[04:43] meh. mesa is so hard to cross-compile [04:44] you have this stupid builtin_compiler thing that needs to be built using the --build compiler, and then chattr +i, make clean to get rid of all the native object files, and make again to finish cross compiling. pfft. [04:45] hyperair: i'm hoping to have multiarch on natty and oneiric in edgers in the next week so it should be a bit easier [10:45] tjaalton, RAOF, Sarvatt, bryceh: do you mind if I upload a new mesa with versioned breaks on proprietary drivers? [10:45] (otherwise my new drivers won't be installable) [10:57] hmm... most of you must be on holiday... [11:00] the git branch looks a little confusing [11:00] tseliot, yes that's fine [11:01] tseliot, in fact Sarvatt and I were planning on trying to figure out what needed done tomorrow [11:01] bryceh: I guess the last commit was not uploaded even though there's no UNRELEASED word in the changelog [11:01] commit a6aaf514c91a593d79e9ae8899798641d52d19de [11:02] i.e. Fix !i386 !amd64 builds by only installing i915g there, ie: where it's actually built [11:02] could be; I didn't talk with raof about it [11:03] bryceh: I'm wondering if it's ok to upload that change if it's not already in [11:03] I don't know; perhaps there was more raof needed to do? [11:03] could he have uploaded it but it failed to build? [11:04] had he uploaded the commit we should've seen at least something in the changelog in git [11:05] bryceh: I guess they won't miss i915g on armel anyway ;) [11:05] heh [11:06] oh, wait there's something in the changelog [11:07] bryceh: yes, the change is already in [11:07] I'll clean up things in git too [11:08] yep. $ apt-cache policy libgl1-mesa-dri [11:08] libgl1-mesa-dri: [11:08] Installed: 7.10.3-0ubuntu3 [11:08] Candidate: 7.10.3-0ubuntu3 [11:08] tseliot, any changes needed for xorg-server? [11:09] bryceh: no, I don't think so. All we need is to version the breaks in Mesa (I've tested things here for a few days) [11:39] jcristau: I'm getting Permission denied (publickey) when trying to push my commits. Any ideas? [11:40] my username is tseliot-guest [11:40] check that you're using the right key? [11:40] jcristau: I've certainly not changed my key for a while now [11:43] when was the last working commit? [11:43] jcristau: also, if I try to ssh into alioth.debian.org, it says that remote host identification has changed, so I'm wondering if anything changed there [11:43] let me check [11:43] yes, something changed there [11:45] http://lists.debian.org/debian-devel-announce/2011/05/msg00007.html http://lists.debian.org/debian-devel-announce/2011/05/msg00008.html [11:46] jcristau: it was b28a90c9a78f605d521a9ba1cda42f23ad487e10 [11:46] that's not a date [11:46] jcristau: sorry, Thu Dec 2 13:34:35 2010 +0100 [11:48] yeah, so long before the move [11:48] try ssh to vasks.debian.org, see what that tells you [11:49] jcristau: still Permission denied (publickey) [11:51] then make sure https://alioth.debian.org/account/editsshkeys.php has the key you're trying to use [11:51] (and that it's not a dsa key) [12:06] jcristau: I've tried that (with RSA) but I'm still getting the same error. Shall I wait a few minutes? [12:06] tseliot: fix the hostname in .git/config [12:06] to git.d.o [12:08] tjaalton: I'm using git+ssh://tseliot-guest@git.debian.org/git/pkg-xorg/lib/mesa.git [12:09] ok.. maybe plain ssh then? [12:10] can't check my config [12:12] I keep getting the following: [12:12] Permission denied (publickey). [12:12] fatal: The remote end hung up unexpectedly [12:13] ok, so it's probably not related to the alioth upgrade [12:14] at least the way i thought it was :) [12:31] jcristau: I've tried creating a new ssh key and setting my ssh config file to use that but it doesn't seem to solve the problem [12:31] * tseliot is starting to think that debian.org hates him [12:47] the cron to sync the keys from the db runs at the hour [13:23] Sarvatt: No more crashes since I installed your PPA package. [13:29] ok, I'll wait then [13:29] jcristau, tjaalton: thanks for your help [13:33] if stuff still doesn't work there's always mailto:admin@alioth.d.o or #alioth on oftc [13:39] jcristau: it works now, thanks again [13:40] ah cool. [17:54] Sarvatt, the 275 blob is still going to be a big problem because of the supported pciid script being broken right? === yofel_ is now known as yofel