Thanks!

This is mainly to test the new macOS patch building. Sorry in advance for the noise...

Could you set svn:ignore to *.cpp on source/third_party/ogre3d_preprocessor/tests (like we do in other tests folders)? That would be great for upcoming svn sts ?

In D2522#105360, @Stan wrote:

Maybe we could have such a file and a test?

Should be good after rP23305.

Thanks for your patch, I have included the first part (with the other one you needed) in our upgraded NVTT.

Hm I just checked on Trac and I was probably confused. The Phab revisions are all from the lobby moderator and references on Trac all point to him as well. I was thinking we had a H_Baraq contributor way back then, in the 2013s or something, but I'm wrong.

Thanks for the work!

In D2501#104715, @Freagarach wrote:

See also D506?

@voroskoi Maybe you should send a pull request upstream. There is already this PR by a friend but it looks like the NVTT maintainer doesn't want it, maybe they'd rather like your solution.

Like Vlad said, I have just uploaded a NVTT upgrade, so it would be better if you'd test the branch (and I would happily include your patch in my upgrade if it's still needed).

I'm uploading it here for Jenkins, though arcanist may have botched the job, we'll see.

Welp, it definitely did...

nvtt will be rebuilt in D2475. nspr was rebuilt with the SM45 upgrade. I used the tool https://github.com/lucasg/Dependencies to ensure that all the other DLLs do not depend on the 2010 runtime.

I'm committing this with a license header and some whitespace fixes. Thanks for all the work!

Updated version using Docker (in place since months now).

Ach, I forgot to accept first, too late now. I applied this to Jenkins, everything works ?

I patched the Jenkins script, but this robocopy call is broken, I had to manually copy files on the slave. This needs to be fixed before committing.

This works on Jenkins and locally. The build failures above are mostly failed tests of the latest iteration of D1946 ?

This is completely outdated, now that the build uses Docker. I will upload the current version of scripts here.

@Stan the error in checkrefs.pl is really strange on Jenkins... ? (https://jenkins.wildfiregames.com/blue/organizations/jenkins/docker-svn/detail/docker-svn/319/pipeline) it's saying it parses units/rome_centurio_imperial, what is this file, I can't find it? Is this a bug in the perl script caused by the empty file here?

Hello! I had not noticed the patch when you created it ? I think it's a good idea.

In D2193#91271, @elexis wrote:

It can only be closed if it is accepted.

Thanks for taking a look ?

The setIsRunOnce addition must be reverted.

Updated RC following comments by historic_bruno, wraitii and elexis.

This looks fine to me, works with SM38 and without any change with SM45, and so much hacky code is removed that I can only be happy ? The autorequest looks needed to me.

Alright, everything is cleared up then! Thanks again for the work :)

Thanks for the answers! Indeed my dictionary must be marking only the vowels where confusion can happen. I still have one word where I do have a short vowel.

This is good now!

Thanks for the work! I went through my own dictionary for these changes (Latin-French by Gaffiot & Flobert, 2001) and I have a few differences. A lot of them are vowel lengths differing. Note that sometimes, my dictionary only says that the vowel is not always long, not that it is always short.

Thanks for looking into this! The concept looks good to me and works, but it has a memory leak. The fix is easy ?

I rebased my branch and added a commit addressing comments: https://github.com/na-Itms/0ad/tree/sm45.

Rebased; this is a RC.

Rebased.

The code looks almost perfect to me (I had too many >> problems in the past, it became an obsession), and this works on SM38 and on SM45. Thanks a lot!

Fix an integer overflow.

Thanks for the review ?

I used the latest one (probably vc12) with the compiler I wanted to build the game with. Recent versions of visual studio can open slightly older solutions without issue ?

You were right, there are a few bogus names here! I tracked them down, they are users who deleted their Transifex account. I added a regex to remove them from the credits.

This is superseded by our new CI system which is using Docker to build the game under clang-only or gcc-only environments.

This will be useful to start supporting VS2019. I would be against updating premake if we have no bugs or missing features in our version though, because that makes heavy commits. What is your opinion? If you agree, we should wait for when we actually enable 2019, at that time maybe that will be a more recent alpha version of premake.

@elexis, in the future, if you want to manually start an autobuild, it's better to use the helper job that triggers the autobuilder. Else, that helper job will start an autobuild as planned on the next day and create an extra commit. The helper job has "timeguard" in its name. If you run a build through it, it will reset its internal state.

Oh, well, I don't know. I usually license my stuff under GPL and it would make sense to put it under GPLv2+ like most of the 0ad code. However, seeing how barebones and easy to adapt this snippet is, maybe MIT would make more sense.

Address comments and use the opportunity to test the new Jenkins setup.

@Stan The script always returns 0, which makes Jenkins not complain about anything. See for instance https://jenkins.wildfiregames.com/blue/rest/organizations/jenkins/pipelines/svn/runs/1085/nodes/46/log/?start=0.

I experimentally added this pipeline to Jenkins with the fix to arc patch, and made Phabricator run it on differentials. We'll see if there are unexpected issues. In the meantime, a few comments...

Vorbis was compiled with VS2010 and still depended on the redistributables I deleted. I upgraded and recompiled libogg and libvorbis in rP22406, which fixes the issue for me. Other libs might create the same issue if built with VS2010 still.