[00:14] directhex, you can't dput to OBS, right? [00:14] AFAIK. they have their own tool, called osc [00:15] which needs mono. hey, we've come full circle! ^_^ [00:29] directhex,its kinda weird [00:29] * NCommander is talking with the OBS Debian guy [00:29] NCommander, i know it's weird ^_^ [00:29] NCommander, i've spoken with him before. they had a critical bug with build-depends-indep previously [00:29] They sorta reinvented the wheel [00:29] They modified the RPM build system to run dpkg-buildpackage vs. rpmbuild [00:30] Crazy hack [00:30] directhex, what do you think of OBS overall however? [00:31] Hello. [00:31] Can someone recommend some automatic tools for creating .deb packages that will make it much easier to make them? [00:32] NCommander, building packages for more than just ubuntu: awesome (launchpad sucks in that it can't be used for debian collab) [00:32] NCommander, workflow: like ass. a big ass [00:32] directhex, well, maybe we can fix that [00:36] Tetracomm: such as launchpad PPA ? [00:36] savvas, i think he wants checkinstall [00:36] or whatever it's called [00:37] I want something to create decent .deb packages. [00:37] I hear that Checkinstall created packages are not accepted into the repositories. [00:37] for decent deb packages you need to know the debian and ubuntu packaging policies :) [00:37] https://wiki.ubuntu.com/PackagingGuide [00:38] * Hobbsee quietly chokes on the idea that checkinstalled packages would be acceptable for hte repos. [00:38] Hobbsee: Tell me more. [00:38] Tetracomm: it's only very good at simple packages, and guesses all the deps (sometimes getting it wrong) [00:38] only really works on the local system well, too [00:39] Tetracomm: https://launchpad.net/people/+me/+archive/ https://help.launchpad.net/Packaging/PPA [00:39] You can build your packages in a personal package archive, but you have to know how to make them first [00:39] Hobbsee: What if I tell it the dependencies when it asks? [00:40] the best way is to play around with deb source packages :) [00:40] Ok. [00:40] Tetracomm: it doesn't ask, iirc. [00:40] but yes, listen to savvas [00:41] Hobbsee: It asked me the last time. [00:41] Tetracomm: which program is it? [00:41] hmmm. maybe i've remembered wrong, then. [00:42] Hobbsee: We really should create a wiki page that explains why checkinstall should not be used to create packages for the repositories. This is a very common question, and it would be much easier to just refer people to a wiki page than to re-explain it each time [00:42] nhandler: that's true. [00:42] nhandler: i'd be surprised if there isn't one already though [00:42] savvas: It was Megatunix, but I will try to package Laoe now [00:42] . [00:42] Hobbsee: I personally haven't seen one before. I'll see what google turns up [00:43] Tetracomm: link please? [00:43] To Laoe? [00:43] http://www.oli4.ch/laoe/ ? [00:44] Yes. [00:45] Hobbsee: This (https://help.ubuntu.com/community/CheckInstall) was the only page I could find on {help,wiki}.ubuntu.com. It doesn't even mention why checkinstall is evil [00:45] oops] [00:45] nvm laoe [00:45] hey Hobbsee [00:45] hey NCommander! [00:45] !checkinstall [00:45] checkinstall is a wrapper to "make install", useful for installing programs you compiled. It will create a .deb package, which will be listed in the APT database and can be uninstalled like other packages. See https://help.ubuntu.com/community/CheckInstall - Read the warnings at the top and bottom of that web page, and DO NOT interrupt CheckInstall while it's running! [00:46] hmmm. that used to have more warning in it, i though [00:46] t [00:46] The package creation process is too complicated. :( [00:48] Tetracomm: That's why we use humans to make and maintain them. :) [00:48] Tetracomm: you just want to make this software available as a .deb package for others or to learn how to package? [00:49] To make it available to others. [00:50] Although it would be nice to learn how to package, it is too complicated. [00:52] and i'm going to package Zinf eventually [00:53] as well as Cmus [00:53] Tetracomm: file a bug at https://bugs.launchpad.net/getdeb.net and use "Create Package: LAoE" in the subject *or* (I'm not sure) file a bug for ubuntu https://bugs.launchpad.net/ubuntu/+filebug and ask for packaging (tag: needs-packaging) [00:54] I did that for Megatunix months ago and it is taking forever. [00:54] and I want to contribute [00:54] I don't see a bug for megatunix in getdeb.net [00:55] I don't think it was getdeb.net [00:55] Then try on getdeb.net :) [00:56] I want to contribute. [00:56] When they make the package, you can grab the source from http://archive.getdeb.net/getdeb/ and study it [00:56] Err. [00:56] ah well then, good luck :) [00:57] What is the main problem with checkinstall? [00:58] I think it's too simple and too my-machine-specific [00:59] I'll try out the new one and see how it works on the live cds. [00:59] Tetracomm: are you using 32-bit ubuntu? [01:00] Yes. [01:00] use your built package on a 64-bit ubuntu made with checkinstall [01:00] 8.10. [01:00] I mean, try to install it, I think it won't work [01:00] Ok. [01:01] A second reason would be that I believe checkinstall can't make source packages (correct me if I'm wrong) [01:02] Ok. [01:02] and source packages are used for easier updates and maintenance, also easier building for various architectures [01:03] Ok. [01:23] does LP have a "blocked by" mechanism on bugs? [01:24] no [01:24] only for blueprints [01:25] damn. I'll just add it in the bug text. [01:25] good god, a feature bugzilla has which is useful :o [01:39] what's the fix for when a package uses docbook and it attempts to contact the OASIS website to download the dtd? [01:39] You can use a local dtd, but I thought there was a better way of doing this than patching the DOCTYPE in each document [01:39] has anyone done it before? [01:40] use proxy to link it back to your machine? heh :) [01:41] hey Hobbsee [01:41] hey NCommander [01:41] don't ask me how though, just brainstorming :p [01:46] nellery: hi, are you around? [01:49] nixternal, NCommander for simple build environments, something like debomatic or falcon might be easier than using some of the available more complex infrastructures. [01:50] ooh I forgot about falcon [01:50] persia, debomatic doesn't have NEW queue [01:50] I dunno about falcon [01:51] persia, so OpenSuSE Build System actually pretty cool [01:51] * NCommander thinks in many ways it works better than PPAs, and in other ways, its epic suck [01:52] james_w: hi [01:53] hi nellery, how are you? [01:53] NCommander, as I said "for simple build environments" :) [01:54] james_w: I'm alright, you? [01:54] persia, reprepro is my favorite [01:54] nellery: good thanks [01:54] nellery: I'm looking at your new debdiff [01:55] nellery: thanks for working on it. I'm wondering why you added inetutils-inetd to the dependencies this time? [01:56] james_w: that package depends on update-inetd [01:56] did I get it wrong? [01:57] nellery: packages shouldn't depend directly on update-inetd, they should depend on inet-superserver and expect that that package provides update-inetd [01:57] as these packages already depend on inet-superserver I don't think you need to add anything [01:57] james_w: so that change should just be removed? [01:57] I believe so [01:58] my other question is why these packages depend on netkit? [01:59] james_w: do you mean netbase? [01:59] yeah, sorry [02:00] james_w: it doesn't appear that this was documented in the changelog [02:00] yep [02:01] james_w: is there any way to figure out where this change was made? [02:03] you can examine all the old packages [02:03] or try and work it out [02:04] it's not too important, but if we don't find out then we are kind of stuck with it [02:05] james_w: I think I've found the change [02:05] bug 123782 [02:05] Launchpad bug 123782 in heimdal "heimdal-kdc should depend on some inetd" [Low,Fix released] https://launchpad.net/bugs/123782 [02:06] which would be in 0.7.2.dfsg.1-10ubuntu2 [02:06] netbase was added before that apparently [02:07] or it might have been in an old version of the Debian package, and then not removed by Ubuntu when Debian removed it [02:09] james_w: it doesn't appear to be mentioned in the Debian changelog either [02:14] james_w: about bug 298496, should i remove the dependency on libxi-dev, or wait? [02:14] Launchpad bug 298496 in anyremote "Please merge anyremote 4.6-1 (universe) from Debian unstable (main)" [Wishlist,Triaged] https://launchpad.net/bugs/298496 [02:14] I'm not sure [02:14] I hit the same problem with another package yesterday, but I think it's a bug in XTest [02:15] I haven't yet checked for bug reports [02:15] A new libxi-dev has been uploaded that includes the required Xinput.h [02:16] (unless this is a different problem with libxi-dev) [02:16] I'm not sure that's the same bug [02:18] Hmm. Test date is the 18th, and https://lists.ubuntu.com/archives/jaunty-changes/2008-November/000827.html is the 19th. [02:19] It was reported in -devel recently that upstream was moving the file around. Mind you, it might also be the way XTest works. [02:22] nellery: I've got to sleep now, I'll review any debdiffs in the morning. Thanks for perservering. [02:22] james_w: alright, and thanks a lot for your help [02:22] wgrant: Did you file a "If I wanted to be contacted I would have published a way to contact me" bug on LP yet? === _Kopfgeldjaeger is now known as Kopfgeldjaeger [04:07] jono: There's a flip side to the hall of fame. People that get left out, can feel demotivated. [04:07] * Hobbsee smashinates akregator into the middle of next millenium. [04:08] Trouble? [04:09] * ScottK guesses he better start insiting that all uploads he sponsors are have all the proper paperwork filled out in LP so he can get 'credit' for sponsoring. [04:09] Much like '5 a day', I suspect the hall of fame will not generally motivate the behavior we actually want. [04:10] ScottK: it dropped my planet.u.c feed again [04:10] ScottK, Given the wide variety of experimentation with sponsoring, especially in light of widely publicised plans to demonstrate alternate defaults for a possible change, you'd do a lot better to support alternate models. [04:11] persia: I've had less than 8 hours sleep in the last two days. You'll need to be clearer. [04:12] ScottK: ah, so i *wasn't* the only one who didn't understand that. Good. [04:12] * Hobbsee notes that it sounds a lot like making Ubuntu more like a school yard. [04:12] ScottK, There's been a lot of talk about things like switching to bzr for packaging. You'd probably get a fair bit of support from that crowd if you support alternate sponsoring workflows. [04:13] ScottK, That said, I've not inspected the hall-of-fame code. Perhaps it uses .changes files to identify sponsoring, rather than bugs. [04:13] persia: I see. If I support that workflow, sure, but I tend to do a fair amount of "pastebin me the debdiff" or mail me the patch. [04:13] ScottK, See, the point is to get the community of people who use any sort of alternate workflow to make sure that it's using -changes parsing rather than bug review. [04:14] persia: True. I haven't either, but given the terminonolgy used it seems likely it's tied to LP bugs. [04:14] Ah. I see. [04:14] You may disagree with the bzr crowd about their workflow, but you support their right to have alternate workflows. [04:14] persia: Absolutely. [04:14] Further, by doing this, they may more support your right to have alternate workflows. [04:14] Yay for pointing out people who do well, but I hate to think how some people's egos (or whatever other terms should be used there) are going to hurt when they don't get a mention. [04:14] persia: I also like how individual source packages get described as 'projects'. [04:16] ScottK, There are several issues. Pointing them out is probably best in mail. I'm just encouraging you to start from a direction where there is already a current of support. [04:17] Hobbsee: Well all it does is point out quantity. This may or may not equate to 'well'. At least one person on the bug triagers list commonly marks bugs "Invalid" with the comment "Already reported", but doesn't bother to dupe them or answer when the reporter asks for a pointer to the bug. [04:17] * persia plans to ignore HoF completely, as 5-a-day. Neither seems to serve a useful purpose, but I strongly support people building tools to better review activity or find issues in Ubuntu. [04:19] persia: I support that too. I just wish the effort that's gone into HoF or 5-a-day would have been used for such tools. [04:20] ScottK, I see those as attempts at such tools. That I don't find them useful is a different issue. [04:21] Similar to gaspa's fairly pretty NBS monitor. It's not widely used because most of the NBS crowd didn't find it useful, but it was worth doing to try to better understand the problem. [04:21] persia: They are attempts at motivational tools. [04:21] Ah. I see them as less than ideal "review activity" tools. [04:22] Any motivational tool which causes individuals to compete doesn't seem ideal to me. If they are indeed intended as motivational tools, I think they support entirely the wrong behaviour. [04:22] Ironically some of these tools can have unintended consequences. I have an LP karma ceiling I use to monitor myself and make sure I'm not spending to much time on bugs. [04:22] persia: Agreed. [04:23] ScottK, the vast majority of the community won't appear on the HoF so I don't think there is a risk of much demotivation [04:23] jono: That's exactly why it's demotivating. [04:23] ScottK, ? [04:24] so you expect that the majority of the community will appear in a bunch of top 10 lists? [04:24] I disagree [04:24] jono: HoF appears to be a list of the top people doing the work that's important to Ubuntu. [04:24] yes [04:24] jono: There's virtually no chance I could appear on such a list. Ergo the work I'm doing, must not be important. [04:25] ScottK, oh come on [04:25] ScottK, you may feel that way, but I doubt others will [04:25] jono: Perhaps. [04:26] jono: How does it calculate sponsorship? [04:26] speak to Daniel about the details [04:26] OK. [04:45] ScottK: I think it's just an excuse for people to be put on different levels in the Ubuntu community. It makes people less equal, and is divisive. The more people care about it, the more divisive it becomes. I don't see why Ubuntu, which has a great appeal as not being a discriminatory project, and doesn't tend to have cliches, keeps moving forward in implementing this kind of thing. Good on them, though, for improving on 5-a-day, and [04:45] making it count more than just bugs! [04:46] +1 Hobbsee [04:46] heh [04:47] Well if I wanted to rank high on 5 a day, it's easy enough to do so. It wouldn't make the distro better though. [04:47] how often is the "Featured Contributors" part supposed to change? [04:47] nellery, no fixed amount, I would hope at least once a month [04:48] I like the idea of "Featured Contributors", but would have preferred it on the fridge, where I'm already looking for stuff. [04:48] jono: I seriously think focusing on listing how many of something (like bug comments) is going to encourage people to churn through large volumes of crap. It's not what we need. [04:48] * Hobbsee notes the golden pony awards had similar problems, towards the end. [04:49] ScottK, where are bug comments the metric? [04:49] you mean sponsors? [04:50] jono: It's also in 5 a day and top karma. [04:50] I think you guys are taking the Hall Of Fame a little too seriously [04:50] we were counting metrics for 5-a-day and the Ubuntu community did not melt [04:50] we are merely showcasing top ranking contributors - it really does help enthuse people [04:50] maybe not you, but others [04:50] jono: the counting metrics, as such, isn't the problem. it's then what you *do* with that information. [04:51] Hobbsee, we put it on a website, called the Hall Of Fame :) [04:51] by producing the Hall Of Fame we have not removed anything from the community [04:51] jono: so all those, along with all the horsemen are famous, and other people are mere mortals. [04:51] if people don't like it, don't look at it [04:51] Hobbsee, crap [04:51] jono, Actually, 5-a-day did a lot to decrease the value of many bug comments, and was a significant factor in the division of "triagers" and "developers" which had previously been more of a continuum. [04:52] Didn't "melt" anything, and has been a useful tool to get new people involved, but came at a cost. [04:52] persia: ++. I saw a *particularly* good one in the last week. [04:52] persia, I disagree [04:52] jono, Fair. [04:52] I disagree that 5-a-day has been detrimental to bug quality [04:52] https://bugs.edge.launchpad.net/ubuntu/+source/apt/+bug/219944 - first comment. [04:52] Launchpad bug 219944 in apt "admin should at least see a warning once for new required packages not installed on the system, even though they've removed the metapackages?" [Wishlist,Incomplete] [04:52] but it has brought more focus on bugs [04:52] jono: I think something certainly has. I'm not sure I know what. [04:53] jono, Well, we get a lot more activity. We get less deep thought. I'm not sure how that balances out. [04:53] whether that's bringing in new people, and not training them up enough, or something else is... [04:53] ScottK, very possibly, but many things affect out community, from new initiatives like the HoF, to blog entries, to the availability of resources, to negative energy [04:53] Hobbsee, don't be a troll [04:53] if you have something constructive to suggest, say it, but don't troll [04:54] jono: She's not. You're just not very open to honest feedback. [04:54] jono: not trolling. Explain how that bug is *possibly* helpful. [04:54] s/bug/response/ [04:54] ScottK, I am open, but accusing us of not training people up sounds like trolling [04:54] maybe I misinterpreting her comments [04:54] jono, The accusation is not to you. [04:54] ScottK, and as for not liking feedback, I have always been open to feedback [04:55] ScottK, is there an example of when I have not been open to feedback? [04:55] jono: well, then, if they *are* being trained, then why do they give bug answers like that? That got a comment, and marked invalid, before it got brought up in -bugs. [04:55] jono, It's that bug triagers aren't being as carefully trained as in the past. That's something for which a *lot* of people share the burden, including all participants in this conversation. [04:55] Hobbsee, there are lots of bad comments in community, irrespective of training [04:55] persia, I agree [04:55] jono: so you're just attributing it to "the community occasionally is shit" [04:56] To me, 5-a-day appears to have led to more bug comments, and to fewer questions in #ubuntu-bugs about how things work. [04:56] Hobbsee, what do you think? [04:56] persia: and it seems to me that questions in -bugs, particularly during the au day, aren't being answered. [04:56] It also appears to have decreased the number of people who are working on the bugs who care about fixing them. They appear to care more about getting the bug into the right commented shape. [04:56] persia: ++ [04:56] Hobbsee, Well, you and I should do more about that. [04:56] jono: Calling people trying to give you feedback trolls is not an example of openness IMO. [04:57] ScottK, it appeared like trolling to me [04:57] if not, I apologise [04:57] and when people do troll, I will call them out [04:57] jono: No one here is trolling, we just feel very differently than you on some points. [04:57] jono: well, I think there has to be a reason for it, and I think that trying to attribute it to "the community makes bad comments sometimes" is pathetic. [04:57] I have to be honest though, with you ScottK and Hobbsee there is always such a lot of negative energy [04:57] jono: i was not intending to troll. [04:58] Hobbsee, then I apologise [04:58] jono: maybe you look in the wrong places, then. [04:58] jono: When was the last time you heard anything negative from me? [04:58] jono: Go look at my posts on planet and find a negative one. [04:58] jono: apology accepted. Please do not always assume that i'm trolling when I disagree with you. [04:58] Hobbsee, ok [04:59] ScottK, I am talking in general - there seems a greater than often sense of criticism and negative energy [04:59] persia: I would have thought that those at canonical should, if they're the ones pushing people into working on bugs, too. [04:59] persia: Often the community members are doing other things, and can't always be relied upon to help. [04:59] Hobbsee, I believe there are reasons for bad bugs, and I am not suggesting that its just "the way it is" - I was asking earlier what you felt could be the cause [05:00] jono: Things don't get improved by not talking about problems. [05:00] I would love to see what you guys think so we could investigate [05:00] jono: I already told you. You already told me I was wrong. [05:00] I note though, there has been some good success on getting the needs-packaging bugs in a more sane state. The discussions between various members of the MOTU, and the bugsquad, seem have been profitable. [05:00] Hobbsee, See, I don't think it's Canonical's responsibility. The more that such an attitude spreads, the less Ubuntu is itself. I like Ubuntu. I want to keep it. That means that I, as a member of Ubuntu, need to do things. I think everyone should feel that way. [05:00] ScottK, agreed - but firstly problems need specific criticism and suggestions for improvements, and secondly, there needs to be a balance - celebrating what we do have and not just focusing on the bad [05:01] People tend to actually check now if the package exists in debian, before marking it as confirmed & wishlist. [05:01] jono: Yes. And as I pointed out I've recently been blogging on planet and actively trying to celebrate good stuff. [05:01] ScottK, in which case, kudos [05:02] persia: I agree with you - but I think that people, or groups, that encourage large amounts of contributions in a particular area ensure that they contribute the resources to ensure that area continues to work well, after they've brought more people into it. [05:02] jono, And on that note, while such initiatives are interesting, it might be better to work with various people to define them in the first place, as opposed to introducing such tools that might get criticism. Building things privately and introducing them as official isn't very Ubuntu. [05:02] persia: whether that be canonical, or another company that invests in ubuntu, or a new team, or whatever. [05:02] persia, I don't think we acted unofficially, we just built a tool and released it [05:02] hrm, there should be a 'should' in the above, too. [05:03] persia, people do that in the community all the time too [05:03] jono, That's precisely my complaint. I'd prefer that you had acted unofficially :) [05:03] heh [05:03] jono: I think he has a good point. [05:03] jono, It's announcing something as official without having worked with the affected teams in an open manner that bothers me. [05:03] Comment from a relative newcomer on HoF: I like the "Latest X" part of HoF, it celebrates new people coming on board and becoming more deeply involved. No so sure about celebrating who is busiest or top 5-a-day or top Lp Karma... Maybe as well as a "featured contributor" there could be a "featured newcomer" -- interview someone who just became a MOTU/DEveloper/whatever, and find out how they did it, what they liked [05:03] about the process, etc. [05:03] Most people who build tools say something like "I put together this. What do people think? Is it useful? [05:04] persia, where do you draw the line - does every project need to consult extensively while being developed? [05:04] I agree in openness, and I cherish and encourage it, but sometimes in the interests of expediency, its a case of just getting it out there and then soliciting feedback and improving === foka_ is now known as foka [05:05] jono, No, just anything that's being introduced as "official". Ubuntu is a very open project. Very strongly based on consensus. Anything that is introduced without consensus should, to me, be introduced as a candidate to build consensus, rather than something official. [05:05] I have been clear that I am seeking feedback for improvements [05:05] persia, I disagrewe [05:05] disagree, rather [05:05] I'm *much* happier with the Harvest introduction than with the HoF introduction. That was presented early, and discussed, and then became more official. [05:05] jono, With what specifically. [05:05] should each package have to be voted on to become official and in the archive? [05:05] Do you believe Ubuntu isn't open? [05:05] jono, Yes. We do that. [05:05] persia: in jono's defence, i'm not sure how you'd cover the entire ubuntu pool of contributors, to askthem for feedback, easily, apart from blogging about it. [05:06] I am talking about packages from ubuntu developers [05:06] jono, so am I. [05:06] an ubuntu developer just uploads [05:06] persia: harvest is just developers, and there are lists to contact them. HoF is everything. [05:06] surely they need to seek concensus [05:06] jono, *every* package local in Ubuntu has been reviewed by at least two Ubuntu developers, and usually three or four. [05:06] persia, what about updates [05:06] jono, There doesn't exist a single Ubuntu-only package that doesn't have peer review. [05:07] if something is uploaded, it becomes officially, those packages are not checked [05:07] jono, Updates are different. See, I'm saying that something new should be reviewed. Once it's there, fixing it is only sensible. [05:07] jono: how is that relevant,sorry? [05:07] what about bug fixes? [05:07] persia, I disagree - I think its reasonable to put out new ideas and then have people review afterwards [05:08] jono, I agree with that. I am only saying that such things should be presented as ideas, and not announced as "official". [05:08] I think for core governance we need review as part of the process, but websites can be put out there [05:08] persia, well if it appears on ubuntu.com, it could be percieved as official [05:08] jono: are you trying to say that "releasing a major community thing" is somehow on par with "uploading a bug fix, or new version of a package", and thus, it's easier to JFDI than solicit opinions from people, so that it gets done? [05:08] but then that makes any content on the wiki official [05:08] or have i missed the relevance? [05:09] Hobbsee, what I am saying is that the HoF was an idea between Daniel and I, and we did [05:09] it [05:09] jono, Right. Most other sites appear somewhere else first. This was true for the ISO tracker. It was true for harvest. It was even true to some degree for 5-a-day [05:09] we don't have a lot of time as it is, and if we would have had to seek concensus and spend time on it, it would have never been a justifiable chunk of time [05:09] as such, we threw it out there and requested feedback after the fact [05:10] jono, So you're saying that because you have access to the ubuntu.com domain, you shouldn't have to bother asking others for input before making something official? [05:10] jono: That's not really a very Ubuntu attitude IMO. "Didn't have time to work with anyone, so just did it". [05:11] persia: who would he ask for input, though? [05:11] Hobbsee, readers of his blog would work for me. [05:11] it was not a case of didnt have time so just did it anyway, it was a case of having an idea, developing it and releasing it [05:11] persia: while it does show some *very* interesting reasoning behind some of the horsemen decisions, i think this one would have been ahrder. [05:11] I am sorry if that does not fit with you expectations, but thats what happened [05:11] jono, Do you understand the source of the complaint? [05:11] if there is significant objection to the HoF then we will revoke it - so far, there has been none [05:12] persia, I can see the reasoning, but as I said earlier, I respectfully disagree [05:12] jono, Yes, but you've yet to identify with what you specifically disagree. [05:12] jono: I guess we're nobody then. [05:12] all communities will at times develop first and ask for feedback later, and while I think that for the purposes of governance and community growth that is not a bad idea, for periphery elements such as the HoF it is not such a big deal [05:12] ScottK, your words [05:13] jono, Is the perception that foo.ubuntu.com should be perceived as official something with which you disagree? [05:13] ScottK: no, it just means that anyone should be able to put forward an idea, implement it, release it, and stick it on ubuntu.com without actually asking the other communities for input. [05:13] persia, I disagree that everything needs concensus before it becomes official [05:13] this is my opinion, not necessarily the opinion of the Ubuntu project [05:13] jono, That's *how* we do things. It's precisely contrary to the basis of my understanding of the opinion of the Ubuntu project. [05:13] I agree that most of *.ubuntu.com is official [05:13] jono: You said there has been no objection to HoF so far. I guess I don't know how else to characterize this conversation. [05:14] ScottK: which, incidently, means that stuff such as the RCbugs, etc, should be very easy to get on ubuntu.com, in future :) [05:14] ScottK, sorry, no issues outside of this discussion [05:14] I don't have much opposition to HoF in and of itself, but I do have strong opposition to the idea that it doesn't take consensus in the Ubuntu project to make an official change. [05:14] jono, So in that light, why does ubuntuwire exist? [05:15] persia, the word "official" is a complex one, and I suspect difficult to define in concensus [05:15] jono, It was specifically rejected as offical, although the tools are very useful to many. [05:15] jono, So it is the perception that foo.ubuntu.com represents the Ubuntu project with which you disagree? [05:15] persia, I think I have been clear [05:16] I believe that most of ubuntu.com and sub-sites are deemed official [05:16] yes [05:16] but some content is not, such as the wiki [05:16] parts of the wiki are official, parts are not [05:16] sorry, I don't have time to debate this right now [05:16] I don't mean to blow you off, but it is 9.15pm here and I supposed to be washing dishes [05:17] if you guys want to discuss this via email, drop me a line [05:17] and I am happy to recieve any and all feedback about the HoF, and for us to discuss it [05:18] jono: As for the earlier question. Here's the answer: "Currently, I think it's a lack of training, and more of an emphasis on getting the bugs in the right states, rather than actually fixing them, like persia said. Like being encouraged to ask if people can reproduce the problem, rather than being trained to report things upstream, if they're relevant, and to at least try to debug it a bit (such as "does the package exist in debian" for [05:18] needs-packaging bugs)". Please respond to me by email, when you have more time. [05:19] what is the procedure to get a package transported to debian once we have it in ubuntu? [05:20] i am currently trying to get a package in jaunty, and iulian is helping me with it, however i would like to get a heads up on above [05:21] Hobbsee, I agree that training is a big issue, do you have any thoughts on how we can fix it? [05:21] bain, Check the BTS for an RFP or ITP. If it exists, work with that bug. If it doesn't exist, file an ITP, prepare a debian revision, and upload to Debian. If you can't upload to Debian, put it on mentors.debian.net, and ask debian-mentors@lists.debian.org for upload. You'll need to commit to maintenance. [05:22] jono, It's mostly a train the trainers thing, and reward the right activities. Long ago, bug days involved lots of developers, and were focused on closing bugs. These days they are more about reviewing new bugs, or pushing stuff upstream. [05:22] persia, thanks [05:22] bain, Note that there's no requirement for a package to be in Ubnutu to use that procedure. It's worth starting in parallel. [05:24] persia, right my thoughts exactly, i was just worried about debian process being too long, so i started with Ubuntu. [05:24] jono: a few. But most of them will probably be unpopular. [05:25] bain: I haven't found Debian to be a terrible pain, partially because much of the work has been in a fairly small team (pkg-cli-{libs,apps}) [05:25] bain, In both cases, it depends on how many developers feel like sponsoring stuff, and either is extended by issues with the packaging. Either can be quick, or long, and which is faster is a hard question to answer at any time. Debian is usually perceived as better though. [05:26] Hobbsee, You registered a retaining-developers spec. Perhaps you have time to put together a bug-to-fix-workflow-improvement one? [05:26] jono: perhaps not *that* unpopular. But will still require people sitting down and admitting there's an issue, and open to suggestions and collaboration on how to fix it - then actually doing the work in fixing it. This would involve updating things like documentation based on the progress that has been made, emailing everyone saying "this is just a reminder on how it should be done. here's the crash course", then making sure people are [05:26] around to actually help out, and people watching to make sure others *are* giving out the revised, now correct, information. [05:27] Especially with the growing popularity of PPAs, we aren't seeing as many patches as patches, and the old procedures aren't throwing a very wide net. [05:27] RAOF, well i am not talking about community/debian side delays per say, ubuntu just felt faster for various things including getting to understand packaging and finding documentation [05:28] jono: If people are unwilling to do the last couple of parts, then the entire thing is null and void. So the first question is...are people (including canonical members, as they're the team pushing lots of people into bugs) willing to? [05:28] persia: that's a point. [05:28] But i am convinced debian process is worth starting early now that i am slightly more comfortable with packaging, thanks you people [05:28] persia: (willing to help out?) [05:29] persia: I guess there would be a lot of current themes between them. [05:29] Hobbsee, In a distracted fashion. I've a stack of stuff to do in the next couple days, including sending an announcement about my fast approaching inability to access the network for a couple weeks. [05:29] persia: well, i've got 3 exams in the next 2 weeks, so i'm not around much either. Will think on it. [05:30] Maybe just a hallway discussion then, or perhaps it could be raised as a topic for someone to grab in the next QA meeting [05:30] * persia wishes this had come up yesterday, as that would have been better timing [05:30] persia: Are they convinced it's actually an issue? [05:31] Dunno. I haven't seen it discussed in a couple months. [05:31] There was a hallway discussion about it in Prague, with a number of interested parties. [05:31] If heno and such don't see it as an issue, then raising it, to try and get it fixed, is a waste of time and energy. [05:31] was heno there? [05:32] heno had some interest in Prague towards improving the fix ratio, but that may not translate into action items without further discussion. [05:32] the fix ratio. Right, hmmm... [05:33] On the other hand, such a discussion could already be planned. I haven't seen too many of the QA track specs on LP yet. [05:33] * bain is away: "work" [05:33] * ScottK is away: "sleep" [05:33] * Hobbsee is here: "studying" [05:34] Hobbsee, Right. The problem is that there exist fixes to bugs that aren't in the repos. There are likely a wide number of opinions about the solution. [05:34] It's that discussion that's likely worthwhile to have, and put together a list of candidate solutions and easy action items to try towards improving the situation. [05:35] Darn. Just remembered.... [05:35] persia: oh, right. I thought the current discussion was that people were 'triaging' things the wrong way, and needed to be trained how to do it right, and be more useful [05:35] * ScottK is away: "cleaning the kitchen - then sleeping" [05:36] this sounds like facebook onto irc... [05:37] Hobbsee, I perceive that to be one element of the larger issue. Adjusting triage training to focus on getting bugs to Fix Released would help with the fix ratio. [05:37] On the other hand, that focuses on the actual problem, rather than on what we perceive to be one of the causes. [05:37] sorry Hobbsee, had to step away [05:37] Hobbsee, I agree, training is where we need to focus our efforts [05:37] persia: er, define "that" please? [05:38]