umläute

well, i don't think that printing values to the Pd-console is especially useful apart from debugging. so i'm really talking about making those values in the patch. but yes, if...

@tytan652 wrone: > Maybe just download tarballs with the right commit hashes for each sub-modules. ah well, we have infrastructure to download tarballs for us, but unfortunately that doesn't work...

here's a script that creates a tar-file with all the submodules included: ```sh prefix=obs-studio-HEAD rm -f release.tar git archive --format tar -o release.tar --prefix=$prefix/ HEAD PREFIX=${prefix} TOPDIR=$(pwd) git submodule --quiet...

I should be able to provide a PR that adds the workflow. however, before I did into that: I haven't found a [workflow (for obs-studio)](https://github.com/obsproject/obs-studio/tree/743117f08060d3d79f92740f46a60cbeed2e2699/.github/workflows) yet that actually creates release...

sidenote: afaik there's no way to check whether the user added something like `JUCE_USE_OGGVORBIS=1` to their build flags from within the JUCE Cmake snippets. that's why i was asking for...

while i understand, that you do not want to support "unmodified, untested, versions of dependent libraries", I still think that a Cmake command to enable specific features (without having to...

something like this? ```diff diff --git a/test/test_socketcan.py b/test/test_socketcan.py index a2c4fae..5d2de5d 100644 --- a/test/test_socketcan.py +++ b/test/test_socketcan.py @@ -276,6 +276,10 @@ class SocketCANTest(unittest.TestCase): ), "Should only run on platforms where sizeof(long) ==...

well, that was the "quick fix". a somewhat nicer fix would be to convert the buffer to native byte ordering first: ```diff From a01d03ad075911496b5a9f4d899dbd9e5f539ce5 Mon Sep 17 00:00:00 2001 From:...

and the same should probably be applied to `test_build_bcm_header_sizeof_long_4_alignof_long_4`. so it's probably worth using a helper-function: ```python def bytes2native(data, item_type="i", item_size=4): import struct fmt = "%d%c" % (len(data) / item_size,...

i found that of the 10 different 64bit platforms supported by Debian, not a single one would support CPU-based clipping (as tested by `m4/clip_mode.m4`. OTOH, all 32bit platforms had some...