Might be a stupid question since I don't understand XMB very well, but isn't that why we usually don't specify extensions, so the game can figure out whether to load the XML or the XMB?

Fair enough, looks good to me.

One could also wrap the constants in a macro?

fix tests.

Readd jschanges tests are still broken

Simpler and faster version

Remove the potential bottleneck, move to graphics

years

LGTM.

Rebase. @tpearson-raptor can you test again now?

Well I'm wondering whether it's spurious because it makes no sense, or whether it's spurious because it's annoying to fix :)

Seems to work fine on Intel 630 and my Nvidia card I asked @OptimusShepard to test on AMD

Works as expected (aka no 501 error).
I'm wondering about the gl. prefix since we do that nowhere else (e.g the novbo is not called gl.novbo)

Seems to fix the warnings for me. I suppose that GCC7 bug is not documented?

Weird case but what if you garrison a skirmish entity? (Those that get replaced on map load) so you have a map which has garrison points in XML

Should I abandon this?

I'm not sure I follow, can you give me a pointer?

Can you fix all those warnings? I fixed everything for A24 ^^"

Changes in Renderer.cpp seems unrelated and can go in separately;

Sweet. Will fix headers and commit now that I know its use.

Is there any perf hit/improvement of doing this?

Can you check D1593 again?

MOAR DEFINES

And another attempt

Definitely test this without pch.

Try GLDEBUGPROCKHR

Fix missing param, and don't be a dummy

Fix GL_APIENTRY on Windows fix some notes and format specifier

Define GL_DEBUG_SOURCE_APPLICATION for macOS

Rebase, need the CI for macOS errors

Should be in checkrefs too.

I will try to provide something yeah but maybe I'll do it after this patch is committed :) I'm supposed to get some rest. Failing completely for now :/

Oh right you should definitely hear them. One question about abuse. Is there a way we can limit spam!

If it's not too much trouble can you provide a little mods with the steps? so it's easy to test? E.g with broken templates maps and whatnot

In D1751#157606, @wraitii wrote:

@nwtour can you confirm it works for you?

I'm fine with either, just thought it was worth mentioning :)

Why is it unneeded ?

We have a user with both libpng12 and libpng16 installed and I wonder why it doesn't use pkgconfig?

Restore to the previous version + rebase

I didn't commit it cause I could not reproduce the bug D1593 looks simple enough for me to commit it. One thing you can do is close this one (abandon) and commandeer the other one so we can finish it.

I really wish we would have gone for alphabetical schema ordering, it's soooo easy to autofix. I did it for hyrule and it's pretty cool.

Did you see D1593? It has a different approach not sure which is better

Also won't this break @nani's autociv workaround for undetectable mods?

Looks alright, shouldn't you ignore the test.cpp files too? I remember someone mentioning it.

Those languages are not present in the release. Check A23b :)

The reason we do not add those characters is that it makes the fonts too big. Instead CJK are provided as mods of around 100MB each.

Copy paste FTL

Use em dash instead. In case I'm not here, and someone can't wait, they have the permission to commit this.

Looks like I made a mistake in my hotfix in rP24856

I tested this patch and it seems to fix the issue. I found another way to reproduce. (It works both in packaged an non packaged and triggers a stack overflow)

Okay so here comes the frustrating and tough part :/ So bear with me cause we're in for a ride. Just know that it's not against you :)

I see now still you could use our define cause it has more things

Well your patch does it anyway :)
And not for all only for x86_x64::Family() == 6 if IA32 is detected

Hey thanks for the patch ! You might want to move that into the check for IsHyperThreading capable.

Can you add yourself to the credits? :)

For me it's A25 material. I will ask maintainers about it but so far the people I asked were really mitigated about it.

Sure. :)

D3520 says it's complete

Sorry, what do you call the manifest file?
Defining it in premake might be a good idea. I could set those define there yes I suppose.