/srv/irclogs.ubuntu.com/2011/05/17/#launchpad.txt

=== medberry is now known as med_out
=== Ursinha is now known as Ursinha-afk
akoskmhi! I'm trying to build my java packages in launchpad, the problem is that at compile time I have to get a jar file from downloads.sourceforge.net, at this point the build always fails with java.net.UnknownHostException. Any ideas to solve this? Here is the build log https://launchpadlibrarian.net/71131967/buildlog_ubuntu-natty-i386.libqtjambi_4.7.2-0ubuntu7_FAILEDTOBUILD.txt.gz07:41
wgrantakoskm: The builders do not have Internet access.07:43
akoskmwgrant: wow, but they had like 3 month ago, right?:d07:45
akoskmanyway, I'll just include that file with the sources, thank you for pointing out the problem!07:46
wgrantakoskm: No, they never have.07:46
wgrantAllowing arbitrary code execution with Internet access would be a very bad idea :)07:46
akoskmuh07:47
spmautomated hack scan farm ftw!07:49
=== henninge changed the topic of #launchpad to: https://launchpad.net/ | Help contact: henninge | Launchpad is an open source project: https://dev.launchpad.net/ | This channel is logged: http://irclogs.ubuntu.com/
ramanahi09:31
ramanaa query about bug watchers in Launchpad .09:31
ramanahttps://bugs.launchpad.net/binutils-linaro/+bug/772200 is a case where a bug watch was added in lp to an upstream binutils bugzilla. How is it that comments made in that bug don't seem to reflect in the lp bug report ? I've seen other bug watchers added to the GCC bugzilla where comments made on the bug entry make it back into lp09:32
ubot5Ubuntu bug 772200 in Linaro Binutils "Large immediate for neon vmov doesn't work correctly" [Medium,Triaged]09:32
henningeallenap: Hi! Can you help ramana regarding bug watches? ^^09:43
allenaphenninge: I'll give it a go :)09:44
henningeThanks!09:44
jfiHello, is there a way to maintain the consistency between 2 http requests using 'next_collection_link'? Is it scoped by the http session?09:45
henningejfi: I always thought statelessness was one of the defining qualities of a restful service.09:48
henningejfi: Or what do you mean by "consistency"?09:48
allenapramana: It seems to be because Launchpad can't detect an XML-RPC at http://sourceware.org/bugzilla/. I'm going to dig a bit further to see if there's something obvious I can do about it. It might be a bug on Launchpad, or it could be that the remote tracker needs a config change.09:49
allenaps/an XML-RPC/an XML-RPC endpoint/09:49
ramanaallenap: ok cool :) thanks for looking at it . It just seemed to be something missing when I looked at it in LP .09:50
jfihenninge, first request get 0 to 74th first items, then something modify the result which shift the 74th item to 75th, so the second request will miss one item and get twice the 74/75th item09:50
wgrantjfi: There's no direct way to maintain consistency there at the moment.09:50
wgrantjfi: Most consumers cope OK.09:50
wgrantBut once bug #752153 is fixed, it will be consistent.09:50
ubot5Launchpad bug 752153 in Launchpad itself " batch navigation uses high OFFSET rather than db constraints" [Critical,Triaged] https://launchpad.net/bugs/75215309:50
jfiwgrant, I guess there is no way to exceed the '300 limit'? no way to request an unlimited result?09:52
wgrantjfi: No, that would sort of nullify the purpose of having batches: to maintain reasonable request time limits.09:52
jfiwgrant, yes, I understand the reason09:53
jfiwell, I will manage the consistence in my code, it may miss (in theory) some items but it is possible to avoid duplicated items09:55
jfithanks for your response09:55
allenapramana: The remote bug tracker is returning an internal error when Launchpad tries to probe for an XML-RPC endpoint. I'll email someone at sourceware.org to let them know.10:04
ramanaallenap: Thanks for looking into this .11:01
allenapramana: No worries.11:03
sfunkeHi, my name is Simon Funke and I am a developer for the fluidity code (https://code.launchpad.net/~fluidity-core). In order to do some tests, I created a experimental branch  (https://code.launchpad.net/~simon-funke/fluidity/automatic_differentiation) to which I would like to commit some changes now. Unfortunately I can not commit to this branch because it is locked. The exact error message is:11:38
sfunkesf1409@doodson:/data/sf1409/src/fluidity_adjoint/adjoint2$ bzr commit11:38
sfunkeUnable to obtain lock  held by simon-funke@bazaar.launchpad.net11:38
sfunkeat crowberry [process #17283], acquired 24 minutes, 50 seconds ago.11:38
sfunkeSee "bzr help break-lock" for more.11:38
sfunkebzr: ERROR: Could not acquire lock "(remote lock)": bzr+ssh://bazaar.launchpad.net/~simon-funke/fluidity/automatic_differentiation/11:38
sfunkeDo you have an idea what is going wrong here?11:39
henningesfunke: it looks like a previous run of bzr died11:40
henningesfunke: or is still running11:40
henningesfunke: did you look at "bzr help break-lock"?11:40
sfunkea previous bzr process on my machine or on launchpad?11:41
henningesfunke: on your machine11:41
henningesfunke: bzr locks the branch when it works on it. If it gets interrupted doing that, the lock will not be removed.11:42
sfunkehenninge: I was not sure if bzr break-lock is safe safe to use.11:43
sfunkehenninge: thanks, i will try it again then11:43
spivsfunke: the lock is owned by your bzr client11:43
spivsfunke: if your old bzr process isn't running, then it's safe to break its lock11:44
henningesfunke: what spiv  said11:44
sfunkeI checked that no bzr process is running and broke the lock. the commit worked fine now11:45
sfunkei thought launchpad is still doing background work and that is was causing the lock, but it was indeed my machine11:46
sfunkethanks11:46
henningesfunke: np11:46
wgrantsfunke: It says "held by simon-funke", so it was a client authenticated as you.11:46
henningeadeuring: ready to take over?12:04
adeuringhenninge: sure12:04
henningeadeuring: cool ;)12:04
=== adeuring changed the topic of #launchpad to: https://launchpad.net/ | Help contact: adeuring | Launchpad is an open source project: https://dev.launchpad.net/ | This channel is logged: http://irclogs.ubuntu.com/
* henninge lunches12:04
=== henninge is now known as henninge-lunch
=== zyga-afk is now known as zyga
=== henninge-lunch is now known as henninge
pfarrellhi! I have a project on launchpad (libadjoint) which has two recipes building from the trunk. the two recipes are the same (I think), except they go to two different ppa's. However, one builds very successfully (https://code.launchpad.net/~libadjoint/+recipe/libadjoint-daily) and the other one failed to upload (https://code.launchpad.net/~pefarrell/+recipe/libadjoint-daily-fluidity).14:19
pfarrellI got an email saying "File libadjoint_0.1-0~207~maverick1.tar.gz already exists in Fluidity Supporting Software, but uploaded version has different contents."14:20
pfarrellbut I don't really understand it14:20
wgrantpfarrell: It's trying to upload a package that has already been uploaded.14:22
wgrantIn particular by an earlier build of that recipe, 21 hours ago.14:22
wgrantThe version number has no changed, so it cannot be successfully built again.14:22
pfarrellah, I see. so this is only caused because I requested a manual build first, and then when the daily build built it, it couldn't upload it?14:22
pfarrellso now, if I leave it alone, it won't happen again?14:23
wgrantThat's correct, as long as your version template includes all relevant revision numbers, so a change to any branch will change the version.14:24
wgrantIn this case it's fine.14:24
pfarrellyep14:24
pfarrellgreat, thanks for clearing that up14:24
pfarrellnext question (sorry about this, still learning launchpad). Is there a way to formally associate a bug report with a branch that fixes it? In this case, I have a bug (https://bugs.launchpad.net/fluidity/+bug/783987) which I think is fixed in a branch (lp:~pefarrell/fluidity/spatialindex-mktemp), and I was wondering if there's a better way of associating them than just saying-so in a comment14:28
ubot5Ubuntu bug 783987 in Fluidity "Spatialindex uses dangerous functions, which causes compile warnings" [Wishlist,New]14:28
spivpfarrell: click "link a related branch" on the bug's page14:28
pfarrellahh14:29
pfarrellsorry14:29
pfarrellit's obvious when you say it14:29
spivpfarrell: (or "Link a bug report" from the branch's page)14:29
=== med_out is now known as medberry
spivpfarrell: also, if you add "--fixes=lp:783987" to your "bzr commit" invocation that will happen automatically as soon as Launchpad sees that commit14:30
pfarrellspiv: good to know, thanks14:51
deryckadeuring, tag I'm it. :-)15:07
=== deryck changed the topic of #launchpad to: https://launchpad.net/ | Help contact: deryck | Launchpad is an open source project: https://dev.launchpad.net/ | This channel is logged: http://irclogs.ubuntu.com/
adeuringderyck: thanks!15:07
pfarrellwhat does being the "help contact" imply?15:09
deryckpfarrell, if you need help here, you can ping the help contact directly.15:10
deryckand we also watch IRC a bit more closely during our shift.15:10
pfarrellah, fair enough, thanks15:11
derycknp15:11
karniAhoy guys! We're writing the Android application for Ubuntu One Files and we're looking for a solution to annonymously collect bug reports (any crashes whatsoever). Would there be a possibility for an annonymous, not logged-in user to submit a bug report with an attachment via POST to launchpad?15:14
karnideryck: hey mate, I see you in the topic :) Do you happen to have an answer to my question ↑ ?15:15
deryckhi karni.  reading it now....15:16
karnideryck: Thanks!15:16
karnideryck: We might use the Oops thingy, just exploring possibilities here :)15:16
deryckkarni, is this question "do we (or would we) allow anonymous users to post bugs" ?15:17
karnideryck: heheh. I know we currently don't allow that. I'm asking if there a possibility if we could :)15:18
deryckkarni, yeah, we don't support it now.  I'm not sure if we would.  it would require some work on our end....15:18
deryckkarni, you probably want to chat with jml about this if you need it.  but this will slow you down obviously.15:19
karnideryck: I believe so. I see. All right then, we'll look into the Oops'es thingy then first :) Thanks deryck !15:19
ScottKkarni: Set up a role account to accept the bug reports and preauthorize the app to use LP for that account.  Then they aren't "anonymous".15:19
karniScottK: interesting :)15:20
deryckkarni, np.  and yeah, as ScottK notes, you just need *an* account to submit bugs.15:21
karnideryck: okey :)! /me looks into lp help15:21
ogra_so i have this open team on LP ... and just thought it would be nice to have a PPA for it ....15:40
wgrantogra_: That doesn't make sense.15:41
ogra_.... i clicked on "Create PPA" and added the details15:41
wgrantOpen team with PPA == free root for everyone.15:41
ogra_to then get told that open teams cant have ppas15:41
ogra_why is the UI offering me to create a PPA if that kind of team cant use it ?15:41
ogra_(i wasnt actually aware its an open team)15:42
wgrantProbably because the restriction was only added a few months ago, and nobody has pointed out that the link should be hidden :(15:42
wgrantCould you file a bug?15:42
ogra_will do15:42
awolfsonderyck, Hi. I can not run my scripts using "-s staging" - getting error 401 unauthorized. They work on production and qastaging. Is there any way to reset my credentials? I tried to remove all authorized apps from https://staging.launchpad.net/~awolfson/+oauth-tokens but it did not help16:28
deryckawolfson, you can also try removing the credentials from $HOME/.launchpadlib16:29
awolfsonderyck, I did that as well16:29
deryckawolfson, ok, that seems weird.  when you run the script you don't get prompted to re-auth?16:30
awolfsonderyck, no I don't16:30
awolfsonderyck, that is what I hoped for to reauthorize, but it did not happened16:30
deryckawolfson, do you have anything stored in the gnome password manager, or whatever it's called?16:32
awolfsonderyck, good question - let me check16:32
awolfsonderyck, at least in passwords and encription keys there is nothing with launchpad16:34
deryckawolfson, and you're running these scripts via terminal?16:34
awolfsonderyck, yes16:34
deryckawolfson, you try logging out on staging?16:35
awolfsonderyck, you mean from the web browser?16:35
deryckawolfson, yes, just log out via web browser at staging, and try again.16:36
deryckwell, that doesn't make sense, sorry :-)16:36
deryckthe credentials have to be cached somewhere16:36
deryckawolfson, can you pastebin me the output from terminal when you run it?16:36
awolfsonderyck, sure one sec16:37
awolfsonderyck, https://pastebin.canonical.com/47541/16:38
deryckawolfson, can you debug the script and figure out what it's using for cachedir?  that "unknown access token" implies cached credentials somewhere on your system.16:40
awolfsonderyck, OK let me do it.16:40
deryckbenji, can you look at the pastebin above ^^ and see if my advice to awolfson is sound?16:41
deryckbenji, I'm guess cached credentials somewhere, but he's blow away .launchpadlib and revoked oauth tokens.16:41
deryckmaybe there's some other obvious thing I don't know?16:42
benjilastlog awolfson16:42
benjiheh, that needs a slash to be a command16:42
deryckheh16:43
awolfsonbenji, what do you mean by lastlog?16:43
benjiI was trying to run the command "/lastlog awolfson" to see what you had said earlier (it had scrolled off my screen) but I left off the leading slash so I said it instead16:44
benjithat's the IRC equivalent of "did I just say that out loud?!" :)16:45
awolfsonbenji, thanks, I did not know that command - good to know16:46
benjiawolfson: what version of launchpadlib are you using?16:48
awolfsonbenji, how to find that out?16:48
awolfsonbenji, deryck For cache i have: cachedir = os.path.join(os.environ["HOME"], ".launchpadlib/cache")16:49
awolfson    lp = Launchpad.login_with("lp-fish-tools", options.lpsystem, cachedir)16:49
benjiawolfson: dpkg -s python-launchpadlib | grep Version16:49
deryckawolfson, and if you ls that dir you don't show any credentials?16:50
deryckbenji and I are doubling up on awolfson :-)16:50
awolfsonderyck, benji 1.9.7-0ubuntu216:51
awolfsonand I cleaned that dir16:51
benjiawolfson: in that case the credentials are stored in the gnome keyring, let me see how you clear those16:51
awolfsonbenji,looks like I found it under network password16:53
awolfsonlet's see if I can clean it16:53
awolfsonbenji, deryck It solved the problem. Thank you for help! Details: in system settings->Passwords and Encription Keys there are 3 entries named "network password" in details you can see which one was for staging and delete it16:57
benjiawolfson: great!16:57
deryckawolfson, awesome!16:57
awolfsonbenji, deryck thanks again16:57
=== beuno is now known as beuno-lunch
pooliethat's a known bug i think17:48
poolieit can get corrupted17:48
pooliehttps://bugs.launchpad.net/launchpadlib/+bug/74580117:48
ubot5Ubuntu bug 745801 in launchpadlib "system-based authorization doesn't store useful credentials in gnome-keyring" [Critical,Triaged]17:48
primes2hHello kiko. I had a look at this bug #176558 and I wonder if it's possible to have (if any) a svg or whatever of those icons.17:49
ubot5Launchpad bug 176558 in Ubuntu QA Website "New icons required" [Medium,Triaged] https://launchpad.net/bugs/17655817:49
kikoprimes2h, I think we do have SVGs for them, but they have a special license -- joey, flacoste?17:50
joeykiko: the only ones I'm aware of the LP Icons that we had created. Canonical owns the copyright for them.17:51
primes2hkiko: in the bug you mention they have (c) Canonical17:52
joeyand flacoste would be the Canonical rep for them at this point... or mrevell17:52
joeythey are not specifically addressed under https://help.launchpad.net/Legal17:52
primes2hkiko: I would like to start working at that bug, so it would be nice to have SVGs.17:54
deryckabentley, handing over to you now.18:01
=== deryck changed the topic of #launchpad to: https://launchpad.net/ | Help contact: abentley | Launchpad is an open source project: https://dev.launchpad.net/ | This channel is logged: http://irclogs.ubuntu.com/
=== matsubara-afk is now known as matsubara-lunch
primes2hjoey: so should I ask flacoste or mrevell about this?18:20
joeyprimes2h: either.  Matt R is probably a good person to email18:27
primes2hjoey: kiko: ok, thank you very much. :-)18:28
=== beuno-lunch is now known as beuno
=== yofel_ is now known as yofel
flacosteprimes2h: you can find the SVG in lib/canonical/launchpad/images/src/ in LP source tree18:35
primes2hflacoste: do I have permission to use them in the tracker?18:37
flacosteprimes2h: as long as you are fine that they are not redistributable18:38
flacosteprimes2h: i can grant the right to use it on the qatracker site, but if that site is free-software users of it wouldn't have that right18:39
flacosteprimes2h: if you want more, i'll have to escalate the request18:39
primes2hflacoste: sure.But how can I handle this for use them in the tracker? Do I have to put the license in the qatracker-website source?18:42
primes2hflacoste: about more, you mean for use them on other projects?18:42
flacosteprimes2h: yes, or making them open source18:43
flacosteprimes2h: see the LP license18:44
primes2hflacoste: I'll have a look at it later. I must go now. Thanks for now. If I need more help I'll ping you here ;-)18:45
flacosteprimes2h: no problem18:46
ryehi, as I see lp has started using support U1 RT, but there appear to be tons of incoming messages to launchpad feedback that do not pass any spam filtering and just end up creating more and more tickets in the queue18:51
=== medberry is now known as med_out
=== matsubara-lunch is now known as matsubara
=== deryck is now known as deryck[lunch]
flacosterye: yep, that's a known problem, we notified IS of this19:55
ryeflacoste, thanks19:55
=== deryck[lunch] is now known as deryck
abentleylifeless: bug 704698 seems to be fixed.  I recently had to deal with fallout of transaction killing in BranchUpgradeJob.  Do you agree?20:05
ubot5Launchpad bug 704698 in Launchpad itself "Scripts should run with a database transaction timeout" [Critical,Triaged] https://launchpad.net/bugs/70469820:05
abentleylifeless: bug 777958 is the one where killing transactions caused oopses.20:07
ubot5Launchpad bug 777958 in Launchpad itself "branch upgrade jobs keep transaction open" [Critical,Fix released] https://launchpad.net/bugs/77795820:07
=== med_out is now known as medberry
keks-nsup20:42
keks-nGuys, I have some problems with uploading to ppa20:42
keks-ndput says "dput says "Successfully uploaded packages.", but I cann't find my package in the build queue20:43
jdobrienbeuno, do you have anyone sharing a folder with you?20:45
jdobrienoops20:45
ftaspam in bug 29646320:54
ubot5Launchpad bug 296463 in prism (Ubuntu) "Ctrl-F "find" does not work in Google Docs under Prism" [Undecided,Confirmed] https://launchpad.net/bugs/29646320:54
abentleykeks-n: I don't know, I'll find someone who does.20:56
keks-ntnx20:56
abentleykeks-n: it looks like the right people aren't available right now.  Could you file a Question, and I'll assign it to the appropriate people?20:59
keks-nWhere can I do it? I cann't find the form at answers.launchpad.net21:02
keks-nfound it21:05
keks-nhttps://answers.launchpad.net/ubuntu/+question/15794821:09
lifelessabentley: looking21:20
lifelessabentley: for bug 704698 I'm thinking larger scale - we have many different scripts (the email ones, rosetta translations, garbo etc21:25
ubot5Launchpad bug 704698 in Launchpad itself "Scripts should run with a database transaction timeout" [Critical,Triaged] https://launchpad.net/bugs/70469821:25
lifelessabentley: all of them need to have some insurance against long transaction21:25
abentleylifeless: are you saying we kill some transactions but not others?21:26
lifelessah, I'm talking about two different mechanism21:27
lifelessthe transaction killer is a global backstop21:27
lifelessI'm talking about a timeout mechanism like the appservers have21:27
lifelesswe might be able to use the transaction killer to implement that, I hadn't thought about that; I'm not sure if its fine grained enough to have different policies for different users21:28
lifelesswhich I imagine we'll need as many scripts would fail today.21:28
abentleylifeless: so far, it seems we want to restrict scripts'/jobs 1. memory use 2. duration 3. transaction duration.21:33
lifelessyes, though with 2 that limit might be several days21:34
lifeless(for the specific case of upgrade jobs)21:34
abentleylifeless: I don't have any examples of upgrade jobs taking longer than two hours, but yeah.21:34
lifelessthats cool21:35
abentleylifeless: for merge proposal diff generation, we limit it to 10 minutes.21:35
lifelessI thought hairy things like mysql5 still took many hours21:35
abentleylifeless: We may be past all those.  The logs don't go back terribly far.21:35
lifelessabentley: I don't know if they have migrated yet; they have some inertia to overcome aiui21:36
lifelessanyhow, most things don't need such high constraints21:36
lifelesswe have a couple of 12ish hour scripts that run nightly21:36
lifelessbut they do lots of little xactions21:36
abentleylifeless: anyhow, perhaps we should be looking at generalizing those things.21:37
lifelessI think we can do 1 and 3 easily because we have appropriate places to hook in already21:38
lifeless2 might be harder for /some/ of our backend things that aren't in the jobs framework, for instance21:38
lifelessI think 1 and 3 matter the most - if they are in place its hard to take out the rest of the system21:39
lifelessI've no objection to enforcing all three21:39
lifelessjust doing a little analysis about it21:39
abentleylifeless: I agree it would be harder.  I suppose we could rig up something to run scripts under ampoule.  But it seems like we want to prevent runaway scripts as much as runaway jobs.21:57
=== abentley changed the topic of #launchpad to: https://launchpad.net/ | Help contact: - | Launchpad is an open source project: https://dev.launchpad.net/ | This channel is logged: http://irclogs.ubuntu.com/
broderhmm...is there any way to suppress e-mails about successful daily recipe builds? they're kind of spammy22:26
lifelessbroder: there is a bug open22:45
lifelessI think poolie has a patch even22:46
broderlifeless: indeed - i see that now, thanks22:47
=== matsubara is now known as matsubara-afk

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!