| Age | Commit message (Collapse) | Author |
|---|
|
|
|
This reverts commit 45b9eed30320c6281efcfb748098a8789b15f15f.
|
|
I've tried using FMOD instead, but CEF didn't work either.
At first I used crow-misia's WebRTC build but it would cause a
segmentation fault, but LL's build seems to break CEF.
Gotta find a way so CM's build doesn't crash the viewer.
PKGBUILD should be moved to indra/newview as an .in to be configured
by CMake for dynamic version numbers, and adjust the instruction
too to run makepkg -R from the folder where the generated PKGBUILD
will be.
|
|
It wasn't inheriting the viewer's BUILD_SHARED_LIBS setting.
|
|
|
|
Also simplify CMake-based dependency projects, the parameters that
have been set for the viewer seem to have been implied all this time
for the subprojects.
|
|
and specialise it just for FreeBSD, along with other fixes to get
it building again on FreeBSD with an empty packages folder.
|
|
|
|
|
|
I thought /opt/local/include wasn't needed since we switched to boost181,
but PCRE headers are there. If you still have "boost" installed (maybe
depended on by some other package, just deactivate it first before cmake).
|
|
so we don't need the boost package or the -no_static variant of
boost181 any more.
|
|
Now that it's only Debian and macOS x86-64 that are left safe to use
LL's WebRTC build, the logic is reversed.
|
|
|
|
System 2.5.2 caused too much rainbow in DeltaFPS. For now, the OpenJPEG
listed in autobuild.xml is 2.5.0. However, LL has recently got 2.5.2
too in their OpenJPEG fork repo, but we switch to that once it's the
one listed in autobuild.xml.
Reverting to the now maintained LL 3p-openjpeg should fix the texture
thrashing problem https://megapahit.com/show_bug.cgi?id=1 starting
from DeltaFPS.
|
|
|
|
Thanks to the Linux x86-64 WebRTC binary from Zenichi Amano (crow-misia).
https://megapahit.com/show_bug.cgi?id=64
Haven't been tested on openSUSE, but it should work.
|
|
|
|
2024.08-DeltaFPS
|
|
|
|
repository.
|
|
Clone the repository by git and make it a cmake submodule.
|
|
|
|
by managing to compile LL's ColladaDOM 2.3 on-the-fly.
|
|
on macOS arm64, the header directory.
|
|
Spaces are really just troublesome in argument passing. They can't
really be escaped, quoting from -D would even replace spaces with
semicolons. Hence making use of every variable that can be used in
Collada DOM's CMakeLists.txt.
|
|
This reverts commit 3bba39cf3e4d56da5aba883479539ed89fa578b9.
|
|
|
|
So that colladadom_installed won't throw USESYSTEMLIBS to the next
scope.
Also libpcrecpp_LIBRARIES doesn't take spaces (might work by escaping
the quotes, but only 1 pcrecpp needed anyway).
|
|
Turns out getting 2.3 compiled on those platforms need a lot more
attention.
Apart from that, the compilation result variable name got fixed,
the prebuilt include directory got fixed, and pkg-config-ing minizip,
libpcrecpp and libxml-2.0 so we can make have generic variable arguments
to pass to Collada DOM's CMake settings.
|
|
https://megapahit.com/show_bug.cgi?id=76
It seems like we have to use LL's Collada DOM 2.3.
Make sure minizip is installed on macOS.
It should be safe to uninstall your system Collada DOM package now.
The CMake arguments might have to be completed for non-Darwin
platforms in a next commit.
|
|
|
|
It breaks CEF on these platforms.
|
|
for macOS arm64.
|
|
Ubuntu, unlike Debian, has /lib64 and /usr/lib64 in their system,
so it misled CMake. This is a quick fix, ideally it's the distro
name that is checked.
|
|
|
|
|
|
cause they're not like 0 valued ${_binary}_installed that it would
run lipo -thin on already thinned binaries.
|
|
|
|
Refer to this commit if you need to recover the steps in compiling
XMLRPC-EPI.
The prebuilt binary for Darwin is distributed as universal, and will
be thinned accordingly. It also makes use of autobuild.xml, but only
partially (the sha1sum isn't used yet in the Prebuilt.cmake reimplementation).
|
|
also to minimise ViewerMisc.cmake diff from upstream.
|
|
Anticipate the possibility that an archive isn't suffixed with a 64.
Parentheses need to be escaped by quotes (backslashing would work too),
otherwise they would be parsed as part of CMake commands syntax.
Regex subexpressions shouldn't contain regex themselves (only verbatim
or preprocessed strings are used within the parentheses).
|
|
|
|
* #2354 Enable Tracy on test builds
* Only enable Tracy by default on windows
|
|
|
|
2024.06-atlasaurus
|
|
|
|
Turns out it was on by default, probably that's why it would fail
on everyone else's systems, building libcurl, cause in the log
I saw libssh2 was requiring the very openssl3 we have to deactivate
until libcurl compilation is done.
|
|
Its own CPACK_RPM_PACKAGE_REQUIRES will catch up soon.
|
|
|
|
|
