[01:01] cjwatson: https://blueprints.launchpad.net/ubuntu/+spec/servercloud-r-fastpath-install [01:01] interesting new blueprint. [01:02] Yeah, I know about it [01:02] ack =) [11:36] Hi any wubi developer's here? [11:44] sure. [11:44] smartboyhw: what's up? [11:44] xnox, we want to add Ubuntu Studio to Wubi [11:45] smartboyhw: ok. [11:45] smartboyhw: that needs correct links to how/where your images are called. [11:45] xnox, just wondering what should I write in the isolist.ini. Should I use the 12.10 metalinks or the 13.04 links [11:46] smartboyhw: well, currently wubi is targetting 12.10 & we didn't update the links for the 13.04. [11:46] xnox, so the 12.10 links then [11:46] smartboyhw: for now yeah, and then we will update all of them to 13.04 in bulk. [11:47] xnox, same for the debian/changelog? [11:47] smartboyhw: what do you mean for the debian/changelog? $ dch -i and keep it UNRELEASED. [11:48] xnox, I mean the version number in debian/changelog. Is it that I should use 13.04? [11:48] smartboyhw: don't worry about debian/changelog, it's a bit out-of-date and we need to fix it up =)))) [11:48] =)))))0 [11:51] xnox, is it that we must have lupin-support package in our images in order to install Ubuntu Studio on Wubi? [11:51] From Bug 1070682 [11:51] Launchpad bug 1070682 in Ubuntu Studio "wubi and ubuntu studio" [Undecided,In progress] https://launchpad.net/bugs/1070682 [11:53] smartboyhw: sounds about right, have you seeded it already? [11:53] xnox, that is the problem . No [11:53] Let me find the guys in my team to add it [11:55] xnox, told them to add...Just wait till it seeds to ubuntustudio-meta [11:59] xnox, hmm I now got the code in https://code.launchpad.net/~smartboyhw/ubuntu/raring/wubi/bug-1070682 but I can't see how to propose merge proposals [12:00] smartboyhw: huh.... wubi predates package branches. base your changes on lp:wubi & push to lp:~/wubi/bug-1070682 and propose into lp:wubi. [12:00] smartboyhw: ubuntu archive does not build windows binaries yet.... [12:00] xnox, so I do like lp:~smartboyhw/wubi/bug-1070682 right? [12:01] smartboyhw: you have just created wubi package in ubuntu =((((( because of that branch.... [12:01] xnox, oops [12:01] xnox, immediate delete [12:01] smartboyhw: in the future make sure you push back branches to the same project you got them from...... [12:02] Phew got it now [12:02] iirc using push :parent actually makes suggestions for branch names if you cant push due to missing credentials [12:03] that should give a proper "to-merge" branch name [12:03] Got it now xnox https://code.launchpad.net/~smartboyhw/wubi/bug-1070682/+merge/134446 [12:05] It doesn't need to go in ubuntustudio-meta [12:05] The seed it should probably be in is not one that generates a metapackage [12:06] (Kaj did it right) [12:07] cjwatson, yeah [12:18] cjwatson, sorry on the edubuntu typo [12:18] fixing [12:22] cjwatson, resubmitted proposal [12:23] https://code.launchpad.net/~smartboyhw/wubi/bug-1070682/+merge/134453 [12:45] FWIW you should never resubmit proposals [12:45] At least not for minor corrections like this [12:45] The way you use the system is that you simply commit a fix and push it [12:45] Resubmitting is awkward and overcomplicated and basically just for the situation where you proposed against the wrong branch [12:47] cjwatson, hi, is bug 220961 that Ubiquity never checks for enough disk space, or that it normally checks but is sometimes mistaken? [12:47] Launchpad bug 220961 in ubiquity (Ubuntu) "[MASTER] ubiquity crashes instead of notifying the user of not enough disk space" [High,Triaged] https://launchpad.net/bugs/220961 [12:47] I'm just trying to work out whether it needs a design for the error message. [12:48] smartboyhw: At the risk of blowing my own trumpet, see https://code.launchpad.net/~cjwatson/launchpad/refactor-cron-germinate/+merge/84624 for an example; if you scroll down you can see review comments nicely interleaved with revisions fixing them [12:48] mpt: It normally checks but is (probably inevitably) sometimes mistaken. The bug is that it falls over in varying different kinds of heaps rather than dying gracefully. [12:48] mpt: there are conditions/code-paths in ubiquity which allow you to proceed to installation and later it will trip over itself that there is not enough disk space. [12:49] ok, thanks [12:49] cjwatson: the classic is Laney's "typical install": 8GB disk + 8GB RAM in a VM, with swap taking over most of the space and / being the minimal size of ~800MB (whatever hardcoded in partman-auto recipe as minimal size) [12:50] which falls flat on it's face. [12:50] Yeah, we have separate bugs for lots of the special cases of this kind of thing [12:50] The "ridiculous amounts of swap" case is one [12:51] cjwatson: interesting this blueprint got superseeded: https://blueprints.launchpad.net/ubuntu/+spec/servercloud-r-default-swap-disk-size-calculation [12:51] *shrug* I lose track [12:51] not sure what happened there.... [12:51] But that's the sort of thing that's stupid as a blueprint anyway [12:51] It's just a bug, albeit a complicated one [12:51] yeah.... blueprints are funny. [12:51] cjwatson, have you considered using RecoverableError for that? Then you could pick off the causes in order of frequency [12:51] No [12:51] Blueprints are bug reports with delusions of grandeur [12:52] Because I've only spent a fairly minimal amount of time on ubiquity at all since errors.u.c et al were created, and that really only in support of other project [12:52] s [12:52] fair enough [12:52] I should have addressed that to xnox :-) [12:52] There's a point to blueprints when we need complex design across multiple pieces of software, but they do get massively overused [12:52] That said, it should only be a recoverable error if it's actually recoverable; half the point of the bug is that it isn't [12:52] mpt: I like that definition =) [12:53] (Except by restarting the installation from scratch, which I don't think counts) [12:54] cjwatson, that might be reading a little too much into the name ... Recoverable in that sense only means "doesn't crash". [12:54] Maybe the name is wrong? :-) [12:54] maybe. [12:54] I'm reading the dictionary words [12:54] (I didn't choose it, therefore it must be, right?) [14:53] cjwatson, so you are just not going to merge the fix since I (stupidly) resubmitted the proposal? [15:10] smartboyhw: Where did I say that? [15:10] smartboyhw: I'm not merging it since it's my day off today :-) [15:11] smartboyhw: I was just recommending that you use an approach that's both easier and better in future [15:13] cjwatson, ah OK:P