| 20 Oct 2020 |
 tar | i got your .dsc however it's downloading stuff at build time, that's a no-go | 06:04:21 |
| tar | mooff: find my build logs here: http://phd-sid.ethz.ch/debian/fractal/2/ | 06:22:24 |
| tar | and debian/changelog and debian/copyright clearly look wrong (stopped looking after those two) | 06:22:47 |
| tar | my builds took about 12' each (one in a sbuild, one unclean sid environment) | 06:23:29 |
 emorrp1 | tar: haha, no I'm just good at translating between people with different domain knowledge | 07:41:33 |
|  phaze left the room. | 08:52:42 |
|  get_me_dog joined the room. | 10:15:09 |
| *  uhoreg waves at tar | 14:42:06 |
| tar | uhoreg: did you see https://github.com/trunkmaster/nextspace/issues/325 ? | 16:19:11 |
| uhoreg | No, I didn't. I've mostly lost touch with the *Step world, but that looks neat. | 16:23:18 |
| 21 Oct 2020 |
|  init joined the room. | 13:01:52 |
| 23 Oct 2020 |
 RSS Bot [@hubert:uhoreg.ca] | Debian package news for matrix-synapse:
A new upstream version is available: <a href="https://github.com/matrix-org/synapse/archive/v1.22.0rc1.tar.gz">1.22.0~rc1</a> | 04:31:59 |
|  roland left the room. | 06:18:41 |
| 24 Oct 2020 |
| emorrp1 | Hi all, I've written up my Synapse experiences on my Single Board Computer (via freedombox). I'd appreciate any further performance issues or recommendations you've come across: https://github.com/matrix-org/synapse/issues/8428 | 13:03:38 |
| 25 Oct 2020 |
| uhoreg | jonas: thank you for updating emscripten in sid | 13:55:49 |
 jonas | I am trying... - it still doesn't succeed in compiling olm :-/ | 13:58:15 |
| uhoreg | :( | 13:59:02 |
| 26 Oct 2020 |
|  pearguin joined the room. | 15:24:22 |
| uhoreg | Looks like the only issue is that it can't find google-closure-compiler. But if I change the Makefile to use --closure 0, then it seems to succeed. So I guess we "just" need to tell it how to find google-closure-compiler. | 19:27:49 |
| jonas | I am aware of the closure issue - and have now pushed the patch I made for that: https://salsa.debian.org/matrix-team/olm/-/commit/5a5f4a5d032e01195bc1fb25a706803635f6182f | 19:55:30 |
| jonas | the issue I am facing now is a different one - when new emscripten hopefully succeeds to build within an hour, I will try make a release to experimental of olm - both so that others like you can help figure out what fails (if it still fails), and to get the package queued up for ftpmaster approval (introducing the javascript library will trigger a visit to NEW queue) | 19:58:05 |
| uhoreg | Too old closure-compiler? checks apt-cache policy Oh. version is 20130227. :-/ | 20:03:05 |
| uhoreg | Well, I ran a "make js" in a debian:unstable Docker with emscripten installed, and it built something that passed the unit tests. | 20:06:38 |
| uhoreg | What issue are you running into? | 20:07:05 |
| jonas | undefined symbol | 20:11:09 |
| jonas | error: undefined symbol: __stack_chk_fail (referenced by top-level compiled C/C++ code) | 20:12:05 |
| jonas | I am not familiar with docker - did you run as root or as a separate user without write access to /usr/share/emscripten? | 20:16:34 |
| uhoreg | I'm running as root | 20:17:15 |
| jonas | my pbuilder environment builds as a separate user and if build fails it provides me a root account - I have experienced failure to build but when then repeating as root it succeeds | 20:18:26 |
| jonas | ...so I supect there is something broken in emscripten related to resolving system-shared librarie | 20:19:06 |
