[00:03] penguin42, around? hggdh, you too [00:03] nod [00:04] can you read and comment on the bugsquad mailing list discussion put forth by AG Restringere? [00:04] TheLordOfTime: Which what? [00:05] I can't seem to form a logical argument right now... [00:05] https://lists.ubuntu.com/archives/ubuntu-bugsquad/2013-November/004347.html from the first to the last in the thread [00:05] TheLordOfTime: here [00:05] TheLordOfTime: Sigh is this the whole we need to change the way stuff is done one? [00:05] hggdh, same request I made of penguin42 [00:06] they're asking to add a status system for "actively triaged" as well as fields to identify who exactly is triaging a given bug [00:06] TheLordOfTime: Can I admit to having hit the 'd' key after reading the 1st couple of lines? [00:06] sort of like a "Triage Assignment" thing that adds overcomplexity to the system [00:06] penguin42, if you want. HOnestly, I can't tell you what I"m thinking about AG right now because it doesn't conform with the CoC [00:07] but that may be because of the massive headache I have... [00:07] TheLordOfTime: sounds like overkill (this is, thou, quite common on commercial BTSes) [00:07] o/ phillw [00:07] hggdh, "commercial BTSes" [00:07] let's make that point clear. [00:07] TheLordOfTime: just the person I need :) [00:07] phillw, you can wait [00:07] you're next on my list of things [00:07] TheLordOfTime: TBH I have had bugs I've been triaging where other triagers have come along and done something weird to it and wondered WTH? But there again most of the time I only triage at weekends and thus it's right someone picks stuff up - so it's non-trivial [00:07] hggdh, commercial BTSes are different than a community-driven BTS [00:08] a business would need such a setup to identify who is working on a given task [00:08] whether for a project or for helpdesk or whatever else. [00:08] TheLordOfTime: a commercial Bug Tracking System -- a triager would usually assign his/herself to the bug as a triager [00:08] hggdh, true. [00:08] well LP ain't too bad as bug trackers go, it's got many features I've not seen in others including commercial ones [00:09] hggdh, but here's a question that is highly relevant: in the Launchpad system for bugs, how would we implement that and *prevent* the abuse of the permissions, short of having bugcontrol or bugsquad having specific permissions [00:09] but there again I've seen commercial ones set up in ludicrously complex ways [00:09] or, better idea, can you just comment on the mailing list [00:09] while i help phillw with whatever problem he has [00:09] * hggdh thinks of Remedy [00:10] TheLordOfTime: My understanding anyway is that LP is in a strictly maintenance mode, so I'm not sure anything can be changed if people wanted it to be [00:10] I believe that it's overkill here for Launchpad/Ubuntu, hggdh, and made my opinions pretty clear so far, but I'd like more than just me and Alberto making the comments. [00:10] TheLordOfTime: can you mark https://bugs.launchpad.net/ubuntu/+source/firefox/+bug/933726 as what ever is needed to kill it [00:10] Launchpad bug 933726 in firefox (Ubuntu) "Firefox, incorrect language pack on PPC" [Undecided,Incomplete] [00:10] TheLordOfTime: I do not know how to prevent abuse (in an easy way). But we do not need, I think, this field [00:10] phillw, and by "kill it" you mean make it so the bug is just "Dead"? [00:10] It's outdated, will not be fixed etc. [00:10] TheLordOfTime: yup [00:10] TheLordOfTime: wont-fix [00:10] hggdh, i was just about to say that :p [00:11] :-) [00:11] phillw, give me a moment to relogin to llaunchpad, i just wiped cookies [00:11] then i'll Won't Fix it [00:11] TheLordOfTime: that's the one! we mere mortals cannot set that flag! [00:13] phillw, done [00:14] phillw, i would add a comment that it won't be fixed though, and state the reasons therein [00:14] since i just set the status, you requested it :P [00:15] TheLordOfTime: is https://bugs.launchpad.net/ubuntu/+source/firefox/+bug/933726/comments/5 okay? [00:15] Launchpad bug 933726 in firefox (Ubuntu) "Firefox, incorrect language pack on PPC" [Undecided,Won't fix] [00:16] phillw, works for me. any issues with that, hggdh? [00:16] none at all [00:16] cool. [00:17] It was Ibere who flagged it up, i just got to ask :) [00:17] TheLordOfTime: Iwill look at all email from bugsquad as soonas I have recovered my system (just got a new laptop) [00:17] hggdh, ack [00:18] hggdh, actually, robru just responded to the mailing list [00:18] i think he made the point faster than you did :) [00:23] is there some QA team which I can subscribe to a bug report? shipping a broken library in a final release isn't that great, no one seems to care to actually fix it [00:23] bug 1183580 [00:23] Launchpad bug 1183580 in librcc (Ubuntu) "librcc segfaults on latest saucy" [High,Triaged] https://launchpad.net/bugs/1183580 [00:27] hmm that's in main as well [00:28] brainwash: I'd suggest pinging bdmurray so that he can have a chat with the release team as to how this bug should be allocated. But, that is only I.M.H.O. approach to a bug that seems to be affecting many. [00:29] bdmurray seems to be busy all the time :) [00:30] well there is lots of stuff to do [00:30] brainwash: It's a shame that being the main bug it doesn't have any backtrace in it [00:30] brainwash: he has to put up with all of us :D [00:31] bdmurray: hi boss, [00:32] penguin42: the source code is not buggy, the way the package is build is somewhat broken [00:32] bdmurray: https://bugs.launchpad.net/ubuntu/+source/librcc/+bug/1183580/comments/23 does seem to have a solution, is it worth trying out and getting SRU'd? [00:32] Launchpad bug 1183580 in librcc (Ubuntu) "librcc segfaults on latest saucy" [High,Triaged] [00:32] brainwash: Not necessarily [00:33] brainwash: The fact it doesn't seg on debian doesn't mean it's a build issue; it just means that whatever happened didn't trigger in debians build [00:34] I got told that syncing a working deb package is not an option [00:34] phillw: I wouldn't call comment #23 a solution more of a workaround or hack [00:35] brainwash: It needs to *understand* why the seg happens and then fix it [00:37] somebody should try what was asked in comment #19 [00:37] bdmurray: rebuilding does not help [00:39] I just want to make sure that someone (a dev with the needed knowledge) takes a look at this issue [00:40] xnox is a developer with that knowledge [00:40] right, he already commented on this issue :) [00:40] bdmurray: I'm away on Wednesday, but will happily do testing for it if needed from Thursday. [00:41] phillw: its really trivial to test so I'll just have a poke at it [00:41] phillw: comment/23 is not a solution. [00:42] * penguin42 gets builddep and kicks a build [00:42] penguin42: rebuilding the package did not solve segfault. [00:42] xnox: Yeh not surprising [00:43] xnox: Problem is looking at the code it's got a data structure and it's landed at a place where it's got a null pointer - without pulling it apart and understanding wth it's doing it's non-trivial; it's not a trivial screw up at the point of the seg [00:44] it's almost entirely uncommented as well - albeit it with flowery function names [00:47] bdmurray: I hope I have not been "out of order" by raising this personally to you, but the bug does seem to affect quite a few people and I guess what they would like to see is "Assigned to" == " Unassigned" to actually be assigned to some one :) [00:50] penguin42: hm, maybe recompile with -O0 ? [00:52] xnox: It's not fair to blame the compiler yet - it might just be a good old fashioned screw up in the code (says he noting that arch have librcc 0.2.10-3 and we're on a 0.2.9) [00:55] mmh, take a look at http://moc.daper.net/node/928 [00:56] "Recompiling moc --without-rcc makes moc working in (X)Ubuntu." [00:57] would fix moc, but not the library =S [00:58] xnox: I mean if you think it is a compiler bug (which IMHO there is no proof of since it's just falling into a NULL pointer that's set elsewhere and it seems to have various places it could set that) then you could try -O0'ing it and see if it helps [00:58] xnox: If it does help then you try -O0ing one file at a time and find which file does it [00:59] penguin42: it's not a compiler bug, but rather that optimisations may expose bad programming bugs. [00:59] gawwwd.1700 new emails [00:59] xnox: maybe, but yeh you could try that, but Debian uses different compiler versions and flags so it could be any one of a bunch of things triggering the difference [01:00] * penguin42 hands hggdh ctrl-D [01:00] as in complier says "i got so smart, and this code says i can skip that part/inline/etc and it should work none the less", where actually it's a programming bug. [01:00] * hggdh grabs Ctrl-d before it vanishes [01:00] penguin42: sure, it's called bisect =) we have a good one and a bad one ;-) [01:01] xnox: Yep, and the newer compilers do get smarter and let less dodgy code through [01:01] gah now you have to focus to see whether you're going to click on amd64 or arm64 [01:04] xnox: Might also be worth seeing if the ./configure on debian ends up taking the same paths - I can't see a debian buildlog for the package [01:05] xnox: There is also a warning in the build about a write above array bounds with the code just being straight wrong [01:06] * penguin42 doubts that's the problem but it's a trivial fix, rccstring.c line 67 [01:16] nope, doesn't seem to help - oh well, bed soon === Jikan is now known as Jikai === Jikai is now known as Jikan === Jikan is now known as Jikai === Jikai is now known as Jikan [04:01] Please could someone accept my nomination for Precise in bug 1248394? It's fixed since at least Saucy, but I need an SRU. [04:01] Launchpad bug 1248394 in libvirt (Ubuntu) "Python virStorageVolUpload binding fails on i386" [High,New] https://launchpad.net/bugs/1248394 === mapreri_ is now known as mapreri === Ursinha_ is now known as Ursinha [11:46] Hello, I am trying to fix one minor bug in launchpad, I am trying to merge my branch in to another branch, but not able to propose a merge, option for proposing a merge is not seen , Can anyone please help me [13:36] rbasak: approved === LordOfTime is now known as TheLordOfTIme === TheLordOfTIme is now known as TheLordOfTime === brainwash_ is now known as brainwash === shadeslayer_ is now known as shadeslayer [15:22] hi, does anyone know the utf8 filename issue of unzip in 12.04-13.04 ? [15:32] maxiaojun: I have seen it, a long time ago. There is a bug on it [15:34] i'm discussing this with someone in #ubuntu-devel [15:42] hggdh: you in -devel? [15:56] TheLordOfTime: yes [15:56] hggdh: yeah just noticed, i might've pinged you once or twice, i know bdmurray's alive and kicking in there, too :) [15:57] heh [16:25] hggdh: thank you! [18:49] hggdh: or anyone with knowledge, what needs to be removed from a private crash bug or a private bug in order to make it able to be "public" in most typical cases? [18:50] relevant question: http://askubuntu.com/questions/368817/what-to-do-with-private-bug-reports-on-launchpad [18:50] TheLordOfTime: memory dump, and check if any potential private data is shown in the backtraces [18:51] hggdh: and if there is data in the stacktrace that's private, then what? [18:52] trying to write a good answer for this xD [18:52] TheLordOfTime: then pain: download the affected trace, edit & masquerade the offending data, upload the new version, and remove the old one [19:02] heheheheh [19:02] that's what i thought xD [19:03] hggdh: in the offchance you have an Ask Ubuntu account, i've posted an answer on that question, feel free to review and comment, i can add anything else there that's missing or revise things. [19:04] TheLordOfTime: I *had*. Will have to find it, and the password :-). But will look [19:04] you can look now if you want, it's public, the answer [19:04] if i done goofed on something, let me know [19:04] I did make a note that private security bugs might have different policies, because i'm not on the security team... [19:05] and also that bugs against projects that aren't Ubuntu or a specific ubuntu package have policies set by those projects' managers or teams [19:05] since that is, obviously, true. [19:05] hggdh: are the procedures linked anywhere on one of the Bugs/something wiki pages, though? [19:05] maybe under triage procedures or something? [19:06] they should be. Just a sec [19:11] ok [19:12] TheLordOfTime: https://wiki.ubuntu.com/Bugs/HowToTriage#Apport_reports [19:15] int_ua: hiya! [19:15] hi :) [19:15] hggdh: int_ua is the guy who posted the question on Ask Ubuntu [19:15] TheLordOfTime: reading the wiki [19:15] as well as a sizable list of bugs... [19:16] 'course i trimmed that from the question so it becomes a more general, canonical quesiton, and I can pastebin the list of links or bug numbers if you want. [19:16] TheLordOfTime, the list now includes two more firefox crashes, sadly [19:16] int_ua: FYI, I am multitasking... so i might not be here the whole time. [19:16] TheLordOfTime, ok, acknowledged [19:17] especially since some of my tasks include moderating some other channels... [19:17] whcih get a little rowdy (outside hte Ubuntu namespace) [19:17] TheLordOfTime: awww c'mon, everybody knows moderation is easy ;-) [19:17] hggdh: when you have to balance it with a nearly-full banlist, it gets tricky [19:17] because then I have to trim out old bugs.. [19:17] :/ [19:18] s/bugs/BANS/ [19:18] ... fooey [19:18] heh [19:18] as long as you do not ban a bug, that's OK [19:18] hehehehehe [19:18] however if I bug a ban, that's also bad [19:19] int_ua: if you have a question, please feel free to ask [19:19] TheLordOfTime: indeed [19:19] =D [19:19] indeed, any questions, even about private bugs, are OK here [19:19] and hggdh and I are both... well... alive to help :) [19:20] hggdh, currently I'm reading the wiki link from the answer [19:20] whic is the same one I gave you on the Stack Exchange chat too :) [19:20] you can thank hggdh for digging up the link, i'm still rebuilding my set of links for bug information xD [19:20] since newcomputer and all [19:21] w00t, +40 rep from that answer xD [19:22] TheLordOfTime, you totally won the bounty, but I'll just leave it there for a while to collect some more, ok? ;) [19:23] if you want, i think you have to wait a day before assigning the bounty anyways [19:23] or ask a mod to force-assign [19:23] but i'll gain rep from upvotes either way [19:23] although, kudos to hggdh reminding me about what i forgot about crash bugs xD [19:23] because it's been a WHILE since i've had to handle a valid crash bug [19:23] (there was a PEBKAC crash bug on nginx but that was Invalid because of PEBKAC) [19:26] hggdh: http://paste.ubuntu.com/6372293/ is the list of bugs he had, plus link structures added by me so you can copy-paste them in and just go to the links, in case he has specific questions about each bug [19:26] * TheLordOfTime has to call Comcast and tell them to go die in a fire again for breaking the interwebs [19:27] int_ua: you don't mind that I trimmed the list of bugs off of your question, right? I did that because I wanted to make the question more general so it could be referenced as a general Q&A about private bugs and handling them... [19:27] TheLordOfTime, Can you please very briefly explain how stacktrace can have sensitive information? [19:27] hggdh can, i can't, the comcast person just picked up the phone on their end. [19:27] TheLordOfTime, yeah, no problems with that, of course :) [19:29] stupid comcast put me on hold again... [19:30] s/how stacktrace can/how can stacktrace/ [19:30] int_ua: if a crash has a stacktrace, it can sometimes contain the information entered into the system which then triggered the crash [19:30] in rare cases, I think some stacktraces can contain entered passwords, or other sensitive information, in case of firefox or other crash bugs, or crash bugs where it crashed when transmitting auth data or stuff [19:30] but i personally have never seen those kinds of bugs. [19:31] (most nginx crash bugs, which I primarily handle, end up being because of code issues in the packages, or weird edge-case configs being loaded rather than the transmission of sensitive data) [19:31] TheLordOfTime, can I skip (retraced) attachments while checking for private info? [19:31] int_ua: think of it this way, if I crash a program while giving it a password... [19:31] looking [19:31] the coredump will probably have that and other things [19:32] hggdh: you can answer their last question, i'm going to go punch something hard to get my aggression towards comcast out of the way while i'm on hold) [19:32] wait, there is no such thing [19:32] sorry to dump things on ya, hggdh, but when comcast breaks my internet... : [19:32] :/ [19:32] there are two (basically) options to stacktraces -- I am talking about GDB stacktraces): they can contain variable data values, or not [19:34] to get a stacktrace on gdb, you do 'bt' <- no variables values are shown; or you do 'bt full' <- variable values are shown [19:34] it is the 'bt full' (or the usual cousin, 'thread apply all bt full') that can contain private data [19:36] can anyone increase the importance level of bug 1205384 please? [19:36] Launchpad bug 1205384 in lxsession (Ubuntu) "Lock can be circumvented by switching to console" [High,In progress] https://launchpad.net/bugs/1205384 [19:36] I have just made bug 762087 public, and you can see the effect of 'bt full' there [19:36] Launchpad bug 762087 in freeciv (Ubuntu) "freeciv-sdl assert failure: *** glibc detected *** freeciv-sdl: munmap_chunk(): invalid pointer: 0x0aad0a38 ***" [Medium,New] https://launchpad.net/bugs/762087 [19:37] brainwash: why does this need an increase, out of curiosity? [19:37] (BTW, back from punching the wall, i hurt my hand in the process but meh) [19:37] brainwash: Julien is already working on it, no need to increase importance [19:37] ^ that [19:38] hggdh, sorry, but where exactly can I see the effect? [19:38] right, but nothing happened until now and this is a critical issue which you don't expect in a final release [19:38] brainwash: But it's assigned so that should be enough [19:38] int_ua: the last two attachments -- Stacktrace.txt and ThreadStacktrace.txt -- show variable values [19:39] hggdh, ok, I think I see them, yes [19:39] int_ua: so, depending on which program crashed, these variables may contain private data. (I have seen user and password shown, for example) [19:39] brainwash: Critical is normally reserved for corruption, turning the machine into a brick or the like - see https://wiki.ubuntu.com/Bugs/Importance [19:40] hggdh, what about cookies in firefox stacktraces? Is it possible? Because it would be much harder to notice [19:41] penguin42: isn't this the case if the anyone could gain access to the system if screen locking does not work as intended? [19:41] brainwash: well, there is comment#29 about that, and gilir is -- theoretically -- working on a bypass [19:41] data corruption by simply deleteing files, e.g. [19:42] brainwash: Right, but that's in the High category 'Prevents the application or any dependencies from functioning correctly at all ' [19:42] int_ua: I do not deal with firefoxx, so I cannot answer about that. But I would be really surprised if we were uploading cookies [19:42] brainwash: I agree it's a pretty grim bug [19:43] brainwash: a good place to make your case is #ubuntu-desktop. I *do* agree it is a bad one, but I cannot help making it fixed sooner [19:44] (I have seen arguments in the tone "if They have access to your computer, all is lost". I do not agree, but... [19:45] TheLordOfTime, hggdh, I see my old phone BT address in UbiquitySyslog at bug 817250. But it looks too old to be any useful, can I just delete the syslog? [19:46] brainwash: Given it's 2 weeks since Julien's comment you could ask (gently) if there's an ETA [19:47] int_ua: if you feel the syslog is not needed to help solve the issue, yes. Otherwise, download the syslog, edit out offending data, re-upload, and delete the original. THEN make the bug public [19:50] penguin42: I'm not even using lubuntu, and would never consider using it after the disastrous 13.10 release, I'm just concerned about the users who don't get warned about this issue [19:50] brainwash: your interest in trying to solve it *is* welcome [19:51] even more given that Lubuntu needs help (as all of us) [19:51] brainwash: I agree, but I think 'high' is the right place for it, and in principal with it assigned that should be enough [19:52] penguin42: ok, thanks for clarifying :) [19:53] brainwash, what's so disastrous about it? Just interesting [19:53] brainwash: anyway, addded a comment asking Julien about it [19:53] int_ua: Breaking screen lock trivially? [19:53] oh, just added, not addded... [19:55] int_ua: the screen locking issue is just one major issue, there are still other bugs left and the release feels unpolished and requires the user to manually change default settings [19:57] brainwash: There are always lots of bugs - especially in the non-main release [19:57] brainwash: Remember Ubuntu is a release-on-day release, so if bugs exist it still gets released [20:03] * slickymaster is away: (Dinner time) [20:04] brainwash: Not that's necessarily a bad thing - the alternative is something like debian where you keep going until there aren't any more criticals and get stuck for 2 years [20:04] slickymaster: please turn that off. [20:10] i just had a look at bug 649139 . what can be done to help with a bug with the tag "bugpattern-needed"? i can reproduce the bug reliably, so maybe i can help get some progress... [20:10] Launchpad bug 649139 in usb-creator (Ubuntu) "[12.04] usb-creator-gtk crashed with SIGSEGV" [Medium,Confirmed] https://launchpad.net/bugs/649139 [20:44] TheLordOfTime, hggdh, I've made public all reports except firefox ones, will check them tomorrow, thanks a lot for your assistance :) [20:48] int_ua: you're welcome! [20:48] int_ua: sorry I handed you off to hggdh for primary assistance, i had to fight off COmcast :/ [20:48] glad to see all was handled though :) [20:49] ^_^ [20:52] int_ua: also, feel free to lurk around here and ask any questions you have on bugs, that's what we're here for :) [20:53] (I very rarely work on anything outside server packages, other than generic triage duties, so like hggdh, my ability to help on some bugs is... limited.) [20:53] (but i'll try and help whenever I can :D) [20:53] TheLordOfTime, my last firefox report was marked as duplicate of 1200272. But the last one is private too AFAIU. Can you check that it had some attention? [20:54] one momeny [20:54] moment* [20:55] it's had a "Confirmed" autochange because multiple users are affected, but other than that not really... [20:57] i could look through the stacktrace but i have other higher-priority-attention-needed things at the moment :/ [20:57] int_ua: what was your bug one that was marked as a dupe of 1200272 [20:58] TheLordOfTime, https://bugs.launchpad.net/ubuntu/+source/firefox/+bug/1248673 [20:58] int_ua: Error: Could not gather data from Launchpad for bug #1248673 (https://launchpad.net/bugs/1248673). The error has been logged [20:59] you can ignore ubot2 he can't see the private bugs like bugcontrollers can >:D [21:00] int_ua: yeah the apport retracer can id certain crash patterns and set dupes like that, i'll keep an eye on 1200272 if you want [21:00] (also there's a good number of dupes of this crash so... yeah...) [23:40] Hi good people, is there someone available to look at the status of bug 1219708 it is a major issue for those struck with it, but has no one assigned to it. I'm really sorry to come back for a second request in 24 hours; but these are 'zero' day bugs that I've been made aware of. [23:40] Launchpad bug 1219708 in xserver-xorg-input-synaptics (Ubuntu) "Asus synaptics/elantech touchpad not detected during or after fresh install" [Undecided,Confirmed] https://launchpad.net/bugs/1219708 [23:49] * penguin42 looks [23:50] phillw: Are you one of the people affected? Fancy trying a 13.10 ? [23:51] penguin42: I do not have that hard ware configuration. Had I had, it would have been squealed about a long tine ago! [23:53] phillw: So that looks like a report from one person, are there multiple? [23:55] so far, so one reporter, I'm not sure of dupes for it. Looks to my untrained eye like a regression; hence my asking here. [23:58] penguin42: as I have the OP in IRC, I'll ask that he try 13.10. Thank you guys and gals again for taking the time to chase up a bug. You are all stars. [23:59] phillw: I've added some questions to it