=== Brownout_ is now known as Brownout === asac_ is now known as asac [04:28] chrisccoulson: you still around? === Nafallo_ is now known as Nafallo [13:21] chrisccoulson, are the ff crash dumps resolved in breakpad? [13:21] i mean, in the UI [13:21] like here: https://crash-stats.mozilla.com/report/index/410b7810-9278-40b2-92da-846972110217 [13:22] everything i see is unusable (at least from my p-o-v) [13:23] fta2 - yeah, those crashes are awkward because there are no symbols for system libraries [13:23] most crashes are ok though [13:23] the crashes are resolved by sorocco btw [13:23] oh [13:23] http://code.google.com/p/socorro/ [13:23] sorocco? [13:24] perhaps we need to figure out a way of generating breakpad symbols from gtk and glib ;) [13:24] although, most crashes don't really need them [13:25] for chromium, it seems it's possible [13:33] well, seems difficult to achieve anyway [13:33] http://www.chromium.org/developers/decoding-crash-dumps [13:34] we'd need to generate those for all versions of all the libs in deps [13:34] and something will have to merge the right combination of files [13:35] chrisccoulson, ^^ === m_conley_away is now known as m_conley [15:07] fta2 - interesting. it actually looks quite trivial to get breakpad symbols from our existing -dbgsym packages then [15:07] fta^^ [15:07] heh ;) [15:07] i could probably quite easily do that for the major system libraries used by both firefox and chromium [15:08] i'm not sure whether we'd really need to do it for every library though [16:34] chrisccoulson, yep, i wanted to ship those breakpad binaries in a package, somewhere; but i didn't do it, so far [16:36] those various symbols need to be merged after each crash, matching the user setup [16:47] dpm, fyi: http://code.google.com/p/chromium/issues/detail?id=72749 [16:48] hey fta, looking... [16:56] fta, I understant that apart from the *_settings.grd issue you are mentioning (IIRC we talked about this a while ago), this is not an issue for the Ubuntu builds? Do we need to ask for feedback to e.g. Basque and Galician translators? [16:58] dpm, no. the bug is about ICU, i don't use the system one, so we're impacted. but our system icu doesn't have all the improvements google added to its copy [16:58] the problem is we can't have both. [17:02] for example; chrome's icu support word wrapping for chinese, ours doesn't [17:03] ours = the system lib [17:23] did firefox use to have an option to view individual cache items? like sort/clear cache from individual websites? [17:23] * vish cant seem to find that option in FF4 [17:24] vish: about:cache [17:24] err, individual sites, idr [17:25] or maybe I'm just mixing it up with the cookies options.. [17:50] jdstrand, hi. [UFW BLOCK] IN=eth0 OUT= MAC= SRC= DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=126 TC=0 HOPLIMIT=255 FLOWLBL=0 PROTO=UDP SPT=5353 DPT=5353 LEN=86, it's mDNS, right? [17:55] fta: yeah [17:55] jdstrand, shouldn't it be allowed & silenced by default? [17:56] that is probably bug #720605 [17:56] Launchpad bug 720605 in ufw "UFW drops sane multicast traffic" [Undecided,New] https://launchpad.net/bugs/720605 [17:57] I need to review it more. there is an (unreviewed) work-around in the bug [17:58] i see no side effect to my drops, i don't expect any mDNS on that box [17:59] in the log, i see 1 pkt denied per hour [17:59] fta: what is your log level? 'sudo ufw status verbose' [18:01] Logging: off [18:01] but still, it spams my logs [18:01] that is weird [18:02] # dmesg | grep -c UFW [18:02] 1489 [18:02] with AUDIT, ALLOW and BLOCK [18:02] fta: do any lof your rules have 'log' in them? sudo ufw status verbose|grep -i log [18:03] none [18:03] # ufw status verbose | grep -i log [18:03] Logging: off [18:03] weird. AUDIT suggests loglevel of medium [18:03] (or higher) [18:03] actually, maybe high [18:03] fta: what is LOGLEVEL in /etc/ufw/ufw.conf [18:04] LOGLEVEL=off [18:04] :) [18:04] i checked that already [18:04] fta: did you restart the firewall after adjust that? (or rebooting) [18:05] restart [18:05] fta: if you used 'sudo ufw logging off' you wouldn't have to [18:06] i don't think i did [18:06] fta: did you change /etc/ufw/*rules ny? [18:06] s/ny/any/ [18:06] manually? no [18:06] fta: what version of Ubuntu is this on? [18:06] and it's not supposed to be changed manually, it shouldn't be in /etc.. [18:06] natty [18:07] +if [18:07] fta: it can be changed manually, you just have to manually reload the firewall for it to take effect [18:08] (using the ufw command does that for you) [18:08] fta: is this only for ipv6 packets? [18:08] no, everything [18:10] fta: I'd have to look at your configuration to debug. feel free to email me 'tar -zcvf /lib/ufw /etc/ufw /var/log/ufw.log' [18:10] err [18:10] tar -zcvf fta.tar.gz /lib/ufw /etc/ufw /var/log/ufw.log === fenris is now known as Guest29209 === Guest29209 is now known as e-jat [18:35] chrisccoulson: ping [18:35] hi m_conley [18:36] chrisccoulson: hey - so things with the messaging menu are starting to come together, and I was wondering how I might about about adding it to the Ubuntu packages? [18:36] chrisccoulson: for people to test out - like globalmenu-extension. [18:36] m_conley, i can take care of that [18:36] although, probably not right away. i'm knee deep in libdbusmenu crashes right now ;) [18:37] chrisccoulson: yeah - no worries, no rush on it right now. [18:40] hi nonix4, can you access other sites? === davida is now known as davidascher [19:40] micahg: most other sites yes. Dns seems to be working a bit flakey atm. w3m does connect to bugs.launchpad.net, firefox doesn't. It has been running for 20 days. Trying to connect to "bugs.launchpad.net." with a dot gives an error, since server doesn't understand such request. [19:41] why are you using a . [19:41] For debugging :) [19:42] chrisccoulson: ^^ is this expected?/ [19:42] Since it gives an error, guess it manages to establish SSL session properly with that extra dot [19:42] without the extra dot, it keeps "Loading..." forever. [19:43] Tried changing to private browsing mode, still have that "Loading..." hang when trying without trailing dot on hostname [19:44] nonix4: and using the same host works in w3m? [19:44] yes [19:47] chrisccoulson, http://code.google.com/p/chromium/wiki/LinuxCrashDumping (some specs, you may already know most of it) === BUGabundo is now known as BUGa_TopGear [21:01] ok is anyone running gnome/unity or gnome/classic that can test a sound in gnomne-terminal for me. i would like bug 723936 confirmed or at least find out if it is just me [21:01] Launchpad bug 723936 in gnome-terminal "Gnome-terminal no longer plays sounds" [Undecided,New] https://launchpad.net/bugs/723936 [21:11] micahg: I suspect that restarting firefox could "resolve" the issue, but I'd rather find the cause if possible. Quite interesting that switching to private browsing mode and back doesn't make any difference though. === Brownout is now known as Guest23263 === Guest23263 is now known as Brownout [21:24] version is 3.6.13+build3+nobinonly-0ubuntu0.10.10.1, when pressing enter on https://bugs.launchpad.net/?test url, it only sends a syn, receives synack and sends rst. After that, when pressing "esc" to stop the "Loading..." from spinning, it sends a bunch more - 13 more packets sent & received then. [21:25] nonix4: can you test the version in the security PPA? it's 3.6.14, link in /topic [21:26] well I think the issue disappears once I close the browser, as it worked earlier during those 20 days the browser has been up. [21:26] could try copying profile to a VM or some such though. [21:27] or other user on same host... [21:29] nonix4: did it update while it was running? [21:30] Browser itself not, but libssl yes. Guess that could be it? [21:31] hmm, do you have nscd instaled? [21:33] nope. lsof on firefox-bin doesn't list deleted files besides /etc/passwd and XUL.mfasl; shouldn't dependancies be listed as deleted if they were updated? [21:53] don't seem to be able to reproduce the issue on another user with a copy of the firefox profile. Where was that list of -dbg packages to install in order to attempt gdb:ing the beast? [21:53] nonix4: the list is probably out of date === fta` is now known as fta === debfx_ is now known as debfx === mdeslaur_ is now known as mdeslaur [22:53] micahg: looks like some kind of dns caching issue, but no nscd installed... http://paste.ubuntu.com/571410/ is an LWP right after pressing enter on bugs.launchpad.net URL. Only that one seems to be stuck in non-resolving state, others like code.launchpad.net work properly. [22:55] nonix4: they both resolve to the same place [23:01] micahg: well it seems like negative dns replies/timeouts from around 7 hours ago are somehow being cached by the running firefox process. === brosnan is now known as kbrosnan [23:25] when does the visual refresh hit the ppa? [23:25] or rather, what commit was it? === m_conley is now known as m_conley_away