[00:22] <desrt> hi.  i'm having trouble uploading to my PPA
[00:23] <wgrant> desrt: What's it not doing?
[00:24] <desrt> it's uploading, but then i get a rejection email saying that the file is already there
[00:24] <wgrant> The file is already there.
[00:24] <desrt> i deleted it
[00:24] <wgrant> That doesn't matter.
[00:24] <wgrant> You can never upload the same version again.
[00:24] <desrt> it's no longer showing on the PPA page
[00:24] <desrt> that sucks a lot :(
[00:24] <wgrant> Why?
[00:25] <desrt> is it possible for an admin to go in and delete the files?
[00:25] <wgrant> Uploading the same version would be very confusing to users, and confuse apt even harder.
[00:25] <wgrant> Why?
[00:25] <wgrant> Why not just bump the version?
[00:25] <desrt> because nobody downloaded it
[00:25] <desrt> it was a botched upload and now i've fixed it
[00:26] <wgrant> => it's a new version => new version number
[00:26] <wgrant> It really does make things a lot less confusing for everybody.
[00:26] <desrt> is it possible to nuke the entire PPA?
[00:27] <wgrant> You could get an admin to rename and disable it for you.
[00:27] <wgrant> But doing that just to upload the same version again seems like a very bad idea.
[00:28] <desrt> hmm
[00:28] <desrt> "  Package deletion can take some time before the packages are actually removed. See the full PPA help page. "
[00:28] <desrt> shame the help page says nothing on the topic :p
[00:28] <desrt> maybe when i wake up tomorrow it will be gone
[00:28] <wgrant> You still won't be able to reupload.
[00:31] <desrt> it seems exceptionally silly
[00:32] <wgrant> Why?
[00:32] <desrt> because it's my PPA
[00:32] <desrt> why should i be prevented from deleting something that i just uploaded?
[00:32] <wgrant> You were able to delete it.
[00:32] <desrt> what if i accidentally uploaded my password or something
[00:32] <wgrant> You just can't upload something immensely confusing.
[00:32] <desrt> it's still on the archive
[00:32] <desrt> that's the problem
[00:33] <wgrant> It should be removed from the archive disk within a day.
[00:33] <desrt> right.  that's what i hope for.
[00:33] <wgrant> However it will remain accessible from the web UI forever unless you get an admin to remove it.
[00:33] <desrt> the email makes it very clear that the fact that it's on the archive is a problem
[00:33] <desrt> it's already gone from the UI
[00:34] <wgrant> Not if you search for deleted packages.
[00:34] <wgrant> The publishing status refers to the status on ppa.launchpad.net. Nowhere else.
[00:35] <desrt> oh interesting
[00:35] <desrt> there's a bug in the web interface
[00:35] <wgrant> Where?
[00:35] <desrt> you can't search for superseeded packages from the delete page
[00:35] <desrt> so you can delete packages, but only if they haven't been superseeded
[00:35] <desrt> that seems to be the problem
[00:36] <wgrant> That's not the problem.
[00:36] <desrt> it's quite reproduceable
[00:36] <wgrant> +delete-packages will show any packages that haven't yet been removed.
[00:36] <wgrant> Why would you want to delete a package that has already been removed because it was superseded?
[00:36] <desrt> it's still on the archive
[00:37] <desrt> the superseeded package is still on the archive.  the new one is already deleted.
[00:37] <desrt> the web interface lets me search for 'superseded' from the delete page
[00:37] <desrt> but it returns no results
[00:37] <desrt> so there's definitely a bug here.  perhaps 2.
[00:37] <wgrant> No.
[00:37] <wgrant> It's just confusing.
[00:37] <wgrant> Let me explain.
[00:37] <wgrant> .
[00:37] <desrt> ok
[00:38] <wgrant> It's a bit strange, but it does make sense.
[00:38] <wgrant> Packages will show up on +delete-packages if there are bits of them that haven't been removed yet.
[00:38] <wgrant> If a package is marked Superseded or Deleted, it is not removed from disk immediately.
[00:38] <desrt> on the archive or in the PPA webapp?
[00:38] <wgrant> Archive.
[00:39] <wgrant> In the case of a Deleted publication, it will be eligible for removable immediately.
[00:39] <wgrant> In the case of a Superseded one, it will become eligible for removal in 24 hours.
[00:39] <desrt> so the reason that this pacakge is not showing on the +delete is because it's already queued for delete by virtue of having been superseeded?
[00:39] <wgrant> Yes.
[00:39] <desrt> that's very very strange
[00:39] <wgrant> Now, on my PPA some deleted and superseded packages do show up.
[00:40] <desrt> you remove your ability to immediately delete a particular release by making a new release
[00:40] <wgrant> Er, what?
[00:40] <desrt> here was my exact sequence of actions:
[00:40] <wgrant> It's not shown on the page because you don't need to remove it!
[00:41] <desrt> 1) upload pkg0
[00:41] <desrt> 2) upload pkg1
[00:41] <desrt> (ie: pkg1 supercedes pkg0)
[00:41] <desrt> 3) delete pkg1
[00:41] <desrt> 4) try very hard (and fail) to delete pkg0
[00:41] <desrt> 5) note that the archive contains pkg0 (but pkg1 is completely gone)
[00:41] <wgrant> What is the timeline?
[00:41] <desrt> all today
[00:42] <desrt> a few hours between pkg0 and pkg1 since the builders were slow with the alpha build today
[00:42] <desrt> now here's why i think there is at least one bug:
[00:43] <desrt> even if i accept that it makes no sense to want to delete packages that have already been marked as superseded then why does teh +delete-packages page have 'superseded' as an option in the pulldown menu to search for packages to delete?
[00:44] <wgrant> I'm looking at that now.
[00:44] <wgrant> But why do you care?
[00:44] <desrt> (and i don't even accept that premise, which is why i think there may be 2 bugs)
[00:44] <desrt> because it's exceptionally misleading?
[00:44] <wgrant> That still won't help you upload the same version.
[00:45] <wgrant> The main use-case for deleting a superseded package is reducing your disk usage to get under your PPA quota. Normally you'd have to wait for the quarantine period.
[00:46] <desrt> the whole "we refuse to delete your data" mentality is a bit odd to me, i must admit
[02:30] <Flannel> Hi all, what steps do we have to go through to get two teams merged?  One is now defunct, it'd be nice to have it not sitting around.  Do we file a question against launchpad still? or is there a better method now?
[02:31] <wgrant> Flannel: File a question.
[02:31] <Flannel> wgrant: Alright thanks
[02:56] <aboSamoor1> I have personal branch. can I separate to different personal branches ?
[03:00] <Meths> You can branch branches
[03:01] <aboSamoor1> Meths: my work is saved under folders, I want to make a new branch for each folder without loosing the history
[03:01] <wgrant> aboSamoor1: Ah. Have a look at the bzr-split plugin. Also, #bzr is better for this sort of question.
[03:02] <aboSamoor1> wgrant: thanks :)
[03:50] <desrt> wgrant: thanks for your help earlier
[03:50] <desrt> you were right.  launchpad rejected the upload of all but a totally unseen before version number
[03:50] <desrt> weird, but whatever :p
[03:51] <desrt> nite
[05:22] <ub3rst4r> does anyone know how to convert .resx to .pot so they can be uploaded to the translations section
[05:26] <mwhudson> do ppas check if a built package can be installed during the build process?
[05:27] <wgrant> mwhudson: No.
[05:27] <mwhudson> hooray
[05:27] <wgrant> Why?
[05:28] <mwhudson> i have a package that fails to install
[05:28] <wgrant> I see.
[05:28] <mwhudson> i'm uploading a hacked up version of it to launchpad
[05:28] <mwhudson> but it's for hardy, which i don't have lying around
[05:31] <ub3rst4r> does anyone know anything about .pot or how the translations work?
[05:34] <spm> ub3rst4r: the translations guys aren't around yet; perhaps another 2-3 hours?
[05:34] <ub3rst4r> k thanks
[07:12] <Q-FUNK> somehow, it seems that recent chnages in LP have obfuscated the location of the menu to change which package is affected by a bug and help.lp.net isn't helping. would anyone know where that went?
[13:02] <mat_t> hey LP people
[13:03] <mat_t> Been trying to report a bug in Software Store, but "Report a bug" button takes me to https://help.ubuntu.com/community/ReportingBugs
[13:04] <james_w> mat_t: yes, you are now asked to follow the instructions there to report the bug
[13:04] <james_w> ubuntu-bug software-store
[13:04] <mat_t> james_w: oh
[13:05] <mat_t> james_w: does it mean we're actively discouraging people from using LP?
[13:05] <james_w> no
[13:05] <james_w> ubuntu-bug opens an LP page to report the bug, it just attaches lots of juicy information that saves you typing it in
[13:05] <james_w> package versions and the like
[13:05] <mat_t> aaah
[13:05] <mat_t> I see
[13:05] <mat_t> clever :)
[13:06] <james_w> which a lot of people skip giving, so the first interaction is "what version of Ubuntu are you using?"
[13:06] <mat_t> cool
[13:06] <mat_t> I still think this instructions page should be *within* LP
[13:08] <mat_t> james_w: OK, so I'm in Karmic, and I'm trying to report a design/usability bug in Software Store... How do I do it?
 ubuntu-bug software-store
[13:09] <james_w> or "Help -> Report a problem" if there is an option in software-store
[13:10] <mat_t> james_w: yeah tried it, but it doesn't let me enter any information, just sends the report
[13:10] <james_w> https://help.ubuntu.com/community/ReportingBugs#Filing%20a%20bug%20with%20ubuntu-bug
[13:10] <kklimonda> hey
[13:11] <mat_t> james_w: ok thx
[13:11] <kklimonda> is it normal that I don't get emails for new bugs that are set to private for packages I'm subscribed to even if I'm in ubuntu-bugcontrol ?
[13:11] <james_w> mat_t: http://screencasts.ubuntu.com/2009/09/16/Reporting_Bugs
[13:12] <james_w> kklimonda: yes unfortunately
[13:12] <james_w> mat_t: it is supposed to open a browser for you to complete the process, but firefox may be crashing for you as it is for me
[13:12] <wgrant> kklimonda: That's an awkward thing to fix, unfortunately.
[13:18] <z3l> hi everybody
[13:25] <z3l> how can i add a project host on launchpad to the file sources list ?
[13:27] <z3l> i mean the file /etc/apt/sources.list
[13:28] <deryck> wgrant, ping
[13:28] <wgrant> deryck: Hi.
[13:35] <z3l> I read this : https://help.launchpad.net/Packaging/PPA/InstallingSoftware but i can't find "the first line from the apt sources.list entries section of the PPA overview page."
[13:35] <noodles775> z3l: click on the 'Technical information' link...
[13:35]  * noodles775 checks for exact wording...
[13:35] <noodles775> z3l: in the 'Adding this PPA to your system' section.
[13:36] <wgrant> z3l: How did you get to that page?
[13:36] <wgrant> noodles775: Quite a few people seem to get to the help wiki without ever seeing the PPA overview page. Maybe some places just link to the ppa.launchpad.net address.
[13:37] <mat_t> james_w: yeah, that's exactly what happened
[13:37] <mat_t> james_w: thanks :)
[13:37] <noodles775> wgrant: Ah, I see.
[13:37] <z3l> wgrant : from here -> https://help.launchpad.net/FrontPage
[13:37] <wgrant> z3l: How did you know that you wanted to install a PPA? Did you have a particular one in mind?
[13:38] <doctormo> Is there any way to use launchpadlib without having it oauth every fickin time? It's making debuging a nighmare and slow
[13:39] <wgrant> doctormo: You can save and load credentials.
[13:39] <wgrant> doctormo: There's an example at https://help.launchpad.net/API/launchpadlib, IIRC.
[13:39] <wgrant> doctormo: And consumers like ubuntu-dev-tools have helpers.
[13:39] <doctormo> wgrant: thank you, I just don't understand why people are still using oauth.
[13:40] <z3l> wgrant : i would like to install ktoon 0.8.4.1 -> https://launchpad.net/ubuntu/+source/ktoon
[13:40] <wgrant> doctormo: ... why shouldn't they?
[13:40] <doctormo> wgrant: firefox isn't an operating system, reinventing authorisation and then trying to man handle a website cross auth into a client desktop auth system. It's just not right.
[13:41] <wgrant> doctormo: That's just the current implementation of the authorization part of Launchpad's OAuth support.
[13:41] <z3l> I must be blind, i don't see 'Technical information' link 'Technical information' link
[13:41] <wgrant> z3l: That's not a PPA.
[13:42] <wgrant> There is no ktoon in any PPA.
[13:42] <wgrant> And there is no ktoon 0.8.4.1 package known to Launchpad.
[13:42] <doctormo> wgrant: Aye, hopefully there'll be a future token based system which intergrates the authorisation into the whole os and not just for launchpad.
[13:42] <doctormo> Or just for firefox
[13:43]  * doctormo dreams about web browsers not having cookies any more
[13:43] <wgrant> doctormo: What is your exact complaint with the way things currently work?
[13:43] <wgrant> How would it be done better? At some point the provider (LP) has to interact with the user almost directly, or everything collapses.
[13:44] <doctormo> wgrant: none, except for a web browser requirement, none at all. You've done what needed to be done with OAuth. Ignore.
[13:44] <z3l> what is this link : https://launchpad.net/ubuntu/+source/ktoon, just a hosted source page ?
[13:44] <wgrant> doctormo: Well, LP is a webapp...
[13:44] <noodles775> z3l: so you don't need to do anything complicated to install ktoon 0.8.1-4... Just go to Applications->Add/Remove...
[13:44] <wgrant> z3l: That's an Ubuntu source package.
[13:44] <noodles775> and search for it there.
[13:45] <wgrant> z3l: That shows the versions of ktoon available by default in each Ubuntu release.
[13:45] <doctormo> wgrant: And say if I wanted to talk to auth on a text only desktop. Just for sake of asking.
[13:46] <wgrant> doctormo: Then I use links to go to the authorization URL.
[13:46] <doctormo> wgrant: If it were me, I'd take the username and password, have those boxes in my os wide service add, do the hoke with the tokens and forget the password once done. No need for web browsers, only for http requests.
[13:46] <wgrant> doctormo: Perhaps a non-browser official client could be provided, but consumers can't just ask for email address and password.
[13:47] <doctormo> wgrant: Web browser is a consumer, something has to act as client, I'd rather it wasn't an easily fooled web browser.
[13:47] <wgrant> doctormo: The browser is not the consumer.
[13:48] <doctormo> wgrant: Only the operating system, I understand what your getting at, but I don't quite agree.
[13:48] <wgrant> doctormo: Huh?
[13:48] <wgrant> I never said anything about the operating system.
[13:49] <z3l> wgrant & noodles775 : thank you for answer, i understand better now :)
[13:50] <doctormo> wgrant: typical OAuth treats the web browser as soemthing special, I have no idea why, it's not like it's very secure. It's just a replacement for propper os based auth sessions, hence mistakenly treating it like it's the os by calling it like a core service for service auths.
[13:51] <wgrant> doctormo: Because OAuth is normally used as an alternative access method to webapps, so the browser already has a secure copy of the master credentials.
[13:52] <doctormo> wgrant: It only has that because it's got it's own set of credentials that the os isn't privy too. If it was backed onto the same system, it wouldn't matter.
[13:53] <wgrant> doctormo: Ah, so you are asking for a fundamental shift of all Web authentication?
[13:54] <Daviey> doctormo: err, if you try and use launchpadlib on a box without firefox it falls back to w3m.. So ANY browser.. it's not fair to mention firefox specifically.
[13:54] <doctormo> Daviey: Your right, I'm being unfair to firefox. ANY web browser.
[13:57] <doctormo> wgrant: Well yes and no, I don't think the industry is going to shift. But it doesn't mean that devels should follow the line completly. Having a second lp auth that allows a http, non-brwoser client would be good, use something simple to specify the required fields and don't use html.
[14:16] <popey> gmb: HELP!
[14:22] <gmb_> popey: Did you actually want my help or did you just shout "HELP" to crash my laptop? 'Cos that's what happened.
[14:22] <popey> heh
[14:23] <popey> No, just testing my crashscript
[14:24] <gmb> popey: Thanks for that, good to know.
[14:24] <popey> yw
[14:44] <flacoste> BjornT: did bzr-git worked to push to github?
[14:45] <BjornT> flacoste: no. bzr-git doesn't support pushing yet (at least not according to the docs, i didn't actually try)
[14:45] <flacoste> BjornT: ok, so you used git itself to create a branch and apply the patch, i guess
[14:46] <BjornT> flacoste: yeah. wasn't much problem, since it was only one commit
[15:00] <doctormo> Are there ways to list all the available bzr branches (code) for a given project using launchpadlib?
[15:06] <james_w> doctormo: https://edge.launchpad.net/+apidoc/
[15:06] <james_w> search for getBranches
[15:07] <james_w> you can call it on a project, team or person
[15:41] <doctormo> thanks james_w
[16:03] <lantash1> QUESTION: Can somebody tell me why the automatic synchronization and one-time import don't work here: https://translations.edge.launchpad.net/lottanzb/0.5 I requested an import several times during the past days and nothing happened
[16:04] <lantash1> ah, it just looks like they haven't been reviewed. Can somebody with the required privileges do that for me? https://translations.edge.launchpad.net/lottanzb/0.5/+imports
[17:26] <dpm> lantash1: it seems that you didn't import a POT template together with the PO files. I believe that's why they are stuck in Need Review. LP cannot figure out where to import those translations without a template
[20:24] <tomfriedhof_> I can't access this url http://bazaar.launchpad.net/~davidstrauss/pressflow/6/
[21:05] <pwolanin> tomfriedhof_: hmm, I was just there a little while ago
[21:05] <pwolanin> I can access it now
[21:11] <tomfriedhof_> thanks pwolanin, I just grabbed the gz file and ran a diff
[21:25] <niemeyer> Hey guys
[21:26] <niemeyer> I'm wondering if there's some issue relative to the ordering of releases, or if I'm missing some step
[21:26] <niemeyer> https://edge.launchpad.net/image-store-proxy
[21:26] <niemeyer> The latest release is 0.9.1, as shown in the "series and milestones" graph
[21:26] <niemeyer> But the Downloads link in the right hand side actually shows 0.9
[21:26] <niemeyer> I'm not sure if I did something stupid or if Launchpad is a bit lost there
[21:53] <pwolanin> I'm trying to push a copy of an existing branch to a new branch, but launchpad is giving me a bzr error
[21:53] <pwolanin> Using default stacking branch /~vcs-imports/drupal/main at lp-45211920:///~acquia/drupal
[21:53] <pwolanin> bzr: ERROR: Server sent an unexpected error: ('error', "KnitPackRepository('lp-45211920:///~vcs-imports/drupal/main/.bzr/repository')\nis not compatible with\nKnitPackRepository('lp-45211920:///~acquia/drupal/1.x-6.x/.bzr/repository')\ndifferent rich-root support")
[21:54] <pwolanin> I guess "different rich-root support" is the key item here?
[21:56] <james_w> pwolanin: yes, that's the key
[21:57] <pwolanin> james_w: ok, maybe I got it to work by deleting the empty branch I'd created on lp and then pushing
[21:57] <james_w> that's possible
[21:57] <james_w> at least it is working :-)
[21:59] <pwolanin> james_w: can you explain what a stacked branch is?
[22:00] <james_w> sorry, got to dash
[22:00] <james_w> someone else will be able to
[23:05] <quentusrex> Why does an orig get counted multiple times for the size limit in a ppa?
[23:06] <quentusrex> each version increment uses the same orig file, but it's counted for each package version... I have a 160MB orig file with sound files and images, but for each package version it fills up the space quickly...
[23:06] <quentusrex> even though it should only be stored once.
[23:08] <wgrant> quentusrex: File a bug at https://bugs.launchpad.net/soyuz/+filebug. It's a pretty easy fix.
[23:08] <quentusrex> ok, thanks.
[23:08] <wgrant> Oh.
[23:08] <wgrant> Actually, it's not so easy.
[23:08] <wgrant> But file it anyway.
[23:09] <quentusrex> I'm filing
[23:10] <quentusrex> https://bugs.launchpad.net/soyuz/+bug/432152
[23:12] <quentusrex> Also, for building for multiple releases, do I need to rebuild the whole package with the ~release added to the end?
[23:30] <quentusrex> I've already deleted the orig file from a previous attempt, but it says: File freeswitch_1.0.4.orig.tar.gz already exists in PPA for PBXbuntu Drivers, but uploaded version has different contents.
[23:31] <quentusrex> it was a corrupted orig file...
[23:31] <quentusrex> the original was
[23:32] <wgrant> quentusrex: You can't ever upload one with the same name.
[23:32] <wgrant> quentusrex: How was it corrupted? Why was it changed at all from upstream>?
[23:32] <quentusrex> there is no upstream package yet.
[23:32] <quentusrex> and the files inside it were corrupted.
[23:32] <quentusrex> I'm building the ubuntu package first,
[23:33] <quentusrex> but someone else is working to get the debian package later
[23:33] <wgrant> Well, you'll have to change the version of the orig.tar.gz.
[23:33] <wgrant> Either by making it 1.0.5, 1.0.4+repack, or something else.
[23:33] <quentusrex> ok
[23:38] <quentusrex> that doesn't work... :(
[23:39] <wgrant> quentusrex: What goes wrong?
[23:40] <quentusrex> it doesn't detect the repack orig properly...
[23:40] <quentusrex> during the build process
[23:40] <wgrant> quentusrex: Did you specify the right version in the changelog?
[23:41] <quentusrex> hmm
[23:41] <quentusrex> that could be why....
[23:42] <wgrant> Um.
[23:42] <wgrant> Your last couple of uploads were native.
[23:42] <wgrant> Oh, no, that was a couple of days ago.
[23:42] <quentusrex> yeah...
[23:43] <quentusrex> what does native mean?
[23:43] <wgrant> It means that the diff.gz and orig.tar.gz are merged into one tar.gz, which is different every time.
[23:45] <quentusrex> ok
[23:45] <quentusrex> makes sense
[23:45] <quentusrex> awesome, that worked.
[23:47] <wgrant> Great.
[23:55] <quentusrex> is there a page for naming and what you can do with the package naming?