Gered wrote:

Disassembled the Toshiba 430CDT I picked up a week ago. Removed both the backup and RTC batteries, neither of which had begun leaking... but better safe then sorry. Also wanted to clean the mouse buttons which were dirty and kind of sticky. Had to pretty much disassemble the entire laptop to get at it only to discover that the buttons are actually part of the very same plastic mold as the top part of the laptop's frame (the same frame which the keyboard fits in) and so there's no way to physically detach them. So I had to remove all the metal bits on that end of the frame and try to very carefully run some water over it, take it away, scrub, rinse, and repeat for a bit to get it mostly clean. Not perfect, but clean enough. This was incredibly annoying to do because I wasn't confident that I could detach the LCD hinges without breaking something, so I couldn't just immerse the part of the frame I wanted to clean in water and scrub away because the LCD assembly was still attached to it. Ugh.

amadeus777999 wrote:

Worked on the C/Assembler Tutorial that I have been planning to do for a long time.

WatcomC + inline + external assembly with TASM.

Nothing too fancy, but should give a nice base for somebody starting his own project. Getting all those little details right is quite a challenge.

Cool, looking forward to reading it when it's done. 😀 I wrote a little something along those lines a while back but it is a bit light on details and only focuses on the "Watcom C + external assembly with TASM" aspect. Mainly wrote it for myself because I kept mucking some little detail up with the calling convention and got tired of flipping through pages in the compiler manual to find whatever detail I was forgetting at the time, heh.

I'll be looking through your ramblings thoroughly - this is looking good... especially the calling conventions and the "Exploring Watcom" article(wish I had such a lofty physical copy).

The essential details are often hard to come by - meaning having the minimum solution at hand that works and is correct. Symbols being all uppercase via TASM/MASM or WatcomC's calling convention can be filed under "deadly details".(devcpp + yasm was easier in this regard). Also getting the convention regarding Watcom's inline assembly right can be bothersome too - fpu related for example.
I also could not use the rdtsc instruction with TASM so I used the "infamous" db 0fh 31h in combination with equ.

Another one where less is more and everybody seems to have a different solution is makefiles... especially Watcom's with their .lnk file "magic".

I tried an example on a site and it just does not work. Do you happen to have a wasm specific makefile with more than one .c file which one could use as a base?

makefiles in general are easy if you do everything explicitely but I would prefer to use the capabilites of the mechanisms behind it... and then the trouble seems to be starting.

Installing Windows 95C on a freshly formatted 10GB Western Digital drive using the XT-IDE Universal BIOS version R596 a fellow VCFed.org user has sent to me. The next thing to figure out is how to silence a 52x CD-ROM drive.

That didn't work out too well. Plugged in the Quantum Fireball ST 3.5 Series and the full 3.2GB capacity was detected, but, the HDD isn't working very well (beeps and clicks).

the_ultra_code wrote:

Dude, that's dope! Would you be kind enough to please share with me pics of the rest of the machine? I am truly fascinated and amazed. 😀

Short work log with parts list: https://smallformfactor.net/forum/threads/wir … rm-factor.8236/

amadeus777999 wrote:

I'll be looking through your ramblings thoroughly - this is looking good... especially the calling conventions and the "Explorin […]
Show full quote

I'll be looking through your ramblings thoroughly - this is looking good... especially the calling conventions and the "Exploring Watcom" article(wish I had such a lofty physical copy).

The essential details are often hard to come by - meaning having the minimum solution at hand that works and is correct. Symbols being all uppercase via TASM/MASM or WatcomC's calling convention can be filed under "deadly details".(devcpp + yasm was easier in this regard). Also getting the convention regarding Watcom's inline assembly right can be bothersome too - fpu related for example.
I also could not use the rdtsc instruction with TASM so I used the "infamous" db 0fh 31h in combination with equ.

Another one where less is more and everybody seems to have a different solution is makefiles... especially Watcom's with their .lnk file "magic".

I tried an example on a site and it just does not work. Do you happen to have a wasm specific makefile with more than one .c file which one could use as a base?

makefiles in general are easy if you do everything explicitely but I would prefer to use the capabilites of the mechanisms behind it... and then the trouble seems to be starting.

Indeed, finding some of these details in the form(s) of a minimal working solution can be challenging. I suppose one could argue that is half the "fun" of using these antiquated development tools. 😉 I tended to scribble a lot of this stuff down in my notebook but then decided that I might as well write posts about them, even if only just for my own benefit. Glad I'm not the only one intending on doing this type of thing.

I found the Watcom "Tools User's Guide" helpful when writing my own makefile for WMAKE (also this one too for completeness sake). I generally use that Makefile as a template and just rename a couple bits for each new project. I'm sure there's many aspects that could be improved (for one, I personally dislike having to list out all the object files explicitly, but DOS doesn't give me the tools I need in order to not do this, heh), but it works for me *shrug*.

Regarding WASM .. I initially started using that when I started to move away from mostly inline assembly. I liked that it was by default aware of Watcom's register-based calling convention. Of course, it's very similar to MASM (if you like MASM syntax, I prefer TASM's "ideal" mode myself) but it's not 100% the same, and the documentation for it is severely lacking, which is ultimately what pushed me away from using it in the end. I guess they just assumed most people knew MASM syntax and that was good enough, no need to write any detailed documentation of their own.

I didn't have any trouble using it in a makefile though. In fact I used it successfully in the previously linked Makefiles a while back... I just had "wasm" instead of "tasm" (and also appropriate command line switches). I just looked to see if I still had an older copy of one of those Makefiles where I had used it but it seems I don't anymore ... unfortunately I tend to be a bit less diligent about committing changes to version control because there is no MS-DOS version of git 😀.

Recently someone on Amibay had some potentially interesting hardware, but the price was a bit higher than I was prepared to pay. However he was looking for various things including a Panasonic 5.25" 360k FDD. Now, I had one of those stuck in the back of a cupboard for months, but had never tested it due to no media and no reason to get any. But now I did order some online.

Yesterday I received some floppies in the post, so I hooked it up, told BIOS what was connected, booted, slipped a floppy in, did dir b: and... it just worked. Smoothly, quietly and quickly - with far less worrying noise than any of my 3.5" drives. The next two floppies also worked fine for good measure.

Now I'm agonizing over whether I really want to offer it up for exchange, or keep such a beautifully engineered piece of history...

So after multiple tries with spray can clear coat on my steel case cover I have concluded this is not the best approach. It comes out blotchy and uneven even with multiple coats. I also found those hairs I mentioned in an earlier post look to be impurities in the clear coat mix itself. I am getting black specks, what look to be fine micro fibers, etc. coming from the can itself as I spray it on. I triple checked the case before spraying to make sure it was clear of any debris. So my next move is to roll on a clear coat and I will be consulting with a paint dept. for advice. This computer case restore will get done right or not at all.

I’m quite interested in this because I have a couple of case covers that need a paint job. i havn’t experimented yet but getting the right paint gloss level and colour is probably key.

Recorded Tom and Jerry off a tape I got yesterday.

You gotta love the quality 🤣

Also tested out how an OPL3 soundcard goes with Mario Minix (whoever played Mario Forever knows it - I should add most Minix songs ARE MIDI! Yes people, they're MIDIs.) and I can say one thing - missing instruments. 🤣

Gered wrote:

Indeed, finding some of these details in the form(s) of a minimal working solution can be challenging. I suppose one could argue […]
Show full quote

amadeus777999 wrote:

I'll be looking through your ramblings thoroughly - this is looking good... especially the calling conventions and the "Explorin […]
Show full quote

I'll be looking through your ramblings thoroughly - this is looking good... especially the calling conventions and the "Exploring Watcom" article(wish I had such a lofty physical copy).

The essential details are often hard to come by - meaning having the minimum solution at hand that works and is correct. Symbols being all uppercase via TASM/MASM or WatcomC's calling convention can be filed under "deadly details".(devcpp + yasm was easier in this regard). Also getting the convention regarding Watcom's inline assembly right can be bothersome too - fpu related for example.
I also could not use the rdtsc instruction with TASM so I used the "infamous" db 0fh 31h in combination with equ.

Another one where less is more and everybody seems to have a different solution is makefiles... especially Watcom's with their .lnk file "magic".

I tried an example on a site and it just does not work. Do you happen to have a wasm specific makefile with more than one .c file which one could use as a base?

makefiles in general are easy if you do everything explicitely but I would prefer to use the capabilites of the mechanisms behind it... and then the trouble seems to be starting.

Indeed, finding some of these details in the form(s) of a minimal working solution can be challenging. I suppose one could argue that is half the "fun" of using these antiquated development tools. 😉 I tended to scribble a lot of this stuff down in my notebook but then decided that I might as well write posts about them, even if only just for my own benefit. Glad I'm not the only one intending on doing this type of thing.

I found the Watcom "Tools User's Guide" helpful when writing my own makefile for WMAKE (also this one too for completeness sake). I generally use that Makefile as a template and just rename a couple bits for each new project. I'm sure there's many aspects that could be improved (for one, I personally dislike having to list out all the object files explicitly, but DOS doesn't give me the tools I need in order to not do this, heh), but it works for me *shrug*.

Regarding WASM .. I initially started using that when I started to move away from mostly inline assembly. I liked that it was by default aware of Watcom's register-based calling convention. Of course, it's very similar to MASM (if you like MASM syntax, I prefer TASM's "ideal" mode myself) but it's not 100% the same, and the documentation for it is severely lacking, which is ultimately what pushed me away from using it in the end. I guess they just assumed most people knew MASM syntax and that was good enough, no need to write any detailed documentation of their own.

I didn't have any trouble using it in a makefile though. In fact I used it successfully in the previously linked Makefiles a while back... I just had "wasm" instead of "tasm" (and also appropriate command line switches). I just looked to see if I still had an older copy of one of those Makefiles where I had used it but it seems I don't anymore ... unfortunately I tend to be a bit less diligent about committing changes to version control because there is no MS-DOS version of git 😀.

THANKS, I'll go through these today!

Set up one of the the SUN Blades that I got - before this I added one cpu and some ram(may add a better graphics card too). Nice system but never did anything with these. SUN Solaris took an aweful long time to install and this was only version 8 - on top of that no C Compiler and the second cpu is not correctly recognized. Guess i have to test Solaris 9 tomorrow.

I was reading a forum thread from back in 2007 about Pentium Dual Core E2xx0 overclocking (on another computer themed forum).

Many posts were misguided and comprised mostly of erroneous information. other posts were borderline retarded. Getting more and more irratated over the fact that people asking questions diddn't get any useful help in the thread I started to wonder if we really were that clueless back then. Suddenly there were a few good (or at least better) posts setting everything right and my mood brightened. I thought finally a sane person among all these shit posters and poor misinformed kids, I would have written something just like this! Then I checked who the author of the posts was... Then I thought Oh...

This is borderline "modern" as the performance is nearing useful levels but as I have written about the not so great Asus Striker Extreme nForce 680i motherboard here already I might aswell continue.

It seems the 333 MHz FSB northbridge strap is alot less problematic than the 266 MHz FSB northbridge strap. Its a pity you can't control these straps on this motherboard in any other way than pin- (pad-?) modding the CPUs or use a CPU that runs the preferred strap/FSB natively.

Even with an unsupported 45nm Q9650 3GHz (9x333) quad core most issues I had with older 65nm 266 MHz FSB quads disappeared. I could even overclock the Q9650 to 9*400 = 3600MHz with full stability without pushing high VTT, 401 MHz FSB wouldn't even POST though. This is still a crappy result compared to other motherboards and chipsets but alot better then my earlier experiences with nForce 680i in combination with quad core CPUS.

the_ultra_code wrote:

Oh, I promise you, I'm not. :) […]
Show full quote

dionb wrote:

An EEPROM flasher is never a bad investment - but don't get your hopes up too much for this board.

Oh, I promise you, I'm not. 😀

I've had it pretty lucky up until now, so I would think it would make sense that I would finally start running into... bad luck, eh?

Besides, this just screams "Problems that only the gods know how to fix.", it seems just that strange. I've never in my whole career of this hobby of my (a whole year 🤣; it's a drop in the bucket compared to you guys) anything like this before.

I was right. I finally got my TL866II Plus EPROM programmer today from China, so I followed the nice little guide that PhilsComputerLabs made to set everything up, flashed the BIOS chip on this Asus TX97-XE, reinserted the BIOS chip back into the motherboard, hooked up the motherboard to the StarTech PSU, inserted a PCI graphics card a some SDRAM, removed the CMOS battery and then reinserted it just to "make sure" I was on a clear slate, turned on the power supply, hit the power button, and... same behavior as before.

RIP Asus TX97-XE (1997-ish - 2018) 😢

Because I am desperate to finish this build, I found on ebay an Intel LT430TX motherboard for $166-ish and got that. There was a similarly speced Intel AN430TX (basically the same board from what I can tell) that also had onboard audio on sale on ebay for around $100, too, but, continuing with my bad luck lately, when I checked the listing again today, some smart fool bought it the day before. 😢 x 10

The only good thing to come out of today was that I learned while listening to The Crystal Method's Community Service album that I recently acquired that one of the songs on the album ("Dude in the Moon (Luna Mix)", by Dastrix), was a prominent song in Need for Speed: High Stakes. When I heard the song, I was like, "Wait, I heard that before. Wait, that's from NFS:HS, no?!?", and, with a quick Google search, I found that indeed a "radio edit" 3-minute mix of the song is in the game's soundtrack (http://nfs.wikia.com/wiki/Need_for_Speed:_Hig … akes/Soundtrack). Is it just me, or did a lot of songs found on Crystal Method albums find there way into early Need for Speed games (up to, say, around the Underground period)? 😀

Oh, bonus post of dubious retro quality. Been listening to this Art of Noise song ("Paranoimia") a lot. https://youtu.be/6epzmRZk6UU

Any thoughts? 😀

amadeus777999 wrote:

[...]

Set up one of the the SUN Blades that I got - before this I added one cpu and some ram(may add a better graphics card too). Nice system but never did anything with these. SUN Solaris took an aweful long time to install and this was only version 8 - on top of that no C Compiler and the second cpu is not correctly recognized. Guess i have to test Solaris 9 tomorrow.

Solaris 8 is definitely too ancient for a beast like this. Solaris 9 is a bit more resource-heavy, but with two late model SPARC CPUs that's not an issue really. A lot of interesting hardware (such as SunPCI x86 cards with K6-2 onboard to run DOS/Windows in a window in Solaris) is only supported up to 8, but will work fine under Solaris 9, all you need to do is edit the config files and add Solaris 9 (SunOS 5.9) in the list of supported versions.

Been doing some minor updates to a few of my PCs. Currently trying to settle on a companion 2D/3D card for my K6-2/SLI PC; after selling off my Mystique 220 (which turned out to be a mistake, no pun intended) so now I have to dig into my stash of contemporary (1996-1998) PCI cards and pick one. I tried a Virge/DX but the 2D was pretty bad and caused rolling lines all over the place on my LCD even though the black levels were right. Now I need to pick my poison between a Rage Pro PCI, Matrox Millennium PCI or a TNT2 M64 PCI (I know it's a 1999 card but it's a passive 16MB OEM version that I can pretend is a TNT PCI until I can actually find one). I'm leaning towards the TNT2 M64.. Open to suggestions, however.

dionb wrote:

amadeus777999 wrote:

[...]

Set up one of the the SUN Blades that I got - before this I added one cpu and some ram(may add a better graphics card too). Nice system but never did anything with these. SUN Solaris took an aweful long time to install and this was only version 8 - on top of that no C Compiler and the second cpu is not correctly recognized. Guess i have to test Solaris 9 tomorrow.

Solaris 8 is definitely too ancient for a beast like this. Solaris 9 is a bit more resource-heavy, but with two late model SPARC CPUs that's not an issue really. A lot of interesting hardware (such as SunPCI x86 cards with K6-2 onboard to run DOS/Windows in a window in Solaris) is only supported up to 8, but will work fine under Solaris 9, all you need to do is edit the config files and add Solaris 9 (SunOS 5.9) in the list of supported versions.

Interesting - do you happen to have a SUN running and been using it for some time?

I'm a little bit of lost on these machines due to never having had anything to do with them. I have two, so I will sell one and keep another for testing & experimenting.
Regarding Solaris 9 - I installed it today and everything went fine except for the companion CD not being recognized... which meant I had to install a C Compiler by hand... exactly what I wanted to avoid

I have heard of these CPU boards and maybe I'll "bite" if I happen to come across one.

I installed gcc later on and it was a total hassle(pkgadd'ing didn't work because of some nifty oversights by me) due to not knowing my way around - in hindsight a snap but, as always, those little details! On one hand I hate it on the other I'm always thinking of K & R unix'ing around, so I kinda feel indebted to at least gain a little bit of "expertise".

I have an Ultra10 on the way which interests me way more than these newer machines. BUT, at least they are really Sparc based - and not x86 powered which would be simply boring.

I'm planning on running Quake and Doom on them - any tips regarding Solaris/Unix based "ports"?

I Started working on my Tualatin machine today. Been wanting to do a BX based Tualatin build for a long time. Never had any luck running a Slot 1 BX board @ 133mhz no matter how hard I tried or combination of Slokets etc! I did pick up a while ago a Abit BX-133 board but the caps were shot and I need to re-cap it. I was lucky to pick up a DFI-ITOX GCB60-BX-Rev C which has turned out to be the Perfect motherboard for what I've wanted to do. It is able to run the 1.4-S@133mhz no problem, Is a socket 370 board so better cooling options, UDMA-100 capable and 3 ISA slots!!

So far it is looking like this

Penitum III 1400-S
256mb PC-133 Ram
DFI-ITOX GCB60-BX-Rev C Motherboard
Hitachi 80GIG Hard Disk
Pioneer DVD Rom
Geforce 2 Ultra AGP GFX Card
Yamaha 724 PCI Soud Card (With SB-LINK)
Turtle Beach Pinnacle REV-F with Daughter board and SPDIF bracket

I plan to add a AWE64 to handle the Soundblaster Stuff and Maybe another wavetable card in the mix.

I bought another Acme Desktop Case just for this build. It feels like its all clicking into place.

Nice build.

appiah4 wrote:

Been doing some minor updates to a few of my PCs. Currently trying to settle on a companion 2D/3D card for my K6-2/SLI PC; after selling off my Mystique 220 (which turned out to be a mistake, no pun intended) so now I have to dig into my stash of contemporary (1996-1998) PCI cards and pick one. I tried a Virge/DX but the 2D was pretty bad and caused rolling lines all over the place on my LCD even though the black levels were right. Now I need to pick my poison between a Rage Pro PCI, Matrox Millennium PCI or a TNT2 M64 PCI (I know it's a 1999 card but it's a passive 16MB OEM version that I can pretend is a TNT PCI until I can actually find one). I'm leaning towards the TNT2 M64.. Open to suggestions, however.

Is S3 Savage4 outside of your preferred date range? The picture quality, compatibility and a quirky API are all there 😀

As for me - I've purchased a Covox Speech Thing clone from the Spanish e-shop that keropi has been mentioning. This is to improve the DOS sound compatibility of my P3 PC - where ESS Solo-1 in TDMA mode is the only (PCI) sound card I'm willing to go with.

Revolter wrote:

Is S3 Savage4 outside of your preferred date range? The picture quality, compatibility and a quirky API are all there 😀

Savage4 is 1999 so yeah it's one year too far in the future.. That said, I would be willing to cheat abit to get the Savage4 in there like I am willing to do for the TNT2 M64. Ideally I would be very happy to have a Savage3D PCI in the system, but I can find neither the Savage3D nor the Savage4 PCI cards for cheap 🙁

As for me - I've purchased a Covox Speech Thing clone from the Spanish e-shop that keropi has been mentioning. This is to improve the DOS sound compatibility of my P3 PC - where ESS Solo-1 in TDMA mode is the only (PCI) sound card I'm willing to go with.

ESS Solo-1 is my favorite PCI sound card for PCI only P3 systems, great choices there.

amadeus777999 wrote:

Not much of a dev myself, so not sure whether this is typical or not... […]
Show full quote

dionb wrote:

Sadly, no. My Sun systems (I actively used a Sparcstation 20 for a few years and a maxed-out Ultra5 briefly as well, although I […]
Show full quote

amadeus777999 wrote:

[...]
[...]

Interesting - do you happen to have a SUN running and been using it for some time?

Sadly, no. My Sun systems (I actively used a Sparcstation 20 for a few years and a maxed-out Ultra5 briefly as well, although I was disappointed with that one, as well as an IPX, a 5, a Voyager and an Ultra1) were some of the casualties of moving from a house with cellar and shed to a small apartment in the big town 10 years ago.
[qiote]I'm a little bit of lost on these machines due to never having had anything to do with them. I have two, so I will sell one and keep another for testing & experimenting.
Regarding Solaris 9 - I installed it today and everything went fine except for the companion CD not being recognized... which meant I had to install a C Compiler by hand... exactly what I wanted to avoid

Not much of a dev myself, so not sure whether this is typical or not...

I have heard of these CPU boards and maybe I'll "bite" if I happen to come across one.

They are fun stuff, very similar to comparable boards for Amiga and Mac, Sun just supported them for longer. The last versions were So370. Note that particularly video performance is bad - integrated chipsets and limited overlay possibilities - so don't expect to go gaming on them. Also the different versions have different OS requirements. The ones that work under Solaris 8 are easy to get to work under 9, but the ones only supported in Solaris 7 are a different kettle of fish.

I installed gcc later on and it was a total hassle(pkgadd'ing didn't work because of some nifty oversights by me) due to not knowing my way around - in hindsight a snap but, as always, those little details! On one hand I hate it on the other I'm always thinking of K & R unix'ing around, so I kinda feel indebted to at least gain a little bit of "expertise".

I have an Ultra10 on the way which interests me way more than these newer machines. BUT, at least they are really Sparc based - and not x86 powered which would be simply boring.

I'm planning on running Quake and Doom on them - any tips regarding Solaris/Unix based "ports"?

Not really, wasn't into gaming on them at the time. Iirc it was never really their forte.

That Ultra10 sounds marginally more interesting, but like the Ultra 5 (its desktop version) it was from the time when Sun started using commodity PC ODMs to manufacture their boxes - and it shows. Build quality is way down from the old Sparcstation & Ultra 1 / 2 level. Also iirc the fastest CPU you could put in it was 440MHz, but even though it was a highly clocked 64b CPU, it felt more sluggish than the K6-2 400 I had on a PCI card in it. Still, back then I had piles to choose from that were almost being given away. Now beggars can't be choosers and an Ultra10 is still very much old-school pre-Oracle real-SPARC Sun 😀