| 30 Sep 2021 |
 John Ericson | like yocto does it | 02:22:04 |
| John Ericson | the gnome reputation about not givinga fuck about outsider's patches sadly seems pretty true though | 02:22:31 |
| John Ericson | but yeah could just take the second commit and do a patch | 02:23:00 |
| John Ericson | the whole thing is so shitty though | 02:23:13 |
 colemickens 🏳️🌈 | That's too bad, I certainly understanding wanting to do it in the right place, for more than one reason. | 02:23:19 |
| John Ericson | it's fundamentally a terrible design | 02:23:22 |
| John Ericson | and relying on binfmts as it does can go wrong, though hopefully not too bad | 02:23:43 |
| John Ericson | colemickens 🏳️🌈: https://gitlab.gnome.org/GNOME/gobject-introspection/-/merge_requests/224/diffs that is the second PR | 02:27:19 |
| John Ericson | it's actually pretty simple semantically | 02:27:28 |
| John Ericson | just moving some things and reindenting some things | 02:27:33 |
| 2 Oct 2021 |
|  cdepillabout joined the room. | 04:48:39 |
| 4 Oct 2021 |
 siraben | What should I change /usr/lib/gcc/arm-none-eabi/6.3.1/libgcc.a to in a Makefile? | 03:16:41 |
| siraben | Looks like the program needs to link against a static cross-compiled libgcc | 03:16:54 |
| siraben | GCC_LIBR = "${pkgs.libgcc}/lib/libgcc.a"; fixed it | 04:11:47 |
| siraben | where pkgs ? (import nixpkgs {}).pkgsCross.arm-embedded | 04:12:01 |
 trofi | $gcc -print-file-name=libgcc.a might also get the path. Or use LDFLAGS += -Bstatic -lgcc -Bdynamic if it's used just for linking. (both untested) | 06:34:37 |
| trofi | ideally -lgcc should be an internal detail of the compiler. | 06:35:37 |
| trofi | Also, perhaps there is no libgcc.so at all (libgcc_s.so should be an equivalent). Thus for LDFLAGS -lgcc should be enough. | 06:37:06 |
 sterni | we don't have ready-made solutions for compiler bootstrapping like lib.makeScopeWithSplicing, right? | 14:58:09 |
| sterni | I think at some point we'll want to make a similar abstraction for staged package sets which contain the compiler and run time libraries | 14:58:51 |
| sterni | ideally this could also work for the stdenv boostrap / reuse parts of that? | 14:59:18 |
| 6 Oct 2021 |
|  Rosario Pulella changed their display name from rosariopulella to Rosuavio. | 10:38:41 |
| Rosario Pulella changed their display name from Rosuavio to Rosario Pulella. | 10:45:04 |
|  kamadorueda joined the room. | 22:33:48 |
| 7 Oct 2021 |
| sterni | John Ericson: seems like we need to bring back the old school clangStdenv which was using lld regardless of hostPlatform.linker | 16:32:48 |
| sterni | since building any C++ just fails hillariously when linking using ld.bfd | 16:33:06 |
| sterni | the question just is what to call it | 16:33:13 |
| sterni | I suspect libcxxStdenv should always have lld, but we want libstdcxx + lld as well probably | 16:33:52 |
| sterni | plain clangStdenv with ld.bfd is fine more or less since it works fine for C | 16:34:05 |
| 8 Oct 2021 |
 pie_ | (xposting from main) | 00:48:22 |
