[00:03] There's a spelling error on help.lp.net [00:03] MTecknology: or take the tour from https://launchpad.net [00:03] MTecknology: which is? [00:03] suite of tools that help ;; s/help/helps/ [00:04] no, the do help [00:04] they don't helps [00:04] oh [00:04] I was thinking of suite, not tools [00:05] :) [00:26] Mez: not exactly an API, but: wget -O- -q http://ppa.launchpad.net/maxb/ppa/ubuntu/dists/jaunty/Release.gpg | gpg --list-packets | sed -nr 's/.* keyid //p' [00:29] then wget -O- -q "http://keyserver.ubuntu.com:11371/pks/lookup?search=0x$keyid&op=get" [00:57] MTecknology: (re paper) you could probably also google around for any reports of mark's talks on launchpad. they tends to have his thoughts on what/where/why etc. [02:29] How do I drop a blueprint? [02:31] MTecknology, You mean, indicate that you don't care anymore? [02:33] persia: no, I want them gone [02:33] we got two somehow that should be bug reports and are entirely irrelivant to us [02:35] I don't think you can do that. I think "Obsolete" and "Superseded" are the closed to "gone" tha blueprints get. [02:36] persia: https://blueprints.launchpad.net/loco-drupal/+spec/blue-screen-of-life [02:36] I'll be right back, we're going to run to the library quick [02:36] https://blueprints.launchpad.net/loco-drupal/+spec/ctrl-alt-backspace-timer [02:44] persia: is there anything you can do about those? [02:45] MTecknology, Me? No. I'm just a user. [02:45] err - a rubber ducky can do * [02:45] No idea. [02:48] Ursinha: hi [02:49] Ursinha: got a minute? [02:49] MTecknology, I'll pass the message tomorrow, because mrevell is sleeping now :) [02:49] in person, I mean, because already forwarded to him [02:49] your mail, that is [02:49] MTecknology, what can I do for you? :) [02:49] https://answers.edge.launchpad.net/launchpad/+question/61282 [02:49] and thanks [02:50] Ursinha: I imagine you guys don't have much downtime, so I can see where he might prefer to wait a week to do it [02:51] MTecknology, I'm pretty sure he'll return to you tomorrow with an answer :) [02:51] * Ursinha pokes spm and hides [02:52] Ursinha: howdy, how may I assist today? :-) [02:52] spm, do you have little time to delete two blueprints? [02:53] spm, https://answers.edge.launchpad.net/launchpad/+question/61282 [02:53] Ursinha: not atm, but will reassign to losa's so one of us will get it in the next day or so [02:53] spm, thanks :) [02:53] spm: Ursinha: thanks :) [02:53] np [02:54] MTecknology, usually you just file a request on answers and the person on CHR day will take care the day next [02:55] Ursinha: sorry, I've been on a rampage lately and bugging people in here a lot. I'm slowing down so I'll settle back to waiting for people to get to it :) [02:56] MTecknology, just telling you how it works, so you can be sure it'll be accomplished anyway === stub1 is now known as stub === error404notfoun2 is now known as error404notfound [04:44] spm, is it just me or are downloads from the librarian really slow today? [04:44] like eg http://edge.launchpad.net/bzr/1.12/1.12rc1/+download/bzr-1.12rc1-2.win32-py2.5.exe [04:45] poolie: 4,443,533 683K/s in 9.2s [04:45] flat out as far as my adsl goes. ? [04:45] iow "just me" [04:45] :/ [04:46] poolie: is it possible your PC has been infected by a virus/trojan and turned you into a zombie? :-P [04:47] quite possible [04:47] poolie: you're a 'node customer: try downloading a test file from their mirror: wget http://mirror.internode.on.net/pub/test/10meg.test [04:48] that should eliminate any pacific/atlantic crossing issues. and focus on the syd<->adl link(s) [04:48] yes i was just trying an iso from them [04:48] about 250kB/s [04:48] :-( [04:48] so kinda slow but not like 27kB [04:48] are you adsl1 or 2+? [04:48] 2+ [04:49] hmmm [04:49] I know telstra was killing those on RIM's to a max of 3mb/s - but you can't get 2+ via a RIM. so that ain't it.... [04:50] poolie: mtr? is that showing slower than usual links for you? [04:51] it looks reasonable [04:52] about 16ms to their router, 170ms to San Jose, and 350ms to London [04:52] that's what's strange [04:52] very [04:53] should probably reset my router :/ [04:53] I hate to suggest that. but it might help.... [04:53] just sounds too much alt-clt-del'ish. :-( [04:54] failing that, try snarfing a download via tcpdump/wireshark. have a squiz in tcptrace/xplot and see if you're getting full tcp windows or something? [05:16] poolie: any joy? [05:16] no [05:16] i'm going to restart my pc [05:16] hmmmm [05:16] about 50% because it might help and 50% because their support desk will ask me to [05:18] actually - you haven't go tthat magic tcp slowdown stuff enabled have you? that you use for bzr testing? Going for the long shot here... [05:18] hee hee [05:18] pretty sure not [05:18] cool :-) thought I'd ask the obvious, just in case :-D [05:19] why yes, the gun _was_ pointing at my foot [05:19] actually no. [05:29] spm, no good :/ [05:29] also, not your problem :) [05:30] heh [05:31] poolie: may be relevant: "16:28:41] <-> great, optus cable break again" - near coff's apparently [05:32] i did check their status site and i don't see anything very relevant [05:32] maybe it's Stephen Conroy's fault [05:32] everything else digital is === MTecknology is now known as mteck-mobile [07:39] maxb: might be able to play with that [07:55] I have a basic bazaar question. Can I ask it here? [08:03] keithcu: sure, or you can ask in #bzr === MTeck is now known as MTecknology [09:26] He [09:26] *Hey. I'm failing to check out a branch from launchpad [09:26] Is it down or something? [09:27] i've had problems for several days.. but thought it was just the project that was failing (squid3) [09:27] trying to run bzr branch lp:do but nothing happens [09:28] very annoying. [09:33] kallepersson, there's ni problem with codehosting [09:33] what's the error? [09:35] beuno: there is no error. run bzr branch lp:do just never finishes. [09:35] and it has worked before [09:35] beuno: does it work for you? [09:35] kallepersson, try: bzr branch lp:do -Dhpss [09:35] and look in: ~/.bzr.log [09:35] it will tell you exactly what it's doing [09:35] okay, thx. [09:36] kallepersson, pastebin the relevant bits of the file, and I can try and help [09:37] what version of bzr are you running? [09:37] 1.10 [09:38] kallepersson, do you have a shared repository? [09:38] huh? I'm only trying to check out a branch [09:38] no, I don't [09:38] ok, good, less things to think about :) [09:39] what is .bzr.log saying? [09:39] at the moment it's doing a lot of [09:39] 102.667 RemoteSSHTransport.readv 3 offsets => 1 coalesced => 1 requests (1) [09:39] 102.668 hpss call w/readv: 'readv', '/~do-core/do/trunk/.bzr/repository/packs/0eab7911ca7b40aef205f456ea068c04.pack' [09:39] and by lots I mean that's what it has been doing since i ran the command [09:39] so it's downloading [09:39] but it's never been this slow [09:40] it branched in 32 seconds here [09:40] 46 seconds actually :) [09:40] So it works for you. Good [09:40] still not finished for me, 5+ minutes now [09:41] yeah, maybe your ISP is just misbehaving? [09:41] perhaps [09:41] but launchpad.net is working [09:41] ah, now it shows some status at least [09:41] Copying Revision texts 1/5 [09:42] Extremely slow, but as long as it moves I'm fine [09:42] sounds like you're lacking a local repo [09:42] Thanks for your help [09:42] a local repo of the thing I'm checking out? yeah, I am [09:42] Would that speed things up? [09:42] oh *yes* :) [09:42] Good good [09:42] bzr --init-repo [09:43] well, see the help first, there's options [09:44] erm bzr init-repo --1.6 should be enough [09:44] then it will use stacking, which makes it very quick [09:45] thx [09:45] np, let us know how you get on [09:45] still awfully slow but I guess it will be done sooner or later :) [09:47] kallepersson, do a "bzr info -v" for me in the dir you're branching in [09:47] (as in, in the dir you're runing bzr branch...) [09:48] Location: [09:48] shared repository: . [09:48] Format: [09:48] control: Meta directory format 1 [09:48] repository: Packs 5 (adds stacking support, requires bzr 1.6) [09:48] Create working tree for new branches inside the repository. [09:48] Repository: [09:48] 0 revisions [09:48] Ow, sorry for pasting so many lines. === kiko is now known as kiko-afk [09:56] kallepersson, si you are using a shared repo :) [09:56] s/si/so [09:56] yes, I added that since bigjools suggesed that [09:56] ah [09:57] anyway, gonna try later [09:57] thx for help [09:58] :) [10:06] guys I was wondering how I get a SSH key pair on my Mac? [10:06] are there instructions on how to do this for Mac OSX? [10:09] CBro2007, same instructions as for Linux I think [10:09] verterok may know [10:09] ok === _neversfelde is now known as neversfelde [10:48] CBro2007, in the terminal, it's the exact same method as Linux. [11:58] Can some of the devs please enlighten me as to the motivation for this annoying change to check for the file name of a changes file in an upload to a PPA? -> bug 315643 [11:58] Launchpad bug 315643 in soyuz "PPA upload fails silently if .changes filename is misformed" [Undecided,Confirmed] https://launchpad.net/bugs/315643 [11:58] and question 51583 [12:05] Laibsch: well, the motivation is correctness, isn't it? [12:05] Laibsch: is there any use case for uploading filenames in other formats ? [12:09] Laibsch: 'reusing bash history commands' ? I don't think it's legitimate. [12:15] Why not? [12:15] Very legitimate use case [12:16] And what is incorrect in tmp.changes? [12:16] It's about the content that I sign [12:16] not the filename [12:17] Where exactly is that correctness defined? RFC? Debian policy? [12:17] There is no such thing [12:17] quite simply put [12:19] Laibsch: I'm sorry, but your are wrong. Who would benefit of downloading 10 tmp.changes files ? [12:20] Who downloads the .changes files? [12:20] I download dsc files [12:20] Or .deb files [12:20] orig.tar.gz files [12:21] diff.gz files [12:21] .changes? [12:21] never [12:22] you say "you are wrong", "bash history is not a legitimate use case", I still wait for *real* arguments from your side [12:22] * Laibsch wouldn't even know where the .changes files are stored [12:22] * Laibsch goes to look [12:23] found it, yet never used it [12:23] Laibsch: that's you, you are not the only LP user. [12:24] Well, I never implied this [12:24] Laibsch: note that you have to convince us that meaningful changesfile names are *not* beneficial [12:24] You demand a use case from me [12:24] I demand a use case from you [12:25] Laibsch: right, understood [12:25] That is what arguing (and finding the best solution) is about [12:25] wait a minute, you say you won't change unless I can prove that meaningful .changes filenames are detrimental? [12:26] Interesting trick [12:26] Laibsch: ehe, it's not tricky, it's correct [12:26] BTW, who corrected the filenames for the situations where I used tmp.changes? [12:26] All the filenames in my PPA seem to be correct [12:27] Laibsch: you only change if what you have is broken or if what is purposed is better. [12:27] Yet, I am sure I uploaded quite a few tmp.changes [12:27] well, you made the change from accepting tmp.changes to silently rejecting (which is still unfixed) [12:27] Laibsch: maybe you don't, but I can look for the upload logs [12:28] I am demanding why that is necessary and you turn around and say that I need to convince you that you won't change unless I can prove that meaningful .changes filenames are detrimental. that is a trick in arguing [12:28] demanding to know [12:28] I'm talking about the first change [12:28] You switch to talk about the second change [12:29] -> rhetorical trick [12:29] cprov: IYO, how should the "upload from other distributions" work in a meaningful way? [12:29] Laibsch: because that's the only thing that matters ;) how the system is and what you are suggesting to change. [12:29] * Laibsch can only think this is to save the uploader some work [12:30] cprov: I'm still waiting for your ex-post reason to change and why that was so necessary to break a valid use case [12:30] "correctness" is not defined anywhere [12:30] Laibsch: okay, and I said that's not a sufficient argument for forcing users to download meaningless filenames from the UI. [12:31] And I have not yet heard anything else [12:31] cprov: Are changes downloadable? [12:31] see, maxb did not know, either [12:31] I thought they specifically needed to not be downloadable [12:31] gets better all the time [12:31] https://launchpad.net/%7Er0lf/+archive/ppa/+files/anki_0.9.7.8-1rolf1_source.changes [12:31] just an example [12:32] because if you can obtain someone's signed .changes, you can possibly forge an upload to somewhere else [12:32] maxb: signatures are striped in upload time [12:32] Laibsch: Whilst I sympathize to some extent with your point of view, I think you're being excessively confrontational to the point of damaging your own argument [12:32] Understood [12:33] I'm trying to really argue [12:33] not be confrontational [12:33] But I'm really mad about this change [12:33] and IRC does not convey so many things [12:33] Just text [12:33] And I dont feel I can type fast enough ;-) [12:33] seems to me like you are ;) [12:34] Laibsch: but we can always 'try harder', right ? :) [12:34] Like everything technical, one is never fast enough ;-) [12:34] Laibsch: what command do you have in your bash history to repeat? [12:35] alright, if you give me some way to extract the information needed without changing the command line every time, I'm happy (you should still fix the silent ignore, soon, cost me quite a few times of "wth?") [12:35] * Laibsch looks at dpkg-parsechangelog for help [12:35] So, I think its fair to say that the root problem is that all the standard tools adhere to a strict format, but it's not written down in policy, so it's left open for interpretation whether it *should* be strictly checked? [12:36] maxb: on your point about about hiding changesfiles. It would make debug of extra tags a nightmare. [12:36] extra tags? [12:36] james_w: I'm trying to document best practice of "upload from other distro" in https://answers.launchpad.net/soyuz/+question/51583 [12:36] maxb: Launchpad-Closes, for instance [12:36] ah [12:36] Laibsch: do you often dput from outside the unpacked source package you are uploading? [12:37] maxb: I'm not too keen about it, just saying it would be hard if the files were not visible to users. [12:37] Laibsch: read "man dput" [12:37] No, I agree, making signature-stripped changes visible is the best option [12:37] maxb: what tool outside of LP is checking for .changes filenames? === kiko-afk is now known as kiko [12:38] Laibsch: oh, no, it's not in the man page, sorry [12:38] Laibsch: What tool outside your shell fragment is generating non-standard ,changes filenames? [12:38] :-) [12:38] maxb: tmp.changes [12:38] Laibsch: they don't check, but they do generate standard filenames. [12:39] cprov: Tell me the command to generate a .changes file with appropriate filename and we're all done [12:39] debuild [12:39] I mean , that is my problem [12:39] Laibsch: debuild [12:39] doesn't help [12:39] I think the real question that needs solving is - how did the person who wrote the feature described here: https://help.launchpad.net/Packaging/PPA#Using%20packages%20from%20other%20distributions anticipate you getting the .changes? [12:39] only 99% [12:40] Laibsch: I always use debuild to do exactly this [12:40] james_w: I did, too [12:40] until this one day when I filed an invalid bug [12:40] Laibsch: Didn't I write you a mkchanges script last time we were discussing this? [12:40] I'll get you the number in a minute [12:40] suspense [12:40] maxb: Not sure, I raised the point numerous times, it was never solved [12:40] IIRC [12:41] Laibsch: also, changestool from reprepro might interest you [12:41] bug 276391 [12:41] http://paste.ubuntu.com/119192/ [12:41] Launchpad bug 276391 in saxonb "FTBFS: "You must specify a valid JAVA_HOME or JAVACMD!"" [Undecided,Invalid] https://launchpad.net/bugs/276391 [12:41] that is the case where debuild will not work [12:42] cprov: Do you happen to know how the "forced suite upload" feature of PPAs is supposed to be used? [12:43] maxb: over pristine changesfiles from debian, for instance [12:43] Yes... but how do you obtain those? [12:43] maxb: requires re-signing [12:44] I don't think Debian publishes signature-stripped changesfiles like LP does [12:44] maxb: mostly target for people who uploaded stuff for debian [12:44] right, so it's really for people uploading directly to debian, only [12:44] maxb: so they don't necessarily have to create a new debian/changelog entry for upload the same debian source to their PPA [12:45] james_w: I'm not sure what you are getting at wrt changestool [12:45] Laibsch: given a source package it can generate you a changes file I believe [12:45] without unpacking it at all perhaps [12:45] * Laibsch studies manpage diligently [12:45] changestools creates a very skeletal changes file, and won't pick the right filename for it automatically [12:47] james_w: yes, could be. [12:47] oh, seems like maxb tested, but came out negative [12:47] I was going along the line of "manipulte existing .changes" file [12:48] but that brings us back to the problem you already raised of "where to get Debian changes files" [12:48] http://paste.ubuntu.com/119192/ [12:48] script. prewritten. works. [12:48] :-) [12:49] Further, while this will mostly be about recompiling Debian packages, there will be other cases [12:49] * Laibsch gone multitasking, overloaded [12:49] I remember that script [12:49] rememberred, even === mrevell is now known as mrevell-lunch === abentley changed the topic of #launchpad to: https://launchpad.net/ | Channel logs: http://irclogs.ubuntu.com | Help contact: abentley | launchpad-users@lists.launchpad.net === mrevell-lunch is now known as mrevell === matsubara is now known as matsubara-afk [15:24] hi there. I'm the owner of a branch of a project and so I registered a team for the separated PPA and code branch. Now I'd like to have a mailinglist for my users, so that a.) non-team members can send mail to it (moderated), b.) anyone can subscribe to the list. Is that currently possible with LPs mailinglists? [15:25] afflux: yes, if with "anyone can subscribe" you mean "any launchpad user can subscribe" [15:26] andrea-bs: do the subscribers have to be in the team then? [15:26] afflux: currently yes [15:27] andrea-bs: then that's not what I want. It would mean that my users would have r/w access on the PPA and the code branch. [15:27] afflux: you can create a separate team for the users [15:27] or should I rather create a -dev and a -users team for a one-developer and at most 20 users? [15:29] afflux: at the moment you can only create a -users team or wait for bug #194126 to be fixed [15:29] Launchpad bug 194126 in launchpad-registry "Let people lurk on mailing lists" [Low,Triaged] https://launchpad.net/bugs/194126 [15:30] if that's okay with the LP guys, I'll do that ;) [15:30] thanks. === glade88_ is now known as glade88 === salgado is now known as salgado-lunch === abentley1 is now known as abentley === kiko is now known as kiko-fud === matsubara-afk is now known as matsubara === bac is now known as bac_lunch [18:19] i want to branch a project, take the source and continue working in a completely new direction... is it that what launchpads branch tool is created to do? or is it better to just create a new project and commit every source file again as an initial commit? === bac_lunch is now known as bac [19:03] joey: ping [19:03] bac: you bac from lunch? [19:05] Where are al the awesome people in the world? I feel like I've been put on a national mental ignore :( [19:08] mizipzor_: did you get an answer? [19:08] hi MTecknology [19:08] bac: how's it going? [19:08] good,thanks [19:08] bac: Do you know if joey or mrevell are around at all? [19:09] I'm here...sorry on phone concentrating. [19:09] ok [19:09] joey: just pong me when you get a chance? [19:10] mrevell seems to have disappeared from the world entirely :P [19:10] MTecknology: it's pretty late in the UK about now so i suspect matthew is gone until tomorrow [19:10] mizipzor_: That's a good use of the LP tools, Just create a new branch under the same project. [19:10] MTecknology: was there something i can help you with? [19:11] mizipzor_: you can call the initial merge w/e you want, I'd call it just <- that though [19:11] bac: sorry, I just saw your nick change and wanted to say hi. It's joey and mrevell that I need to annoy [19:17] Why do I get "same version already has published binaries in the destination archive" when I try to copy and re-build a package from Intrepid to Jaunty? There is no package with the same version for Jaunty in my PPA. [19:17] MTecknology: ah, ok. i'm sure joey will be with you when he gets a spare moment. [19:18] All the busy you people take on makes me want to apply for a job when I get out of college [19:40] abentley, see postalchris Q ^^ [19:40] MTecknology, if you want to type in here or PM that's fine. I'm on back to back calls until 5:30pm EST [19:41] joey: I just never got that email [19:43] MTecknology, ok. I did speak with him via email earlier. Maybe if you send a follow-up to him he'll reply back. Otherwise I can ask him directly to email you. [19:43] joey: is he right around you? [19:43] MTecknology, no he's in the UK [19:43] MTecknology, I'm in Colorado :-D [19:44] ok, i can ping him [19:46] joey: what about those suggestions you had for me for project management? [19:46] MTecknology, you can ask him about it as well. I told him that I told you that I was talking with hIm. [19:46] ok [19:46] thanks [19:53] Is mrevell still alive? [19:58] MTecknology: He is on UTC, I believe. It's 7:58 pm for him. [19:59] I tried getting him when it was 15:00 his time [20:00] postalchris: Sorry, I don't know much about PPAs, but I'll try to get someone who does. [20:03] postalchris: I don't think anyone's available right now. Can you ask a Question on launchpad? [20:05] abentley, sure. [20:06] abentley: what's the right project/product to ask the Question in? [20:06] postalchris: Launchpad [20:06] Or Soyuz [20:10] https://answers.launchpad.net/launchpad/+question/50883 implies this is a known issue [20:13] MTecknology, he's here... [20:13] MTecknology, he just is in the UK. [20:15] postalchris: The archive contains more than one distroseries. [20:15] joey: I've been trying to contact him in channel, skype, and email - never any answer :P [20:15] The message doesn't say anything about Jaunty. [20:15] I suppose he'll get to me when he wants to [20:15] bbiab [20:16] wgrant: I don't think I understand [20:16] postalchris: Both source and binary versions must be unique within an archive. [20:16] Your PPA is an archive. [20:17] If you copy only the source from one series to another, it would have to generate another set of binaries with the same version number. That can't work. [20:17] Twiddling the version number and re-uploading (as suggested in 50883, above) seems to work... [20:17] wgrant: So launchpad can't copy-and-rebuild within the same PPA? [20:17] postalchris: It's nothing to do with Launchpad; it's how Debian archives work. [20:18] MTecknology, you can catch him on here early in your morning. [20:18] I generally upload to Hardy and then copy with binaries to Intrepid and Jaunty. [20:18] But some things need recompiling, in which case I'll upload a ~8.04, ~8.10, ~9.04, etc. [20:18] wgrant: It has to do with Launchpad insofar as Launchpad is giving me a perplexing error message after letting me try to do something it can't do. [20:19] postalchris: It has to let you try, because copying without binaries is a fine thing to do when copying between archives. [20:20] wgrant: I disagree. The selection for "rebuild" could be grayed-out when "This PPA" is selected. [20:20] Perhaps some JS could be used. [20:20] But what is confusing about the error message? [20:23] * wgrant disappears. [20:23] joey, about what time? [20:23] wgrant: your average PPA user is not necessarily a Debian sysadmin. None of the other 140 people in the room could answer the question before you showed up! [20:24] assmuing they read it [20:26] Not sure about what is being discussed here [20:26] Not being able to copy and recompile in the same PPA between series A and B? [20:27] In that case, I ran into that issue once and could not quite figure out why LP would not let me do it. [20:27] Since I was copying from hardy to intrepid IIRC, I figured I wouldn't see any lib linking issues and just copied with binaries [20:27] I was a bit uneasy about it, though [20:27] My 2ยข [20:28] postalchris: Was that what you were talking about? [20:37] Laibsch: That's exactly the issue. I've filed a bug against this behavior: https://bugs.launchpad.net/launchpad/+bug/330711 [20:37] Ubuntu bug 330711 in launchpad "Unhelpful error message on copy-and-rebuild: "same version already has published binaries in the destination archive"" [Undecided,New] [20:38] I agree this can be difficult to comprehend [20:38] although it is a minor issue, I think [20:40] Laibsch: Minor and really, easily fixable [20:43] Laibsch: It explicitly asks you whether you want to copy the binaries. [20:43] * wgrant is gone again [20:43] MTecknology, he's off for dinner around 9:30am MST [20:43] yes, wgrant? [20:43] I did not want to copy them [20:43] so I declined === kiko-fud is now known as kiko-afk [20:45] joey: what is MST? [20:46] MTecknology, mountain time [20:46] -7? [20:46] MTecknology, or 11:30 eastern if that fits better for you [20:46] MST is -7? [20:46] yup [20:47] I'm CST [20:47] -6 [21:09] Is there a way to push a branch for an existing project, without using that project's stacked branch? I would like to push a branch for Banshee, but since mine is created using bzr-svn it doesn't stack with the existing trunk. [21:10] jmillikin, if you use --overwrite [21:10] i won't stack [21:11] or it [21:12] Still gives me the error, and states "Using default stacking branch /~vcs-imports/banshee/main" [21:12] don't believe it [21:12] what error is that? [21:13] bzr: ERROR: RemoteRepository(...) is not compatible with KnitPackRepository(...) different rich-root support [21:14] hrm [21:14] * beuno calls in jml [21:14] hi [21:15] jml, hi. This gentlemen here has an interesting problem for you [21:15] Full error output at http://pastebin.ca/1340445 [21:15] yeah, I see. [21:15] Has Bazaar still not grown an option for "don't stack"? [21:15] If so, I can't find it. [21:15] I am using version 1.12, from the Ubuntu PPA [21:16] poolie: ^^ [21:16] jmillikin: I'm just trying to figure out a workaround [21:17] jmillikin: ok. let's try this. [21:18] jmillikin: bzr init lp:~jmillikin/banshee/trunk; bzr push !$ [21:20] The init seemed to work, but push printed the same error. [21:21] jmillikin: ok. [21:21] * jml pokes [21:21] Ah! If I push with --stacked-on to the same repository, it says "Ignoring request for a stacked branch as repository already exists at the destination location.", but works anyway [21:22] jmillikin: ooh, clever. [21:22] At least, it's preceded to copying content texts now [21:22] jmillikin: that sounds like it's working [21:22] jmillikin: if pain persists, ping me. [21:23] Will do, thanks. [21:24] jml, no there is no option at present [21:24] i didn't realize that launchpad's suggestion of stacking was so strong [21:24] really we need --no-stacked and a reconfigure option [21:25] poolie: no-stacked would definitely be a good start. [21:28] Alternatively, when checking for a stacked branch, perhaps (repo-format, first-revid) could be sent, for finding an appropriate source? I assume there's some sort of logic in Launchpad itself for determining whether a trunk repo is usable. [21:28] that's a good idea too [21:28] it should probably just not stack rather than giving an error [21:35] hello [21:35] why doesn't the translation feature allow me to license under same license as the project? [21:35] I even suspect you can't legally license translations for a GPL software as BSD [21:36] has anyone investigated this? [21:36] AnMaster: Yes we have ;) [21:36] AnMaster: BSD is fully GPL compatible. [21:36] henninge, so if a project declare it's translations to be GPL, how can you legally convert it to BSD [21:36] sure BSD -> GPL is ok [21:36] but GPL -> BSD is not [21:37] you're right there. [21:37] but that is not what is happening [21:37] henninge, what about already translated strings then? [21:37] when the project is added to launchpad [21:37] they keep their license [21:37] the BSD is only for translaitons done in launchpad. [21:37] henninge, and what if you want to keep same license as the project? [21:37] There is no way I'm going to translate with a different license [21:37] than that of the project [21:38] Well, that is where the BSD compatiblity comes in. [21:38] henninge, also since the translation is a derivative work of the English, GPL licensed text, making the translation under BSD would be a GPL violation [21:38] if you disagree, please explain your reasoning [21:38] :) [21:39] Translations done in launchpad as such are under BSD but you can bundle them with your project under a different license. [21:39] henninge, "henninge, also since the translation is a derivative work of the English, GPL licensed text, making the translation under BSD would be a GPL violation" <-- now what about that... [21:39] ASIK they are not deemed derivative of the english text [21:39] AFAIK [21:39] henninge, according to who? [21:40] I'd have to look that up / ask others ... ;) [21:40] henninge, at least if it is a longer translated text, like a lot of background story of an open source game (which is what I wanted to translate) it would be derivative. [21:40] If you translate a book, doesn't the original author still have copyright? [21:41] sure it may not apply to "File" or "Save" or such. But to longer texts like "One day, Tux was out walking when " [21:42] henninge, and I know at least two games that have such in their gettext files. One is on launchpad [21:42] * henninge looks something up [21:42] and I argue that it is a GPL violation [21:44] AnMaster: OK, I found this for the moment: https://help.launchpad.net/Translations/LicensingFAQ [21:44] * AnMaster looks [21:44] AnMaster: but I am not sure if that touches your situation. [21:45] It was written at the time when we introduced BSD licensing. [21:46] My understanding would be that "translations" are licensed under BSD, that is the pairing of an English and a translated text. [21:47] henninge, so you argue you could translate Harry Potter to for example Spanish or something, claim the translation is BSD licensed and then upload it on a website? [21:47] it seems the logical conclusion of what you just said === salgado is now known as salgado-afk [21:47] So if somebody used the translation of a text in a different program it would be because that program contains the same original English text - which would then still be under a different license (which may have been violated). [21:47] henninge, However I doubt the author would like it. [21:49] AnMaster: Admittedly, I am on a shaky foundation here and would like to get back to you when I have solid information. [21:49] henninge, that doesn't really answer the question I just asked. Because sure it depends on how it is used. But there is another issue, and that is that you host translations under a license that doesn't match the original. Which if it was allowed would allow re-licensing any book and or such [21:49] AnMaster: I see you point and I don't have the answer right now. [21:50] henninge, well I'm not likely to hang around all the time in this channel, so I guess you can contact me using the first email at https://launchpad.net/~anmaster [21:50] if that is public [21:50] no idea [21:50] AnMaster: I just checked that ... ;) [21:51] henninge, well I guess you are an admin or something and could see it anywya [21:51] anyway* [21:51] hm I seem to have lots of openids I never use these days [21:51] launchpad too [21:51] AnMaster: It is not public but other Launchpad users can see it. [21:51] * AnMaster never use openid [21:51] now it there was a way to connect these openids to each other.. [21:52] like a key ring or something :D [21:52] (yes I was sarcastic...) [21:52] ;) [22:21] AnMaster: It is not public but other Launchpad users can see it. <-- err [22:21] I have " Hide my email addresses from other Launchpad users" checked [22:22] AnMaster: I just checked and logged in with a test account I have - no address visible. [22:22] ah [22:23] yes as long as the staff can see it when they answer the question I had :) [22:23] AnMaster: I don't have general admin powers but as a LP developer I get to see a little more, I guess. [22:25] AnMaster: btw you should set your location to somewhere in Stockholm, or something similar general but sensible. [22:26] as long as it is unset anybody can set it - where ever they like ... ;) [22:26] henninge, err what? [22:26] why can anyone set it [22:26] I don't want to set my location at all [22:26] I want to not have that info on launchpad [22:26] put it on Antartica then, although that defeats the purpose [22:26] well I'm going to do that [22:27] It is a little gimmick that helps teams to see where everybody is spread out. [22:27] henninge, also Stockholm/Europe is the timezone for the entire Sweden [22:27] *shrug* [22:28] Hey, mine is "Berlin" although that is 300 km away from me ... [22:28] interesting, I was able to move the marker to way outside the map [22:28] check now [22:28] it is like double the map width outside the map [22:29] err, 1.5*map height rather [22:29] henninge, is that supposed to work? [22:29] I am not sure. [22:30] I have seen that it can be moved around in strange ways ... [22:30] google code ... :-/ [22:30] AnMaster: but you did something, the map is completely gone from your profile now. [22:31] henninge, I clicked "hide location info" [22:31] :) [22:31] and the timezone, too. [22:31] mhm [22:31] Ah, it's as easy as that ... [22:31] henninge, was on the map page [22:31] henninge, still I moved it outside the mpa [22:31] map* [22:31] could take a screenshot [22:34] henninge, http://omploader.org/vMTlncg [22:35] henninge, :) [22:35] AnMaster: That does look funny. :) [22:35] henninge, and buggy [22:35] AnMaster: I have seen the map moved that way but not the marker. [22:36] Well, it could mean "outer space" ... ;) [22:36] henninge, btw I used to commit to bzr using a email address on a domain that is no longer valid, is there any way I can get launchpad to credit me for these commits? [22:36] Really, it is missing a star map there ... [22:36] phh [22:36] don't know [22:37] henninge, I'm talking about maybe 500+ commits [22:37] at least [22:37] probably around 1000 [22:38] You could file an answers request (a question) with launchpad-registry about that. [22:38] hm ok [22:38] henninge, anyway I see you at least finally fixed the long standing bug of showing ubuntuwiki as some sort of contact info... Even for FreeBSD users like me [22:38] If those uploads can be linked to you in some way, I am sure it could be hacked somehow... [22:39] henninge, well they were on anmaster@envbot.org instead of a similar address. Also I'm the sole developer on those projects [22:39] Yeah, we have been trying to untangle Ubuntu and Launchpad. [22:39] yeah I couldn't afford the domain... Students sometimes can't [22:39] :/ [22:39] * AnMaster is a student yes [22:40] don't know what to say "I am sorry" or "Lucky you" ;) [22:41] henninge, hey the page https://launchpad.net/~anmaster/+editwikinames still use "Example: https://wiki.ubuntu.com/", I suggest replacing that with the Gentoo Wiki ;P [22:41] :P [22:42] why not, it would definitely untangle you from ubuntu, like making the ubuntu guys angry at you probably. [22:42] That would take it a bit too far. [22:43] I guess ... ;) [22:43] henninge, hm why do you use different navigation bars on different pages [22:43] like the top black one at [22:43] https://launchpad.net/~anmaster/+editsshkeys [22:43] and one on the side at: https://launchpad.net/~anmaster/+editircnicknames [22:43] it seems a bit inconsistent? [22:43] The UI is in a transient state from what I understand. [22:44] um [22:44] in a good design, the navigation bar would be a separate component [22:44] so you would replace it in one place... and it would update everywhere [22:44] or do you mean it is hard coded in each page or something!? [22:45] The side menus (coloured components) will be going away, the gray menu bars are the new style. [22:45] No, but page groups. [22:46] henninge, so why that just replacing a global class NotTopButTheOneBelowNavigationBar: or something (yes I guess you use python, no I haven't heard of PEP 8, and even if I had I wouldn't remember it since I mostly code in C) [22:46] But that is something I had little to do with so far. [22:47] I have only been working on Launchpad since September... [22:47] Yes, it is Python. [22:47] and Zope. [22:48] henninge, hm did it finally go open source yet or= [22:48] s/=/?/ [22:49] I guess not [22:49] "There are no branches for Launchpad itself in Launchpad." [22:49] damn, when I see bzr looping in 'GC_*' functions, I guess I should just kill it ;-) [22:50] domas, sounds like a python bug to me [22:50] or bazaar, or both [22:50] so I guess you should get a backtrace and report a bug [22:50] * domas kicks ddebs [22:50] domas, does bazaar use C modules? [22:50] if not: python, if yes: trickier [22:51] AnMaster: https://dev.launchpad.net/OpenSourcing [22:51] * AnMaster has actually worked more with the python C API than with python itself [22:51] =)) [22:51] henninge, "Soyuz and Codehosting", just wondering, what do they do? [22:52] Soyuz is the "buildds and archive management" bit [22:52] AnMaster: buildes, ppas, hosted branches [22:53] henninge, hosted branches as in those on launchpad? Well I guess without that launchpad is pretty much crippled [22:54] AnMaster: maybe it is just bazaar not behaving nicely on 500MB repo [22:54] domas, well I guess that would make sense. Makes me wonder why on earth anyone use bzr for anything that large. I guess it is images or such? [22:55] AnMaster: mysql [22:55] domas, I see. Hm... [22:56] * AnMaster uses PostgreSQL anyway [22:56] but I guess that too would have a huge repo [22:56] I always wanted something like that "shallow branch" thing that git has... in bzr [22:56] I run 'bzr log -v' [22:57] it goes for ~700MB of memory :) [22:57] ! [22:58] Is that 1.12, or earlier? [22:58] domas, bug [22:58] wait 1.12? Wasn't 1.9 released like um 1-2 months ago? [22:58] maxb: 1.12 on my side [22:58] bzr releases monthly [22:58] so, 3 months [22:58] ...and ads a new repo format bi-monthly [22:58] adds* [22:58] =) [22:59] well.... it's hard to begrudge them playing with new formats, when the basic format is compatible right back to 0.92 [22:59] I wish they could settle for one... I mean I still use pack-0.92 because last I looked it was default... [22:59] well [23:00] it confuses users [23:00] default doesn't work that well once you have people with 500MB trees that get cloned branched merged etc :) [23:00] and you have hundreds of trees based on same knits [23:00] domas, so which format is best? [23:00] newest! :) [23:00] dunno [23:00] I'd pick the one that would work now :)) [23:00] err, I need backward compatibility [23:01] to like 1.5 or so [23:01] why not at least mark most of them as of interest only to developers? [23:02] I'd love if bazaar would be multithreaded! [23:02] I mean, I remember back when bzr added the knit format [23:02] AnMaster: 1.6 provides a valuable new feature. 1.9 provides improved speed. [23:02] would run this thing on fat box :) [23:02] domas, I'd love that it doesn't have a lot of threading overhead on my old single-core CPU :) [23:02] which is the best I have [23:02] a Sempron 3300+ [23:03] abentley, what was that "valuable new feature"? [23:03] AnMaster: Stacking. [23:03] abentley, proper GPG signed revisions? [23:03] oh no... [23:03] `bzr help current-formats` [23:03] abentley, what do I use stacking for? [23:03] abentley: which format does 'bzr log -v' fastest on huge repos? :) [23:04] there are only three formats (and a rich-root variant of each) that are advertised as interesting [23:04] So in a way, most of them *are* marked as of interest only to developers [23:04] maxb, I always looked at bzr help init-repo [23:04] where there are a lot more listed [23:04] AnMaster: Avoiding having to push up the whole history every time to push a new branch to Launchpad. [23:04] maxb, that one says "See also: branch, checkout, init, repositories", but doesn't mention "current-formats" [23:05] abentley, does this only work for launchpad or also for other ones? [23:05] abentley, also "Avoiding having to push up the whole history"... what about bzr init-repo [23:05] doesn't it solve that? [23:05] AnMaster: locally, not on bzr-side [23:06] AnMaster: It works for anything that's got it set up. It works locally too. [23:06] AnMaster: the problem was that people would pull repo from ~userA and upload it to ~userB [23:06] init-repo is fine for a single user, but doesn't work well when the branches are owned by different people. [23:06] AnMaster: before 1.6 it was pain [23:06] ah, that is why the repos should be per-project [23:06] not per-user [23:06] :) [23:06] that wouldn't be distributed :) [23:07] domas, hm? it would just require a major redesign of launchpad [23:07] argh. 'bzr log' : 24556 root 20 0 550m 479m 3244 R 100 12.1 7:42.86 bzr [23:07] but it works quite ok for my setup [23:07] doko, remove it? [23:07] and symlink to /dev/null [23:07] works for me [23:08] AnMaster: That would mean giving users write access to each others repos. [23:08] abentley, hm... wouldn't bzraccess handle that? I mean you can have shared svn repos without clobbering [23:09] AnMaster: No, bzraccess does not handle that. [23:09] so I guess it is just a case of making it possible to add revision files, but not remove or change other users' files [23:09] like group writable directory, user writable files [23:09] (I fail to see why that wouldn't work) [23:09] AnMaster: If you're trying to economize by only having one repo, for a bunch of users, it necessarily follows that you have multiple users writing to the same repo. [23:09] abentley, correct [23:10] abentley, but is each revision stored in a separate file? [23:10] AnMaster: No. [23:10] ah [23:10] then it is indeed more complex [23:11] abentley, btw I can I push fron 1.9 repo to a 0.92 repo? [23:11] oh also what is rich-root good for? And subtree? [23:12] AnMaster: Yes, you can push from 1.9 to 0.92. AFAIK, you can push from 1.9 to knit, and possibly weave. [23:13] hehe [23:13] the oldest I used was metaweave :) [23:13] AnMaster: rich-root is required by bzr-svn. You should only use it if you're using bzr-svn. === matsubara is now known as matsubara-dinner [23:14] AnMaster: subtree enables the experimental Nested Trees feature. You should only use it for experimenting. [23:14] ok, so what *is* rich-root then? I mean, bzr-svn uses it, but what is it actually [23:14] svn:foo attributes or what? [23:14] hehehe, http://p.defau.lt/?p7tmQch8XRJa01AS_QTf1Q [23:15] it spends quite some time walking dictionaries [23:15] domas, that output is from? [23:15] AnMaster: It is a format in which the same data is stored about the tree root as any other subdirectory, and in which the default file-id is randomly generated. [23:15] AnMaster: oprofile [23:15] abentley, aha [23:15] 1.12-preview: [23:15] (native) A working-tree format that supports views and content [23:15] filtering. [23:15] sounds cool [23:15] but whyat is content filtering in this context? [23:16] or views. I guess this means I can do CREATE VIEW foo AS SELECT revid,logentry FROM revisions; or something similar [23:16] that is all "views" mean to me [23:16] AnMaster: Things like native line-endings and CVS tokens. [23:16] really docs need to get better [23:17] abentley, ah finally! [23:17] Hmm... it's not possible to comment on a bug at the same time as adding a tag, is it? [23:17] :D [23:17] abentley, now all I'm missing is cherrypicking like darcs :) [23:17] AnMaster: Docs for a preview feature that hasn't even landed? I think that is unlikely. [23:18] not like that will ever be added I guess === abentley changed the topic of #launchpad to: https://launchpad.net/ | Channel logs: http://irclogs.ubuntu.com | Help contact: - | launchpad-users@lists.launchpad.net [23:19] AnMaster: Like Darcs? Not likely. Darcs is a very different concept from other VCSes. Like Arch? Quite likely. Just needs someone to work on it. [23:19] jml: I am now trying to init another remote branch, stacked on my first branch, and receive the "ERROR ... is not compatible with ... different rich-root support" again. I cannot find any way to override the stacking branch in "bzr init" [23:19] abentley, haven't used arch so no idea [23:22] jmillikin: I wouldn't use init there. [23:22] jmillikin: instead 'bzr push --stacked-on FIRST-BRANCH' === jml changed the topic of #launchpad to: https://launchpad.net/ | Channel logs: http://irclogs.ubuntu.com | Help contact: jml | launchpad-users@lists.launchpad.net [23:23] jml: if I try that, I receive the error: Source format does not support stacking, using format: '1.6.1-rich-root' [23:23] I am using bzr init --1.9-rich-root, however. [23:24] jmillikin: can you please pastebin 'bzr info -v FIRST-BRANCH' and 'bzr info -v SECOND-BRANCH'? [23:25] http://pastebin.ca/1340578 [23:27] jmillikin: ok, can you please do 'bzr info lp:~jmillikin/banshee/trunk' and also pastebin the output of the push --stacked-on command I gave earlier? [23:30] http://pastebin.ca/1340583 [23:32] jmillikin: thanks [23:33] jmillikin: can you try that last command, but to a different URL, like lp:~jmillikin/banshee/test-stacking or something? [23:33] Sure [23:34] jmillikin: fwiw, you are doing two interesting things here. 1. is using Launchpad with non-default stacking, 2. is stacking bzr-svn bugs. [23:34] Would you like me to let it upload? [23:35] jmillikin: what does the output say? [23:35] source format does not support stacking, yada yada [23:35] Same as the pastebin. lp:~jmillikin/banshee/lp-test-stacking [23:35] fwiw, stacking support in bzr-svn is experimental [23:36] jelmer: that's good to know. [23:36] there is a warning but it was accidently disabled between 0.4.x and 0.4.17 [23:36] Once I branch from an svn repo, isn't the result simply a normal bzr branch? [23:36] jml, it *should* work and I'm not aware of any problems, but it's not well-tested [23:36] I assume this would apply to a branch created entirely in bzr with --1.9--rich-root also [23:37] jmillikin: branches from svn aren't quite normal bzr branches [23:37] jmillikin: jelmer can explain in much more detail than I though ;) [23:37] jml: finished uploading to https://code.launchpad.net/~jmillikin/banshee/lp-test-stacking . It seems to have created a --1.6-rich-root repository instead. [23:38] jmillikin: that should be ok [23:38] * jml takes a look [23:39] jmillikin: that looks fine to me. [23:39] * jml keeps poking [23:40] But it's not stacked, right? So trying to push it will require uploading all revisions. [23:41] jmillikin: well, stacked or not, once a branch is up there, only new revisions get pushed [23:41] jmillikin: but if it uploaded in that short a time, I'd stay it's stacked [23:41] I mean that if I were to push another bugfix branch, it would require pushing all revisions. [23:41] particularly since https://code.edge.launchpad.net/~jmillikin/banshee/lp-test-stacking says it's stacked too [23:42] I thought stacking was added in packs 6? [23:42] 5 [23:42] jmillikin: --1.6 flavour formats stack [23:43] Oh, very nice. Was it simply not adopted until post-1.6 then, even though the repository format could support it? [23:43] gosh we need to work on our non-default stacking story [23:44] jmillikin: we made further improvements, and the ability to autostack [23:46] jmillikin: the version marker on the repository format is just a sequence number [23:50] jmillikin: ok, so lp-test-stacking is *definitely* stacked [23:50] jml: Thanks for the confirmation :) [23:50] jmillikin: and if you push up any other branches in the same way, they'll be stacked too. [23:50] jmillikin: but you should use the http:// branch url, not the bzr+ssh one [23:51] Trying the http:// URL prints the error: ERROR: Server sent an unexpected error: ('error', 'Connection closed: please check connectivity and permissions (and try -Dhpss if further diagnosis is required)') [23:53] jmillikin: is this pushing to a new branch or to an existing one? [23:53] (you don't need any of the stacking options for pushing to an existing branch) [23:54] An existing one. Do I have to recreate the branch to change the stacking URL? [23:54] jmillikin: not sure, I've never tried to do it using bzr command-line options :) [23:54] jmillikin: I was just saying with future branches that you push up [23:54] Arr.. An upload to Ubuntu is hanging at 68154k/68155k, is there some magic way to fix this? :P [23:54] RainCT: excellent question! [23:55] hehe [23:57] So if you stack a branch via bzr+ssh, what happens when someone tries to branch from it using http? [23:59] maxb: that's a good question.