> | Joe, | | Ok thanks for that, I'm even less of an HTML expert, anyway | these are just very preliminary questions, its way to early to | even think of doing any such thing. But I'd like to understand | this stuff. I quess I'm going to have to learn HTML one of these | days. | | | ============================================================================ | | Wed, Dec 22, 2004 - 8:38:27 pm | | ray> | Glenn, | | Negative on 'nul' vs. 'NUL' -- same result: stdout is redirected ok | but stderr still goes to screen. | | ============================================================================ | | Wed, Dec 22, 2004 - 9:54:09 pm | URL | | glennmcc> | Well then.... I guess your only option is to install a real OS. | | http://custom.lab.unb.br/pub/os/dos/OpenDOS.701/OD701.EXE | | | | ============================================================================ | | Wed, Dec 22, 2004 - 10:13:26 pm | | ray> | Huh, so opendos is smart enough to redirect STDERR as well? | Hmmm, I've been using M$dos 7.10 for no other reason than it seems | harmless and it will alow me to run '98 when needed, but it is | time I had a look at opendos. Will it read FAT32? If not, it goes | into my DRDOS partition. | | | | ============================================================================ | | Thu, Dec 23, 2004 - 4:28:51 pm | | glennmcc> | Yep, she can access FAT32 when you load DRFAT32.SYS | | ============================================================================ | | Thu, Dec 23, 2004 - 5:07:13 pm | | ray> | Beauty! New OS comin' up. | | BTW, I tried to upload my screenshot last night but Yahoo regected it | as being to big but I've since realized there's not much point -- | it looks fine but I think the reason is that I have 'silver' | as my background color for everything. How do I change my colors | to the same as yours? I think that will cause me to see the error. | | | | ============================================================================ | | Fri, Dec 24, 2004 - 4:39:09 pm | URL | | glennmcc> | Merry Christmas everyone. | | ============================================================================ | | Fri, Dec 24, 2004 - 5:09:32 pm | | ray> | Yes, Merry Christmas to us! | | BTW Glenn, what about those colors? I'll get a lump of coal | in my stocking if I don't duplicate that CSS error ;-) | | | ============================================================================ | | Fri, Dec 24, 2004 - 7:05:42 pm | | glennmcc> | Here are my color settings in arachne.cfg | | BgColor #000000 | Text #FFFFFF | Link #00FF00 | Background NUL | HTMLtext #000000 | HTMLbgColor #C4C4C4 | HTMLlink #0000C4 | CSS Yes | MailIsDesktop Yes | | | ============================================================================ | | Fri, Dec 24, 2004 - 7:07:52 pm | | glennmcc> | BTW, | | Did you 'click the link' in my Xmas greeting above ? | | That's where to find everone's gift that'll be there | ready to unwrap in the morning. | | ============================================================================ | | Fri, Dec 24, 2004 - 9:21:30 pm | | ray> | ... sure did! | | | ============================================================================ | | Sat, Dec 25, 2004 - 9:27:27 am | | ray> | Has Santa come? | | | ============================================================================ | | Sun, Dec 26, 2004 - 5:18:57 pm | | ray> | Yes, he has. Thank you Santa. I'd have left some milk and cookies, | but that's difficult over the internet ;-) | | | | ============================================================================ | | Sun, Dec 26, 2004 - 9:09:52 pm | | ray> | Glenn, | | OK, some progress: now that I know what to look for, I see that | my 'pppglenn.htm' isn't displaying right -- I wasn't aware that the | 'Save Changes' button was supposed to be on a different colored | background. That's a huge relief, now I can start tracking it down. | | | | ============================================================================ | | Mon, Dec 27, 2004 - 3:30:38 am | | phantom> | "Yep, she can access FAT32 when you load DRFAT32.SYS" | | BTW, there's a unofficial OpenDOS project based on the 7.01 | sources that provides native FAT32 (and LFN?) support, no | drivers needed. Can't remember the URL (I can check when I | return to work next year, if required). | | Merry Xmas. | | | | ============================================================================ | | Mon, Dec 27, 2004 - 10:51:47 am | | ray> | Merry Christmas Joe. | | Yeah, native FAT32 would be just the thing with OpenDOS. In a world | where the smallest disk you can buy is 40G, not being able to handle | FAT32 makes an OS seem very dated. | | | ============================================================================ | | Mon, Dec 27, 2004 - 4:19:53 pm | | glennmcc> | 40Gs... no problem. | | 20 2Gig partitions. ;-) | | ============================================================================ | | Mon, Dec 27, 2004 - 6:30:18 pm | | ray> | Yes, but that's with the *smallest* disk you can buy. In a few years | the 2G partition just won't hack it any more. The world doesn't | exactly hang on my opinion, but I like the idea of DOS keeping | right up to date with the latest hardware, after all, DOS is the | OS of the future, right? | | | | ============================================================================ | | Tue, Dec 28, 2004 - 12:54:14 am | | glennmcc> | IMO, | | DOS is the OS of the present. | | Linux is the OS of the future. | | ============================================================================ | | Tue, Dec 28, 2004 - 3:47:05 am | | phantom> | Well, I haven't used Linux in many months; lots of little irritating | problems just take soooo much effort to track down. Initially installed | Mandrake, things simply didn't work out-of-the-box at all, and it took | soooo much effort to get the video and sound working, even then the | performance was unimpressive. Then installed Red Hat instead, however | this proved even worse. | | My plan when time permits (yeah, I know, it's a perennial problem)-: | is to give Free-BSD a go, using it's "Linux compatibility" option. I | hope this will be much less opaque than Linux, so it should be much | easier to configure and get things running properly. | | As for OpenDOS ... there's also DR-DOS 8.0, which has native FAT32 | support. | | BTW, ROM-DOS 7.10 (based on MS-DOS 7.10, AFAIK) has native FAT32 | support too, so probably has PTS-DOS (is this still available?). | Finally, there's FreeDOS, which I'm sure has native FAT32 support. | | | ============================================================================ | | Tue, Dec 28, 2004 - 9:40:54 am | | ray> | Joe, thanks for all that, I've got to check up on some of those. | BTW, my experience with Linux has been similar to yours. I've tried | to make the switch twice, and both times I've come running home to DOS | with a bloody nose. I'm sure if I was running a server with a million | users, or needed 15 passwords or had a machine with 64 processors or | had some other huge computing need, I'd be running Linux, but as it | is DOS is faster and simpler and I hope it never dies. I'd like it | to live on beside Linux as the 'lite' OS. (Sheesh, I remember trying to | desipher Linux's configuration files -- there were dozens of them | scattered all over the place. Bash -- what a monster! It reminds me | of an old 'Chalenger' steam locomotive, yeah it will pull a train | 10 miles long over the Rockies, but friendly it aint! | | | | | ============================================================================ | | Tue, Dec 28, 2004 - 5:29:12 pm | | glennmcc> | Ever try Slackware ? | | Got v7 on this machine right now. | | Got Caldera OpenLinux on the old machine. | | ============================================================================ | | Tue, Dec 28, 2004 - 6:11:37 pm | | ray> | Nope, like Joe, I tried one of the more mainsteam distros. | But I've heard that Slackware has a small but fanatical group of | devotees ... hmmm sounds like a DOS browser that I'm fond off ... | Anyway, I understand that Slackware cuts out most of the bloatware | that can be as much of a problem in Linux as it is in you-know-where. | (Now, once there is such a thing as '4Linux', I'm sold ) | | ============================================================================ | | Tue, Dec 28, 2004 - 6:34:25 pm | | glennmcc> | Unfortunately, it would probably be just as big a cludge as 4DOS | | | But of coarse there are several CLI 'shells' to choose from in Linux. | | Try them all..... you're bound to like at-least one of 'em. | | And since Linux itself and litterally millions of programs for it | are available 100% free...... we got no $ to loose no matter how many | differnt things we try. | | | ============================================================================ | | Tue, Dec 28, 2004 - 8:09:33 pm | | ray> | ha ha. | | Yeah, I tried a few but they were all variations of the same theme -- | all bash-like. But you're probably right, there might be a shell | that I like. I love the idea of Linux, and I want to like it and use it, | but so far I don't. And since I can do anything I want to do in DOS, | I'm a happy camper. | | | | ============================================================================ | | Tue, Dec 28, 2004 - 8:11:40 pm | | ray> | BTW, you never did tell me what you have against 4dos. It is so | powerful and so elegant that (as you tire of hearing me say) I | couldn't live without it. OK, it's huge, but appart from that | what's your beef with it? | | | ============================================================================ | | Tue, Dec 28, 2004 - 8:14:53 pm | | ray> | Oh, one more thing. Can you tell me anything about the | "arachne:restart" label used in the GUI options screen? AFAICT, | in th code this indentifier exists only once in a commented block | of code. It looks to me that however it is processed, it is | by some default of other, not by direct identification. | | | | ============================================================================ | | Tue, Dec 28, 2004 - 8:22:23 pm | | glennmcc> | You hit the nail on the head as to what I hate about 4dos.... | | It's HUGE... therefore it is a waste of precious memory. | | IMO, | Why hold that sucker in memory when a .BAT file can do the exact same | command and uses no memory at-all. | | As you can tell by the fact that I have over 400 BAT files... | | I love BATs :)) | | BRB with data on arachne:restart | | | ============================================================================ | | Tue, Dec 28, 2004 - 8:35:28 pm | | glennmcc> | --- in protocol.c --- | if(!strcmpi(url->file,"restart")) | { | char buf[IE_MAXLEN]; | char *ptr=getenv("ASETUP"); | | if(ptr && !strcmp(ptr,"inst")) //special case - "Finish Setup" button | return GOTO_USEREND; | | arachne.target=0; //!!! | strcpy(buf,"@arachne"); | if(tcpip) | strcat(buf," -o\n"); | else | strcat(buf,"\n"); | unlink("lock"); | *returnvalue=willexecute(buf); | return GOTO_END; | } | _____________________ | | ============================================================================ | | Tue, Dec 28, 2004 - 8:41:38 pm | | glennmcc> | To confirm that this is the correct section of code. | | I just recompiled with this line instead.... | | if(!strcmpi(url->file,"rerun")) | | arachne:restart now results in "Unknown protocol in URL" | | arachne:rerun now results in a restart. | | | | | ============================================================================ | | Tue, Dec 28, 2004 - 9:42:18 pm | | ray> | Nuts, I should have broadened my search for just 'restart'. Thanks | for that. Now I'm wondering why that is handled in protocol.c instead | of in guievent.c like the other simmilar tags. So much to learn. | | As to 4dos, interesting you should mention .bat files, because it is | precicely 4dos's vastly improved .bat file capabilites that weds me | to that program. If it was just about typing on the command line, | 4dos would still be much better, but I could live without it. But | 4dos can do things in batch files that are close to what can be | achived in BASIC. It's like having an interpreter as your shell. | As to memory, sure it uses 200K, but it loads it all in XMS, so you | never see it, and what with virtualy every machine out there having | at least 16M, the memory hit is insignificant, IMO. | | ============================================================================ | | Tue, Dec 28, 2004 - 9:50:50 pm | | ray> | ... Just as an example, here is the output of 'drives.bat' under | 4dos. This output is produced entirely by batch commands, no .com's or | .exe are involved: | | LABEL TOTAL ALLOC FREE ALLOC SUBSITUTIONS | | C: 40G2W98MAIN 3062 2152 909 | D: 15G1WIN98 4186 1502 2684 | E: 40G4SCRATCH 1017 411 606 | F: 40G5TEMP 1017 7 1010 | G: 40G6WORKING 1017 249 768 | H: 40G7STORAGE 29777 9171 20605 | I: 15G2SCRATCH 656 159 497 | J: 15G3WORKING 656 418 238 | K: 15G4CDTEMP 730 2 728 | L: 15G5STORAGE 4941 4427 514 | M: 40G3W98BACK 3062 1667 1394 | N: unlabeled 4186 0 4186 | O: 40G2W98MAIN 3062 2152 909 0.00 O: = C:\APPS\TC\BIN | =============³======================== | R: MS-RAMDRIVE 3062 2152 909 0.00 R: = O:\TEMP | S: 40G7STORAGE 29777 9171 20605 0.00 S: = H:\ | T: 40G5TEMP 1017 7 1010 0.00 T: = F:\TEMP | U: 40G7STORAGE 29777 9171 20605 0.00 U: = H:\USER | V: 40G2W98MAIN 3062 2152 909 0.00 V: = C:\UTIL | W: 40G6WORKING 1017 249 768 0.00 W: = G:\ | X: 40G4SCRATCH 1017 411 606 0.00 X: = E:\MAIN | | | | ============================================================================ | | Tue, Dec 28, 2004 - 9:51:27 pm | | ray> | Shit, I hate the way this board mangles spaces. Pardon my French. | | | | ============================================================================ | | Tue, Dec 28, 2004 - 9:52:25 pm | | ray> | ...LABEL.........TOTAL...ALLOC....FREE......ALLOC....SUBSITUTIONS | | C:.40G2W98MAIN....3062....2152.....909 | D:.15G1WIN98......4186....1502....2684 | E:.40G4SCRATCH....1017.....411.....606 | F:.40G5TEMP.......1017.......7....1010 | G:.40G6WORKING....1017.....249.....768 | H:.40G7STORAGE...29777....9171...20605 | I:.15G2SCRATCH.....656.....159.....497 | J:.15G3WORKING.....656.....418.....238 | K:.15G4CDTEMP......730.......2.....728 | L:.15G5STORAGE....4941....4427.....514 | M:.40G3W98BACK....3062....1667....1394 | N:.unlabeled......4186.......0....4186 | O:.40G2W98MAIN....3062....2152.....909.......0.00....O:.=.C:\APPS\TC\BIN | =============³======================== | R:.MS-RAMDRIVE....3062....2152.....909.......0.00....R:.=.O:\TEMP | S:.40G7STORAGE...29777....9171...20605.......0.00....S:.=.H:\ | T:.40G5TEMP.......1017.......7....1010.......0.00....T:.=.F:\TEMP | U:.40G7STORAGE...29777....9171...20605.......0.00....U:.=.H:\USER | V:.40G2W98MAIN....3062....2152.....909.......0.00....V:.=.C:\UTIL | W:.40G6WORKING....1017.....249.....768.......0.00....W:.=.G:\ | X:.40G4SCRATCH....1017.....411.....606.......0.00....X:.=.E:\MAIN | | ============================================================================ | | Tue, Dec 28, 2004 - 9:53:30 pm | | ray> | ___LABEL_________TOTAL___ALLOC____FREE______ALLOC____SUBSITUTIONS | | C:_40G2W98MAIN____3062____2152_____909 | D:_15G1WIN98______4186____1502____2684 | E:_40G4SCRATCH____1017_____411_____606 | F:_40G5TEMP_______1017_______7____1010 | G:_40G6WORKING____1017_____249_____768 | H:_40G7STORAGE___29777____9171___20605 | I:_15G2SCRATCH_____656_____159_____497 | J:_15G3WORKING_____656_____418_____238 | K:_15G4CDTEMP______730_______2_____728 | L:_15G5STORAGE____4941____4427_____514 | M:_40G3W98BACK____3062____1667____1394 | N:_unlabeled______4186_______0____4186 | O:_40G2W98MAIN____3062____2152_____909_______0_00____O:_=_C:\APPS\TC\BIN | =============³======================== | R:_MS-RAMDRIVE____3062____2152_____909_______0_00____R:_=_O:\TEMP | S:_40G7STORAGE___29777____9171___20605_______0_00____S:_=_H:\ | T:_40G5TEMP_______1017_______7____1010_______0_00____T:_=_F:\TEMP | U:_40G7STORAGE___29777____9171___20605_______0_00____U:_=_H:\USER | V:_40G2W98MAIN____3062____2152_____909_______0_00____V:_=_C:\UTIL | W:_40G6WORKING____1017_____249_____768_______0_00____W:_=_G:\ | X:_40G4SCRATCH____1017_____411_____606_______0_00____X:_=_E:\MAIN | | | ============================================================================ | | Tue, Dec 28, 2004 - 9:55:23 pm | | ray> | :( | | How does one display properly columned text here? | | | | ============================================================================ | | Wed, Dec 29, 2004 - 5:06:54 pm | | glennmcc> | Until Gregy incorperates the changes I made over on my board. You can't. | | ============================================================================ | | Wed, Dec 29, 2004 - 7:05:04 pm | | Ray> | Ah, so it is possible then. | | Sarge, you listening? | | | ============================================================================ | | Wed, Dec 29, 2004 - 8:53:03 pm | | glennmcc> | Yep... no problem. | | Go ahead and post that same stuff over on my board and see what she looks li | | | ============================================================================ | | Wed, Dec 29, 2004 - 9:37:51 pm | | ray> | No need, I believe you. I just wrote to Gregy and asked him. | | Oh, back to the point: try to produce output like that with just | command.com batch files! I've got a hundred others that are just | as powerful. Back before I discovered 4dos, I'd often have to write | an .exe to do something that I can now do with batch files. Heck, | even if I had to use command.com for some reason or other, I'd still | keep 4dos around just as a batch file interpreter ... now, if I could | just figure out why 'system()' won't create a secondary 4dos shell ... | | | | ============================================================================ | | Wed, Dec 29, 2004 - 10:08:48 pm | | glennmcc> | But... my point is that the stuff you are doing in those bats is being | done by 4dos.com for the simple reason that those commands are byuilt | into 4dos.com instead of being in an external .exe or .com | | IMO, | Why have all of that stuff built-in when all of it is available via an | external .exe or .com when needed and it's not taking up any memory | at-all (low mem or XMS), when it's not being used ? | | BRB | | | | ============================================================================ | | Wed, Dec 29, 2004 - 10:18:01 pm | | glennmcc> | ---_wastfile.bat_--- | rem_@echo_off | del_c:\wastfile.txt | c: | echo_c:>c:\wastfile.txt | waste.exe_>>_c:\wastfile.txt | echo_.>>c:\wastfile.txt | d: | echo_d:>>c:\wastfile.txt | waste.exe_>>_c:\wastfile.txt | echo_.>>c:\wastfile.txt | e: | echo_e:>>c:\wastfile.txt | waste.exe_>>_c:\wastfile.txt | echo_.>>c:\wastfile.txt | f: | echo_f:>>c:\wastfile.txt | waste.exe_>>_c:\wastfile.txt | echo_.>>c:\wastfile.txt | g: | echo_g:>>c:\wastfile.txt | waste.exe_>>_c:\wastfile.txt | echo_.>>c:\wastfile.txt | h: | echo_h:>>c:\wastfile.txt | waste.exe_>>_c:\wastfile.txt | echo_.>>c:\wastfile.txt | i: | echo_i:>>c:\wastfile.txt | waste.exe_>>_c:\wastfile.txt | echo_.>>c:\wastfile.txt | c: | ____________________ | | | The_result........... | | c: | WASTED_SPACE_v1.0_by_Dan_Ginsburg | Calculating_drive_statistics,_please_wait... | The_total_size_of_the_disk_is__________:____279470080_bytes | The_total_amount_of_space_used_is______:____170131456_bytes | The_total_free_space_on_the_disk_is____:____109338624_bytes | The_total_number_of_bytes_in_files_is__:____144804613_bytes | The_amount_of_WASTED_SPACE_is__________:_____25326843_bytes | The_percentage_of_disk_space_wasted_is_:__________9.1_% | . | d: | WASTED_SPACE_v1.0_by_Dan_Ginsburg | Calculating_drive_statistics,_please_wait... | The_total_size_of_the_disk_is__________:____328794112_bytes | The_total_amount_of_space_used_is______:____262897664_bytes | The_total_free_space_on_the_disk_is____:_____65896448_bytes | The_total_number_of_bytes_in_files_is__:____255584954_bytes | The_amount_of_WASTED_SPACE_is__________:______7312710_bytes | The_percentage_of_disk_space_wasted_is_:__________2.2_% | . | e: | WASTED_SPACE_v1.0_by_Dan_Ginsburg | Calculating_drive_statistics,_please_wait... | The_total_size_of_the_disk_is__________:____534331392_bytes | The_total_amount_of_space_used_is______:____456687616_bytes | The_total_free_space_on_the_disk_is____:_____77643776_bytes | The_total_number_of_bytes_in_files_is__:____428499769_bytes | The_amount_of_WASTED_SPACE_is__________:_____28187847_bytes | The_percentage_of_disk_space_wasted_is_:__________5.3_% | . | f: | WASTED_SPACE_v1.0_by_Dan_Ginsburg | Calculating_drive_statistics,_please_wait... | The_total_size_of_the_disk_is__________:____534331392_bytes | The_total_amount_of_space_used_is______:____436748288_bytes | The_total_free_space_on_the_disk_is____:_____97583104_bytes | The_total_number_of_bytes_in_files_is__:____431963966_bytes | The_amount_of_WASTED_SPACE_is__________:______4784322_bytes | The_percentage_of_disk_space_wasted_is_:__________0.9_% | . | g: | WASTED_SPACE_v1.0_by_Dan_Ginsburg | Calculating_drive_statistics,_please_wait... | The_total_size_of_the_disk_is__________:____534331392_bytes | The_total_amount_of_space_used_is______:____503439360_bytes | The_total_free_space_on_the_disk_is____:_____30892032_bytes | The_total_number_of_bytes_in_files_is__:____483481623_bytes | The_amount_of_WASTED_SPACE_is__________:_____19957737_bytes | The_percentage_of_disk_space_wasted_is_:__________3.7_% | . | h: | WASTED_SPACE_v1.0_by_Dan_Ginsburg | Calculating_drive_statistics,_please_wait... | The_total_size_of_the_disk_is__________:___1110212608_bytes | The_total_amount_of_space_used_is______:____________0_bytes | The_total_free_space_on_the_disk_is____:___1110212608_bytes | The_total_number_of_bytes_in_files_is__:____________0_bytes | The_amount_of_WASTED_SPACE_is__________:____________0_bytes | The_percentage_of_disk_space_wasted_is_:__________0.0_% | . | i: | WASTED_SPACE_v1.0_by_Dan_Ginsburg | Calculating_drive_statistics,_please_wait... | The_total_size_of_the_disk_is__________:___1110212608_bytes | The_total_amount_of_space_used_is______:____________0_bytes | The_total_free_space_on_the_disk_is____:___1110212608_bytes | The_total_number_of_bytes_in_files_is__:____________0_bytes | The_amount_of_WASTED_SPACE_is__________:____________0_bytes | The_percentage_of_disk_space_wasted_is_:__________0.0_% | | | | ============================================================================ | | Wed, Dec 29, 2004 - 10:19:15 pm | | glennmcc> | Zero memory used when not in use. | | ============================================================================ | | Wed, Dec 29, 2004 - 10:24:06 pm | | glennmcc> | BTW, | | File : /glennmcc/entity.c | Uploaded by : glennmcc1959 < | Description : Quick-n-dirty 'fix' for 'unicode punctuation' | | | ============================================================================ | | Thu, Dec 30, 2004 - 5:07:57 am | | gregy> | hmmm....Glenn...Ray....didn't we have that "formatting spaces" problem | fixed on this board....??? | ... | let's see | four spaces | eight spaces | ten spaces | eight spaces | four spaces | | | | | ============================================================================ | | Thu, Dec 30, 2004 - 5:09:21 am | | gregy> | I'll be doggoned...!!!....I'm sure that we had those fixes "fixed" on | one of these (my) boards.... | ...hmmmm | ... | be back in a bit....I'll check the other boards | | | | ============================================================================ | | Thu, Dec 30, 2004 - 5:21:52 am | | gregy> | Ray...I'll have to sort through my HDDs to see where the board upgrades | went to.... | I KNOW I followed Glenn on those "spacing" upgrades. I was sure that | THIS board, at least (because coders were here, and needed the ability | to paste indented code snipets), had the fixes. | | Glenn has solved this problem loooonnnnnggggg ago........must | have been a year ago... | ... | I'll check around and see where I've stuck them (and never uploaded | them)..... | | If I can't find them...I will hafta re-download Glenn's code and put | the changes into these boards. ....hmmm...but...I'm SURE it was | done about a year ago....and I actually tested it out, too.... | | | | ============================================================================ | | Thu, Dec 30, 2004 - 5:45:50 am | | gregy> | Well, I can't find those "fixed" source files....so....I downloaded | Glenn's "fixed" source, and I'll transfer his fixes over into these | boards. It'll take me a couple days.... | ... | he he....you'll know it when ya see it..... | | | | | ============================================================================ | | Thu, Dec 30, 2004 - 4:10:19 pm | | ray> | Gregy, | | OK, thanks for that, looking forward to it. | | Glenn, | | Yours is Ok, but mine is better, methinks. Anyway, to each his | own; 4dos takes up less low memory than command.com, and even if | it does use some XMS, I couldn't care less since I've got 128M of | it. And I like having all that batch power at my finger tips. | Over the years I've deleted dozens of old utility .coms and .exes | that I no longer need since the power to do the same thing is built | into 4dos -- more memory used, but less disk space. | | To each his own. | | | ============================================================================ | | Thu, Dec 30, 2004 - 4:56:04 pm | | ray> | Glenn, | | That's a nice fix. Every now and then I get some character that's | not shown proberly, so now I think I know how to fix 'em, thanks. | | BTW, my take on your code: | | | ============================================================================ | | Thu, Dec 30, 2004 - 4:58:28 pm | | ray> | Let's see if Gregy has it fixed: | | | //Moved from entity.c | //HTML /2.0 entities (defined in /fontinfo.bin) | unsigned char HTMLentity(char *name) | { | //~179~ | //!!glennmcc: begin Dec 29, 2004 -- 'fix' for puncuation in numeric code fo | //RAY: 04-12-30: Pointless 'elses' removed. | if (!strcmpi(name, "#146") | || !strcmpi(name, "#8217")) | //RAY: Personal pref: ASCII 39 for both opening & closing single quotes. | //if (!strcmpi(name, "#8216")) return atoi("96"); //ASCII 96 == ` | || !strcmpi(name, "#8216")) return atoi("39"); //ASCII 39 == ' | | if (!strcmpi(name, "#8220") | || !strcmpi(name, "#8221")) return atoi("34"); //ASCII 34 == " | //!!glennmcc: end | | | | | | ============================================================================ | | Thu, Dec 30, 2004 - 4:59:16 pm | | ray> | ... not yet. | | | ============================================================================ | | Thu, Dec 30, 2004 - 5:57:38 pm | | ray> | Hey Gregy, while you're tinkering with the board, here's another | request: assuming the others have no objection, I'd like it if | the board automaticaly went to the bottom, not the top, since that's | always where the action is. I know it would just save one keypress, | still, I'd like it. | | BTW, Glenn, this ain't important, but can you explain why the | 'arachne:restart' tag is handled where it is instead of in | 'guievent.c'? The latter would seem to be the logical choice, and it | used to be handled there, but was moved. I have no plans to touch | anything, but I'm curious as to why the change was made. | | | ============================================================================ | | Thu, Dec 30, 2004 - 9:01:53 pm | URL | | little_birdie> | Link to ....... | | http://www.hi-line.net/~gfeig/adev/advindex.php#end | | instead of ...... | | http://www.hi-line.net/~gfeig/adev/advindex.php | | Now you go to the bottom instead of the top. | | | | ============================================================================ | | Thu, Dec 30, 2004 - 10:08:45 pm | | ray> | Now isn't that cute, I had no idea things could be so simple. | (I love things that are simple. Genius is the ability to make | things simple. 'Tis the gift to be simple, 'Tis the gift to be free.) | | Nite Nite. | | | ============================================================================ | | Thu, Dec 30, 2004 - 10:19:27 pm | | glennmcc> | Why in protocol.c instead of in guievent.c ? | | You got me..... I sure du'no why. | | We'd need to ask Michael why he moved it there. | | All I know is.... it ain't broke... so I ain't fixin it. :)) | | | | ============================================================================ | | Fri, Dec 31, 2004 - 7:33:52 am | | ray> | I ain't fixin it neither. It's the same with this 'clr.bat' stuff -- | when something 'obvious' is not done, there's got to be a reason for | it. Still, finding that reason could be instructive. I'm at the stage | now where I'm trying to figure out how Michael thinks. | | | | ============================================================================ | | Fri, Dec 31, 2004 - 7:34:32 am | | ray> | 1 space | 2 space | 3 space | | | ============================================================================ | | Fri, Dec 31, 2004 - 3:50:04 pm | | gregy> | Ray....he he.... | 1 O'Leary | 2 O'Leary | 3 he he he | | I'll try to get this board patched over this weekend (I should be | able to do it...I can sorta remember where Glenn fixed it)...and... | I downloaded his current source, so that should refresh my memory | ... | NOTE - I'll post a notice, when I get the patched board up... | | | | | ============================================================================ | | Fri, Dec 31, 2004 - 5:18:26 pm | | ray> | Thanks Gregy, | | | | ============================================================================ | | Fri, Dec 31, 2004 - 10:19:22 pm | | glennmcc> | Happy New Year !!!!!!!!!!!!!!!!!!!!!!!! | | And I do hope it's a very happy one for everyone. :) | | ============================================================================ | | Fri, Dec 31, 2004 - 10:30:41 pm | | gregy> | ...aaannnnndddd....a H A P P Y N E W Y E A R !!!!!!!!!!!!!!!!!!! | .....................---------------------------...................... | ... | to one and all...... | | | ============================================================================