[GITHUB] Update build-toolchain to use public VC6 from itsmattkc repo #931

fbraz3 · 2025-05-24T19:16:33Z

This pull request updates the build workflow to download the VC6 toolchain from a public repository.

Key changes:

Uses the public itsmattkc/MSVC600 repository to fetch the VC6 portable toolchain.
Removes the need to maintain or use any secrets for toolchain downloads in CI builds.
Simplifies maintenance and improves transparency for all contributors.

This change makes the build process more secure and easier to reproduce for everyone.

TODO

Check if produced executable files are identical

xezon · 2025-05-24T19:25:28Z

.github/workflows/build-toolchain.yml

        shell: pwsh
        run: |
          Write-Host "Downloading VC6 Portable Installation" -ForegroundColor Cyan
-          aws s3 cp s3://github-ci/VS6_VisualStudio6.7z VS6_VisualStudio6.7z --endpoint-url $env:AWS_ENDPOINT_URL
+          Invoke-WebRequest -Uri https://github.com/itsmattkc/MSVC600/archive/refs/heads/master.zip -OutFile VS6_VisualStudio6.zip


If itsmattkc takes the repo private, then this breaks. Are we allowed to fork this repository or is this illegal?

There are already several forks of this repository. If itsmattkc decides to make the repository private or remove it, we can simply switch to another existing fork. Alternatively, we could revert to using the AWS backup if necessary.

https://github.com/itsmattkc/MSVC600/forks?include=active%2Cinactive%2Cnetwork&page=1&period=&sort_by=stargazer_counts

But realistically, this repository has been public for almost five years, and we are talking about end-of-life software. I find it unlikely that it will be taken down at this point.

Regarding the legality of keeping a fork, I am not a lawyer, but as far as I know, simply maintaining a fork of freely distributed executables—without modifying or hacking the binaries—should not be illegal. However, if there are any concerns, we should seek legal advice to be certain.

If allowed, Superhackers should make their own fork

I guess allowed would be based on whatever Microsoft says no? As it is their software, don't know if you still need a license for VS6. If you don't then we could just create our own repo and not even fork this one.

it would probably be better to have our own repo, that ways it's easier for people to get hold of a copy for local use as well and it keeps dependencies within the project as much as possible.

We can do that if someone can find out if that is ok to do.

In legal terms what’s the difference between get vs6 from an s3 bucket and get it from any public repo?

What i can read me to, it looks like we would need a license and run the script locally. I believe the actions run on @tintinhamans's home server? So that should have the license?

xezon · 2025-06-23T20:37:35Z

@OmniBlade what are your thoughts on this?

OmniBlade · 2025-06-23T22:11:41Z

It does make it easier to do builds in our own forks, I'd be inclined to go with it and we either revert/fix it if the repo goes down and hope we don't need VC6 by the time that happens. Github is owned by Microsoft so while someone has directed their infrastructure to create the repo, its being served from Microsoft infrastructure to be run on Microsoft infrastructure. At worst they would just stop it if they took issue with it and its not really any different from what we are doing now? Probably legally more in the clear that having Amazons infrastructure provide copies, but still clear as mud.

Skaronator · 2025-07-07T20:45:29Z

As mentioned, this would be really useful to run the CI in our own forks. Worst case would be we have to fix the CI if the source repository actually goes down, which I agree is very unlikely.

The long-term plan is probably to drop VC6 altogether and move forward with VC2022, right?

aliendroid1 · 2025-07-07T21:01:09Z

As mentioned, this would be really useful to run the CI in our own forks. Worst case would be we have to fix the CI if the source repository actually goes down, which I agree is very unlikely.

The long-term plan is probably to drop VC6 altogether and move forward with VC2022, right?

yes, therefore we shouldn't bother making our own fork/clone of the vc6 portable repo

xezon · 2025-07-07T21:09:57Z

Ok. Shall we create a fork of itsmattkc in this organization and then link to that?

aliendroid1 · 2025-07-07T21:19:01Z

Ok. Shall we create a fork of itsmattkc in this organization and then link to that?

No. Thats additional liability for us and not worth avoiding the small risk that the itsmattkc repo gets deleted before we drop vc6.

worst case scenario even if it does get deleted, its not hard to mitigate. We can just use another repo or fallback to our current method

xezon · 2025-07-07T21:25:05Z

Why is this branch not buildings anything? The CI is not triggered.

Please rebase and push this branch.

CI is not working.

fbraz3 · 2025-07-09T02:20:07Z

Why is this branch not buildings anything? The CI is not triggered.

Please rebase and push this branch.

Done!

xezon · 2025-07-09T07:01:13Z

.github/workflows/build-toolchain.yml

        shell: pwsh
        run: |
          Write-Host "Downloading VC6 Portable Installation" -ForegroundColor Cyan
-          aws s3 cp s3://github-ci/VS6_VisualStudio6.7z VS6_VisualStudio6.7z --endpoint-url $env:AWS_ENDPOINT_URL
+          Invoke-WebRequest -Uri https://github.com/itsmattkc/MSVC600/archive/refs/heads/master.zip -OutFile VS6_VisualStudio6.zip


tomsons26

Regarding vs6 download change, I don't see masm in that repo, that probably isn't sp6 but rtm
[It would mean] Nonmatching codegen, possible new issues cause of compiler bugs fixed in service packs

Can we please make a binary diff of itsmattkc SP6 with our current SP6?

itsmattkc/MSVC600@001c4ba

"upgraded to professional edition sp6"

It claims to be SP6

i checked c1.dll and the version number is 12.0.9782.0 which is the SP6 version

it says its a "Portable download of Microsoft Visual C++ 6.0 command line tools."
Thats why it doesn't have all the files

I compated against OmniBlades portable version:

Mismatching files are

BSCMAKE.EXE 77886 2022.06.04 06:04:32 != 2024.02.05 16:19:58 77886 BSCMAKE.EXE CVPACK.EXE 81979 2022.06.04 06:04:32 != 2024.02.05 16:19:58 81979 CVPACK.EXE FCOUNT.BAT 169 2022.06.04 06:04:32 != 2024.02.05 16:19:58 178 FCOUNT.BAT FCOVER.BAT 169 2022.06.04 06:04:32 != 2024.02.05 16:19:58 178 FCOVER.BAT FTIME.BAT 169 2022.06.04 06:04:32 != 2024.02.05 16:19:58 178 FTIME.BAT LCOUNT.BAT 348 2022.06.04 06:04:32 != 2024.02.05 16:19:58 359 LCOUNT.BAT LCOVER.BAT 489 2022.06.04 06:04:32 != 2024.02.05 16:19:58 502 LCOVER.BAT MAPSYM.EXE 39184 2022.06.04 06:04:32 != 2024.02.05 16:19:58 39184 MAPSYM.EXE MC.EXE 25872 2022.06.04 06:04:32 != 2024.02.05 16:19:58 25872 MC.EXE MIDL.EXE 673616 2022.06.04 06:04:32 != 2024.02.05 16:19:58 673616 MIDL.EXE MKTYPLIB.EXE 91136 2022.06.04 06:04:32 != 2024.02.05 16:19:58 91136 MKTYPLIB.EXE PLIST.EXE 61498 2022.06.04 06:04:32 != 2024.02.05 16:19:58 61498 PLIST.EXE PREP.EXE 77880 2022.06.04 06:04:32 != 2024.02.05 16:19:58 77880 PREP.EXE PROFILE.DLL 94270 2022.06.04 06:04:32 != 2024.02.05 16:19:58 94270 PROFILE.DLL PROFILE.EXE 41022 2022.06.04 06:04:32 != 2024.02.05 16:19:58 41022 PROFILE.EXE PROFILER.INI 2528 2022.06.04 06:04:32 != 2024.02.05 16:19:58 2658 PROFILER.INI REBASE.EXE 59664 2022.06.04 06:04:32 != 2024.02.05 16:19:58 59664 REBASE.EXE VCVARS32.BAT 954 2022.06.04 06:04:32 != 2024.02.05 16:19:58 989 VCVARS32.BAT Include\ RPCPROXY.H 19006 2022.06.04 06:04:32 != 2024.02.05 16:20:00 19007 RPCPROXY.H WINNT.H 252068 2022.06.04 06:04:32 != 2024.02.05 16:20:00 245610 WINNT.H WINPERF.H 28649 2022.06.04 06:04:32 != 2024.02.05 16:20:00 28650 WINPERF.H Lib\ MSJAVA.LIB 549908 2022.06.04 06:04:32 != 2024.02.05 16:20:01 226694 MSJAVA.LIB MFC\SRC\ MFCINTL.MAK 4161 2022.06.04 06:04:32 != 2024.02.05 16:20:02 4312 MFCINTL.MAK MFCISAPI.MAK 6375 2022.06.04 06:04:32 != 2024.02.05 16:20:02 6721 MFCISAPI.MAK

Missing files are

2024.02.05 16:20:03 513536 MasmRef.doc 2024.02.05 16:20:03 227307 procpack.chm 2024.02.05 16:20:03 22709 PPreadme.htm Bin\ 2024.02.05 16:19:58 21880 h2inc.err 2024.02.05 16:19:58 249344 h2inc.exe 2024.02.05 16:19:58 9687 ml.err 2024.02.05 16:19:58 385072 ml.exe 2024.02.05 16:19:58 180276 MSPDB60.DLL Samples\Amd\Asdk\ 2024.02.05 16:20:03 6725 adetect.h 2024.02.05 16:20:03 4534 adsp.h 2024.02.05 16:20:03 3353 amath.h 2024.02.05 16:20:03 2667 amatrix.h 2024.02.05 16:20:03 52438 amd3dx.h 2024.02.05 16:20:03 2097 amdlib.h 2024.02.05 16:20:03 3610 aquat.h 2024.02.05 16:20:03 3296 atrans.h 2024.02.05 16:20:03 3181 avector.h 2024.02.05 16:20:03 20575 detect.cpp 2024.02.05 16:20:03 15772 dsp_lib.cpp 2024.02.05 16:20:03 40781 idsp.cpp 2024.02.05 16:20:03 58188 imath.cpp 2024.02.05 16:20:03 27889 imatrx.cpp 2024.02.05 16:20:03 33664 iquat.cpp 2024.02.05 16:20:03 23507 itrans.cpp 2024.02.05 16:20:03 13042 ivect.cpp 2024.02.05 16:20:03 20939 math_lib.cpp 2024.02.05 16:20:03 16273 matrx_lib.cpp 2024.02.05 16:20:03 20466 quat_lib.cpp 2024.02.05 16:20:03 31119 testlib.cpp 2024.02.05 16:20:03 8089 trans_lib.cpp 2024.02.05 16:20:03 7998 vect_lib.cpp Samples\Amd\Isdk\ 2024.02.05 16:20:03 6725 adetect.h 2024.02.05 16:20:03 4532 adsp.h 2024.02.05 16:20:03 3457 amath.h 2024.02.05 16:20:03 2668 amatrix.h 2024.02.05 16:20:03 3610 aquat.h 2024.02.05 16:20:03 3296 atrans.h 2024.02.05 16:20:03 3181 avector.h 2024.02.05 16:20:03 20575 detect.cpp 2024.02.05 16:20:03 15723 dsp_lib.cpp 2024.02.05 16:20:03 35084 idsp.cpp 2024.02.05 16:20:03 58053 imath.cpp 2024.02.05 16:20:03 32369 imatrx.cpp 2024.02.05 16:20:03 2022 intrinsics.h 2024.02.05 16:20:03 27418 iquat.cpp 2024.02.05 16:20:03 22015 itrans.cpp 2024.02.05 16:20:03 11529 ivect.cpp 2024.02.05 16:20:03 20934 math_lib.cpp 2024.02.05 16:20:03 16273 matrx_lib.cpp 2024.02.05 16:20:03 20466 quat_lib.cpp 2024.02.05 16:20:03 31197 testlib.cpp 2024.02.05 16:20:03 8089 trans_lib.cpp 2024.02.05 16:20:03 7976 vect_lib.cpp Samples\CPUID\ 2024.02.05 16:20:03 8967 cpuid.c 2024.02.05 16:20:03 1041 cpuid.h 2024.02.05 16:20:03 859 main.c

Is this alright?

@aliendroid1 @xezon just for the records, i have this patch running under my fork.

The last build results:
https://github.com/fbraz3/GeneralsGameCode/actions/runs/16178498437

Execution logs:
Generals
GeneralsMD

Thats awesome cuz I had been trying so hard to get it to work without mspdb60.pdb. I want to figure out what makes this version work without it but not mine

I suggest we make binary comparison of the generated executables with the current VC6 and the new VC6. If they match, then it should be all good.

Sigh, xeson's report is wrong. mspdb60.dll actually is not missing and is in fact in the itsmackrepo at it's default location.

Looks like some files are in different folders, so they show as missing.

.github/workflows/build-toolchain.yml

xezon · 2025-07-16T09:50:44Z

So we need to check if the produced executable files are identical to before. If yes, then this is safe.

xezon requested a review from tintinhamans May 24, 2025 19:23

xezon added Minor Severity: Minor < Major < Critical < Blocker Build Anything related to building, compiling labels May 24, 2025

xezon reviewed May 24, 2025

View reviewed changes

xezon requested a review from OmniBlade June 23, 2025 20:37

aliendroid1 mentioned this pull request Jul 7, 2025

[GITHUB][ZH] Validate replays for mismatches in CI #1239

Draft

xezon previously approved these changes Jul 7, 2025

View reviewed changes

xezon self-requested a review July 7, 2025 21:26

fbraz3 force-pushed the public-vc6 branch from fbc5248 to 4e16933 Compare July 9, 2025 02:01

xezon reviewed Jul 9, 2025

View reviewed changes

fbraz3 force-pushed the public-vc6 branch from 4e16933 to 5425357 Compare July 9, 2025 19:07

[GITHUB] Update build-toolchain to use public VC6 from itsmattkc repo

4b630a5

fbraz3 force-pushed the public-vc6 branch from 5425357 to 4b630a5 Compare July 9, 2025 19:30

[GITHUB] Update build-toolchain to use public VC6 from itsmattkc repo #931

Are you sure you want to change the base?

[GITHUB] Update build-toolchain to use public VC6 from itsmattkc repo #931

Conversation

fbraz3 commented May 24, 2025 • edited by xezon Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

TODO

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

fbraz3 May 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

xezon commented Jun 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

OmniBlade commented Jun 23, 2025

Uh oh!

Skaronator commented Jul 7, 2025

Uh oh!

aliendroid1 commented Jul 7, 2025

Uh oh!

xezon commented Jul 7, 2025

Uh oh!

aliendroid1 commented Jul 7, 2025

Uh oh!

xezon commented Jul 7, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

fbraz3 commented Jul 9, 2025

Uh oh!

xezon Jul 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

xezon Jul 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Mismatching files are

Missing files are

Uh oh!

Choose a reason for hiding this comment

Uh oh!

aliendroid1 Jul 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

xezon commented Jul 16, 2025

Uh oh!

Uh oh!

fbraz3 commented May 24, 2025 •

edited by xezon

Loading

fbraz3 May 31, 2025 •

edited

Loading

xezon commented Jun 23, 2025 •

edited

Loading

xezon commented Jul 7, 2025 •

edited

Loading

xezon Jul 9, 2025 •

edited

Loading

xezon Jul 9, 2025 •

edited

Loading

aliendroid1 Jul 11, 2025 •

edited

Loading