[09:38] <iceroot> hi
[09:38] <iceroot> i found a bug in a main-package which is managed through "bzr". what is the way to help fixing? a patch, a debdiff, an upload by me into "bzr"?
[09:39] <iceroot> bug is in command-not-found https://bugs.launchpad.net/ubuntu/+source/command-not-found/+bug/925678  just a string needs to be replaced but i think its a good training for me to work on that
[09:39] <ubot4> Launchpad bug 925678 in command-not-found (Ubuntu) "[12.04] command-not-found is showing a wrong version when its crashing (affects: 1) (heat: 6)" [Undecided,New]
[09:42] <ali1234> iceroot: it depends on the package unfortunately
[09:43] <iceroot> ali1234: command-not-found in this case
[09:45] <ali1234> confirming your bug since i can reproduce it
[09:46] <ali1234> also i feel like it shouldn't crash even if LANG is invalid
[09:46] <ali1234> iceroot: there were some sessions about this in the past few days so you might want to go and read over those
[09:47] <ali1234> iceroot: https://wiki.ubuntu.com/UbuntuDeveloperWeek
[09:47] <iceroot> ali1234: do you have a link for me on that?
[09:47] <iceroot> ali1234: great thank you
[09:47] <iceroot> ali1234: i will fill a new bug on command-not-found for crashing when LANG is set incorrectly
[09:51] <ali1234> the instructions for filing a bug are pretty unhelpful too
[09:51] <ali1234> something obviously caught the crash, yet it provides no useful debugging information for you to repost on the bug report. why?
[09:52] <iceroot> i am remebering a speech at the last "linux day in berlin" where someone has a presentation about using bzr to commit patches, of course i cant remember that but i will do some reading on it, thank you
[09:53] <iceroot> i dont get exactly what you are asking for with the last "why"
[09:54] <ali1234> well if it crashed, what was the error message?
[09:54] <iceroot> ali1234: the one i posted on the bug "with the wrong version number"
[09:54] <iceroot> thats al
[09:54] <iceroot> l
[09:54] <ali1234> i know
[09:54] <ali1234> that's rubbish
[09:55] <ali1234> btw you can strace it like this:
[09:55] <ali1234> strace sh LANG=en vii foo
[09:55] <ali1234> and strace will show you the error message which says:
[09:55] <ali1234> "Can't open LANG=en"
[09:55] <ali1234> which is actually helpful
[09:56] <ali1234> but then that little supposedly helpful instruction about posting a bug report throws that one useful piece of information away
[09:56] <iceroot> ali1234: but that message is coming from sh not from command-not-found
[09:56] <iceroot> imo
[09:56] <ali1234> ah yes, you're right, ok let me retry that
[09:58] <iceroot> do we have something like "set -x" for python? so i can put that into /usr/lib/command-not-found
[09:58] <iceroot> + /usr/bin/python /usr/lib/command-not-found -- vii
[09:58] <iceroot> that is the call of command-not-found
[09:58] <ali1234> no, don't think so
[09:59] <ali1234> looks like command-not-found is actually catching the exception and then hiding it
[09:59] <iceroot> first of all i will open two new bugs, command-not-found not giving usefull output for the bug and crashing with wrong LANG
[10:00] <ali1234> from CommandNotFound.util import crash_guard <- here the bug will be found
[10:02] <ali1234> ok
[10:04] <ali1234> actually i don't understand why it doesn't print the exception/backtrace
[10:05] <ali1234> oh wait, yes i do, there's a double nested try
[10:06] <ali1234> settings the LANG wrong not only crashes command not found but it also crashes the crash dumper
[10:07] <iceroot> so its not a command-not-found bug if i get you correctly
[10:07] <ali1234> it is, cos the crash dumper is part of command-not-found
[10:07] <ali1234> they are trying to be clever and printout a message when it crashes
[10:07] <ali1234> but it is crashing so bad that the helpful message isn't printing
[10:08] <ali1234> but because they already caught the exception, NOTHING gets printed
[10:13] <iceroot> so now we have https://bugs.launchpad.net/ubuntu/+source/command-not-found/+bug/925678
[10:13] <ubot4> Launchpad bug 925678 in command-not-found (Ubuntu) "[12.04] command-not-found is showing a wrong version when its crashing (affects: 1) (heat: 6)" [Undecided,Confirmed]
[10:13] <iceroot> https://bugs.launchpad.net/ubuntu/+source/command-not-found/+bug/926569
[10:13] <ubot4> Launchpad bug 926569 in command-not-found (Ubuntu) "command-not-found crashing when LANG is set incorrectly (affects: 1) (heat: 6)" [Undecided,New]
[10:13] <iceroot> https://bugs.launchpad.net/ubuntu/+source/command-not-found/+bug/926577
[10:13] <ubot4> Launchpad bug 926577 in command-not-found (Ubuntu) "command-not-found is hiding exception/backtrace when crashing (affects: 1) (heat: 6)" [Undecided,New]
[10:13] <iceroot> just because i misstyped LANG and vi :)
[10:15] <ali1234> bugs tend to clump together
[10:15] <iceroot> ali1234: but they are different (imo)
[10:16] <ali1234> the second two are arguably the same
[10:16] <iceroot> not really (imo)
[10:16] <ali1234> you can't really prevent crashing on a bad locale
[10:17] <iceroot> the first is that its not acting correctly on a wrong LANG, the second is when the crash dunper fails its stealing the backtrace (which can happen on other bugs then locale too)
[10:18] <ali1234> can it?
[10:18] <iceroot> and gtk is acting correctly on a wrong locale
[10:18] <iceroot> Gtk-WARNING **: Locale not supported by C library. Using the fallback 'C' locale.
[10:18] <ali1234> it seems to happen specifically because of the locale bug
[10:18] <iceroot> ali1234: but maybe there are other case4s to i can crash it
[10:19] <ali1234> maybe........
[10:19] <iceroot> so there should be a general fix that crash dumper is not stealing anything
[10:20] <iceroot> but that is just my opinion and maybe you are wrong its all just because of the incorrect locale
[10:20] <ali1234> i think instead of trying to manually generate a backtrace it should catch the exception, print the friendly message, and then just reraise the exception
[10:20] <iceroot> ali1234: that is what i wanted to say :)
[10:22] <iceroot> ali1234: maybe you can put that in your words to the bug? maybe that is not confusing so much instead of me guessing something on that (with my english)
[10:23] <ali1234> yeah trying
[10:23] <iceroot> ali1234: thank you
[10:23] <ali1234> it's actually generating all kinds of weird errors rght now
[10:23] <ali1234> the crash dumper actually crashes because: %d format: a number is required, not sys.version_info
[10:23] <ali1234> which makes very little sense to me
[10:25] <ali1234> ah so this has absolutely nothing to do with the locale
[10:26] <ali1234> oh, no wrong again
[10:42] <iceroot> ali1234: thanks for your debug, would be great to append that infos on the bug(s) to its getting more clear then my try to debug it
[10:42] <ali1234> already done
[10:43] <iceroot> ah great then my mailserver/lp mailserver is just slow
[10:43] <iceroot> ali1234: thank you for the help on my issues
[17:03] <cornwall> Hi! Can a triager set bug 735250 to wishlist importance? Thanks!
[17:03] <ubot4> Launchpad bug 735250 in compiz (Ubuntu) "Please disable live viewport previews by default for Desktop Wall plugin (affects: 1) (heat: 6)" [Undecided,Confirmed] https://launchpad.net/bugs/735250
[17:17]  * penguin42 looks
[17:19] <penguin42> cornwall: I will set it to wishlist, but to be honest I like it - what would you prefer to happen?
[17:19] <penguin42> (and to be honest I'm not 100% sure which Desktop wall is - but would that be the one I get when doing ctrl-alt-arrow to switch desktop?
[17:20]  * penguin42 misses the previews on the KDE equivalent
[20:22] <cornwall> penguin42, I like it, too, but it's a feature request
[20:22] <penguin42> cornwall: Nod
[20:22] <cornwall> penguin42, yes, it's the little 'preview' thing.
[20:23] <cornwall> penguin42, thanks for marking it. One day I'll apply for membership....
[20:25] <penguin42> no problem
[20:54] <MxxCon> So do i need to provide any more info regarding bug 909996 to have somebody look at it/fix it?
[20:54] <ubot4> Launchpad bug 909996 in b43-fwcutter (Ubuntu) "firmware-b43legacy-installer errors "17: missing" (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/909996
[22:47] <mfisch> If a bug was filed on an ancient version of some code, and no longer occurs, should it be marked as Invalid or as Fix Released (or something else?)
[22:47] <mfisch> wait!  I found the answer
[22:48] <mfisch> Don't know how I missed that on my first read of https://wiki.ubuntu.com/Bugs/Status.
[22:48] <mfisch> carry on everyone
[23:32] <hggdh> mfisch: fix released, and a comment it does not happen anymore on such and such version