| === nhandler is now known as Guest26812 | ||
| === nhandler_ is now known as nhandler | ||
| === james_w` is now known as james_w | ||
| === Guest31934 is now known as ogry | ||
| === ogry is now known as ogra | ||
| === sanchaz-away is now known as sanchaz | ||
| === sanchaz is now known as sanchaz-away | ||
| === almaisan-away is now known as al-maisan | ||
| === al-maisan is now known as almaisan-away | ||
| njin | hello, just a curiosity, why asterisk comment out lines with ; instead # ,is a devel. decision, language related or what else? | 12:09 |
|---|---|---|
| PetrHH | Hello, I'm trying to create deb package for the first time. A got over a few problems but hang on this: dpkg-genchanges: error: badly formed line in files list file, line 1 | 12:20 |
| PetrHH | Anybody know where could be a problem? | 12:20 |
| zyga | njin: perhaps lisp reference? | 12:24 |
| njin | zyga: thanks | 12:25 |
| njin | these lines are in the conf file | 12:25 |
| njin | *files | 12:25 |
| soren | njin: You could ask in an asterisk channel instead.. | 12:30 |
| njin | soren: thanks | 12:31 |
| Hobbsee | PetrHH: wrong formatting of debian/changelog. You should use dch (-i) to edit it. | 12:57 |
| Hobbsee | rather than by hand | 12:57 |
| PetrHH | Hobbsee, I already found the problem. Interesting is, that problem were in control file, I've chosen wrong category | 13:03 |
| Hobbsee | PetrHH: ah. didn't recall that genchanges looked at the control file | 13:03 |
| === yofel_ is now known as yofel | ||
| PetrHH | Hobbsee, It is OK, thank you for response | 13:06 |
| Hobbsee | np | 13:06 |
| PetrHH | Hobbsee, Where could I find section list for Section in control file? | 13:16 |
| Hobbsee | PetrHH: by googling debian packaging guide section, normally | 13:18 |
| * Hobbsee heads to bed | 13:18 | |
| Hobbsee | it's the most comprehensive, useful place for all those fields | 13:18 |
| PetrHH | Hobbsee, I've read it but sometimes that manual is not so clear for me | 13:25 |
| zyga | james_w: hi, around? | 13:25 |
| === dendro-afk is now known as dendrobates | ||
| === Quintasan_ is now known as Quintasan | ||
| Keybuk | zyga: the open source process does work | 15:15 |
| zyga | Keybuk: :-) | 15:33 |
| zyga | Keybuk: did you commit the patch to dbus? | 15:33 |
| zyga | Keybuk: nothing new on your discussion with Lennart | 15:33 |
| zyga | cr3: hi | 15:34 |
| cr3 | zyga: hey dude, happy holidays! | 15:45 |
| zyga | cr3: :-) | 15:45 |
| zyga | single calm week of the year, eh? | 15:46 |
| cr3 | zyga: yep, enjoying every minute of it, checking email once in a while though | 15:46 |
| cr3 | zyga: otherwise, I won't get any work done before the rally | 15:47 |
| zyga | nice | 15:47 |
| zyga | I'm doing some hacking as always | 15:47 |
| zyga | but focusing on doing stuff for myself and not strictly for work | 15:47 |
| cr3 | zyga: what kind of personal project(s) do you work on these days? | 15:47 |
| zyga | not really personal, just dependencies I created while working on the real stuff | 15:48 |
| cr3 | zyga: have you looked into the go language yet, seems to be an interesting topic on the tech mailing list | 15:48 |
| zyga | I've been moving them to pypi today | 15:48 |
| zyga | cleaning stuff up | 15:48 |
| zyga | using virtualenv for the first time :-) | 15:48 |
| cr3 | zyga: cool, what stuff on pypi? | 15:48 |
| zyga | currently versiontools (just uploaded) | 15:48 |
| zyga | but I'll transition over every other package | 15:48 |
| zyga | so linaro-json linaro-dashboard-bundle | 15:49 |
| zyga | django-testscenarios | 15:49 |
| zyga | django-restricted-resource | 15:49 |
| zyga | django-reports | 15:49 |
| zyga | and finally launch-control and launch-control-tool | 15:49 |
| zyga | pypi gives me two things that I love: | 15:49 |
| zyga | update | 15:49 |
| zyga | and virtualenv | 15:49 |
| cr3 | zyga: you're moving all that to pypi? | 15:49 |
| zyga | yes | 15:49 |
| zyga | but moving != removing lp.net tuff | 15:50 |
| zyga | lp.net still hosts everything apart from tarballs | 15:50 |
| zyga | pypi just has the releases and the meta-data so that I can setup.py any of my work on other systems without bothering with packaging | 15:50 |
| zyga | (untill pkgme is ready for prime time) | 15:50 |
| zyga | oh | 15:51 |
| zyga | and pypi has one extra bit that lp lacks | 15:51 |
| zyga | it can host documentation | 15:51 |
| zyga | so I started building proper documentation for my projects | 15:51 |
| zyga | and I can upload_sphinx them to pypi | 15:51 |
| zyga | killer for me :-) | 15:51 |
| * zyga always regarded lp as a poor place to host one's project | 15:51 | |
| zyga | it has tons of "stuff" | 15:52 |
| zyga | but it lacks in few key details that matter for non-ubuntu-like projects | 15:52 |
| cr3 | zyga: that's unfortunate, couldn't lp.net be extended to do the documentation thing? sounds like a nice self contained project | 15:53 |
| zyga | cr3: I asked for that a few months ago when linaro's input was taken to the stakeholder process for lp | 15:54 |
| zyga | cr3: also the lack of need to open a browser is a big win for me | 15:54 |
| zyga | I currently work from my efika, I removed all the heavy stuff I could and I work with pure console + screen | 15:54 |
| cr3 | zyga: as I once said, ask not what lp can do for you but what you can do for lp :) | 15:55 |
| zyga | (it's crazy fast in that setup for A8) | 15:55 |
| zyga | cr3: I'm not going to fool myself that lp.net can be extended by mere mortals | 15:55 |
| zyga | cr3: especially since it's not a problem that needs solving | 15:55 |
| cr3 | zyga: you are not a mere mortal :) and if you find the project documentation feature useful on pypi, seems useful | 15:56 |
| zyga | cr3: lp.net is driven by ubuntu's needs, that's to be expected, when/if the goals will change to directly compete with other "solutions" like that then it can be changed properly | 15:56 |
| cr3 | zyga: it is indeed driven by ubuntu's needs, but it's vocation is not necessarily limited to it | 15:57 |
| cr3 | zyga: ubuntu or no ubuntu, I agree with you that the documentation feature on pypi is pretty cool :) | 15:57 |
| zyga | cr3: I _might_ be able to do that given lots of effort but frankly the last time I tried to run lp.net on my system it was really complex and I don't think that a hacky solution that I would be happy to work with would fly with upstream's coding quality standards. I don't have resources to go away for a month to fix lp.net to host my projects better. | 15:57 |
| zyga | cr3: I know, I think it's quite strange that there is no 80/20 split | 15:58 |
| zyga | or at least I don't see it | 15:58 |
| zyga | 80% effor for ubuntu | 15:58 |
| zyga | 20% for community | 15:58 |
| zyga | but ask community first :) | 15:58 |
| zyga | having a static website would solve lots of lots of grief I have with lp | 15:58 |
| cr3 | zyga: your concerns have already been expressed by others, you are not alone :) the lp.net folks totally realize the barrier to entry is very high, they call it the typo test, ie how long does it take to fix a typo in launchpad | 15:58 |
| zyga | cr3: other than that there is pypi namespace - if I register my project there there will be no conflict in most python (hopefully) projects _and_ it gets picked up by pip | 16:00 |
| zyga | cr3: lp.net _could_ do that but it's not clear how namespace resolution would work | 16:00 |
| cr3 | zyga: wgrant is one of those community members who's done tremendous work on lp.net, very cool dude | 16:00 |
| zyga | wgrant: thanks :-) | 16:00 |
| zyga | I would love to help in such effort (static pages for projects) but I'd rather help than lead as I'm not experienced with zope and other standard lp.net bits or the whole development policy | 16:01 |
| cr3 | zyga: the lp.net folks are making a lot of work to simplify the code base, we should start reaping the rewards in six months to a year. that's very short, time flies like an arrow... | 16:02 |
| * cr3 hugs lifeless and the rest of the lp.net folks | 16:04 | |
| Keybuk | zyga: at least it has turned into the right discussion :p | 16:09 |
| zyga | Keybuk: yes, but I see that's still a long way from "you and him are friends" | 16:10 |
| Keybuk | heh, we are friends, Facebook says so :-) | 16:10 |
| Keybuk | we just don't agree technically | 16:10 |
| === tkamppeter_ is now known as tkamppeter | ||
| zyga | Keybuk: I'm curious what has to happen till you can commit that? | 16:13 |
| zyga | perhaps can is the wrong word, 'will' | 16:15 |
| * mrenouf|work *yawn* | 16:16 | |
| === SolidLiq is now known as solid_liq | ||
| Keybuk | zyga: oh, next step is someone writes a new mail to start a new thread | 16:37 |
| Keybuk | with the complete proposal | 16:37 |
| Keybuk | I figure I'll do that later this week (so everyone's had a chance to comment) | 16:37 |
| Keybuk | and then we discuss again, and a consensus forms, etc. | 16:38 |
| zyga | Keybuk: complete proposal? Have you already commited the changes discussed in that thread? | 16:38 |
| Keybuk | no, not yet | 16:38 |
| Keybuk | now it looks like the existing systemd patches will be backed out of d-bus | 16:38 |
| Keybuk | in favour of a shared implementation | 16:39 |
| ion | nice | 16:39 |
| zyga | Keybuk: insane in a way :) | 16:41 |
| zyga | Keybuk: why did you not object earlier? | 16:41 |
| Keybuk | at the time Lennart posted the patches, I wasn't subscribed | 16:41 |
| zyga | ah I see | 16:42 |
| * zyga is happy to see stuff go in the right direction | 16:42 | |
| zyga | I though about upstar/systemd duality | 16:42 |
| Keybuk | but in a kooky way, maybe this way is better anyway | 16:42 |
| zyga | that we could actually live with both systems indefinitely | 16:42 |
| Keybuk | we could, it's going to be no different *really* than the fact we live with dpkg and rpm | 16:42 |
| zyga | yeah | 16:42 |
| Keybuk | the fact both are declarative configs is quite useful, it means in theory you could write stuff to convert between them | 16:43 |
| zyga | (except that because of that upstream hate to package stuff for "linux" and just ship sources thus breaking the user-developer loop) | 16:43 |
| Keybuk | (except maybe for the start/stop conditions, which is where the two differ anyway) | 16:43 |
| zyga | ~ that's somewhat possible but it would be ugly | 16:43 |
| Keybuk | zyga: right, but you can't package "for linux" already because of the deb/rpm divide | 16:43 |
| Keybuk | so you're already packaging "for debian" or "for redhat" | 16:43 |
| zyga | yes | 16:43 |
| zyga | and that sucks :-) | 16:43 |
| Keybuk | I'm reactivating my Debian membership and will be doing Upstart stuff there | 16:44 |
| Keybuk | (too) | 16:44 |
| ion | cool | 16:44 |
| zyga | I'm interested if pkgme, our new tool still in progress might change that | 16:45 |
| zyga | so that linux world would move closer to upstream packaging | 16:45 |
| === ogra_ac is now known as ogra | ||
| Keybuk | zyga: I doubt that'll ever happen | 16:58 |
| Keybuk | it's not in distro interests to support it | 16:58 |
| zyga | Keybuk: are you certain? | 16:59 |
| zyga | Keybuk: do you remember mpt's presentation about 3rd party development for ubuntu | 16:59 |
| zyga | Keybuk: if anything it screams 'we suck' | 16:59 |
| zyga | Keybuk: and suggest 'we may have to change how things happen to fix that' | 17:00 |
| ebroder | Keybuk: I could see motivation for distros to support it, if they can be convinced that it would improve the overall Linux app ecosystem | 17:00 |
| Keybuk | sure, I enjoyed mpt's presentation greatly | 17:00 |
| Keybuk | and ev's | 17:00 |
| Keybuk | but distros exist right now to provide lock-in | 17:00 |
| Keybuk | if you could sort the packaging problem such that upstreams did their own packaging | 17:00 |
| zyga | Keybuk: that's not really true, that's a side effect of our technology | 17:00 |
| Keybuk | what would there be left for a distro to *do* ? | 17:00 |
| zyga | Keybuk: package the core platform | 17:01 |
| ebroder | That seems like a narrow view. I'd argue that distros exist to provide cohesion | 17:01 |
| Keybuk | the core platform? | 17:01 |
| Keybuk | there is no core platform | 17:01 |
| Keybuk | Linux is simply a collection of upstreams | 17:01 |
| zyga | Keybuk: there would be, defined by each distro | 17:01 |
| Keybuk | (and you forget, distros invented packaging, not the other way around) | 17:01 |
| zyga | Keybuk: then upstreams would dedice if they want to package for that core platform | 17:01 |
| zyga | Keybuk: not true | 17:01 |
| zyga | Keybuk: distros formalized packaging beyond all hope | 17:01 |
| zyga | Keybuk: the act of distributing software was older than first linux distro | 17:02 |
| zyga | Keybuk: and worked differently | 17:02 |
| Keybuk | the act of distributing source was | 17:02 |
| zyga | Keybuk: not soure | 17:02 |
| zyga | Keybuk: packaging is not about source really | 17:02 |
| zyga | (well unfortunately too I guess) | 17:02 |
| zyga | Keybuk: but nobody is keenly interested in distributing source | 17:03 |
| zyga | Keybuk: do you see anyone being particularly happy that they can distribute source to anyone? | 17:03 |
| Keybuk | upstreams quite like it | 17:03 |
| Keybuk | it's minimum touch | 17:03 |
| zyga | Keybuk: the true distribution is to the end user | 17:03 |
| zyga | Keybuk: upstreams hope to reach their audience but in our current model that involves going via the distro | 17:03 |
| zyga | at least, on linux | 17:03 |
| zyga | Keybuk: what do you think about KDE shipping _official_ packages for windows and osx | 17:04 |
| Keybuk | depends which upstream really | 17:04 |
| Keybuk | that's part of our problem | 17:04 |
| zyga | Keybuk: it's an example of upstream packaging and distribution that bypasses the middleman | 17:04 |
| Keybuk | we think all upstreams want the same thing :p | 17:04 |
| Keybuk | quite a lot of our upstreams are actually Fedora :p | 17:04 |
| zyga | that's true | 17:04 |
| zyga | heh :-) | 17:04 |
| zyga | yeah it's not a simple problem | 17:04 |
| zyga | but I think the common truth is that we have to change something to get that nice explosion curve of new software | 17:05 |
| zyga | we don't have to forfit our current model but we should extend it to new use cases that take distribution out of the way | 17:05 |
| zyga | essentially the way the vast majority of software (open or not) used today is distributed differently and for a reason | 17:06 |
| zyga | while the appstore/gateway concept is cool and might again change things | 17:06 |
| zyga | our appstore sucks :) | 17:06 |
| maco | *snort* | 17:06 |
| maco | i think it's funny how new people whined about repos for years, and then apple does it on the iphone and suddenly repos, oops i mean "appstores" are cool! | 17:07 |
| zyga | nope | 17:07 |
| zyga | maco: apple did not do "a repository" | 17:07 |
| maco | thats all it is | 17:08 |
| zyga | maco: they did something quite different in the way it works | 17:08 |
| maco | the difference being ya pay? | 17:08 |
| maco | to the user, there's no difference | 17:08 |
| zyga | oh please, think not mock | 17:08 |
| maco | except paying ;) | 17:08 |
| zyga | sure there is | 17:08 |
| zyga | the difference is that there is no concept of a depdency on anything except platform version | 17:09 |
| maco | there are two ways users know to get software: 1) search the internet and download from some random webpage or 2) use a bit of software the lists available programs and tell it to fetch them | 17:09 |
| maco | appstore & synaptic are both #2 | 17:09 |
| zyga | and that the upstream ship binaries not source, if your binary matches the rigid form that is required then you are allowed in (I'm referring to filesystem structure, not the review process) | 17:09 |
| maco | oh you're just talking about style of compiling! static versus linked | 17:09 |
| zyga | maco: except that our model makes tons of problems that they don't have | 17:10 |
| zyga | maco: again, a model that we use does not scale | 17:10 |
| maco | and saves gigs of hard disk space | 17:10 |
| zyga | maco: gigs of space is not the problem | 17:10 |
| maco | and tons of bandwidth | 17:10 |
| zyga | maco: I doubt that code is > 1% of the overall app size on mobile platforms | 17:10 |
| zyga | the difference is that you don't solve the packaging problem | 17:10 |
| zyga | there is no insane complex policy | 17:11 |
| zyga | no insane manual review | 17:11 |
| zyga | just automatic and very constrained process | 17:11 |
| zyga | and you just delegated packaging to upstreams | 17:11 |
| maco | i rather like not having each package be several hundred megs due to static compiling. i only install packages that are fewer than 3 mb ordinarily, otherwise i have to go find some other internet to "borrow" if i want it to take less than a day to download | 17:11 |
| zyga | maco: they are not several 100s megs really | 17:11 |
| * ogra curses the lucid kernel | 17:11 | |
| zyga | maco: because the platform has virtually everything that you may want to use in an average app, the rest is just your code most of the time | 17:12 |
| zyga | maco: it's not a model where lots of our stuff fits because we like to ship 4 different things that do foo while they bless one and force you to use it but it seems to work | 17:12 |
| maco | im thinking of those dmg packages on osx. they are usually MUCH larger than packages im used to seeing | 17:13 |
| qb89dragon | hey, I'd be grateful if someone could advise me how to modify cupsd.conf during the install of a deb package in my rules file. Is it simply adding it in there, or is there some process for registering the changes made with dpkg? | 17:13 |
| zyga | maco: but to say that our model is the same is to ignore the problem | 17:13 |
| zyga | maco: I'm not sure that that works on osx, perhaps because they still use fat binaries? | 17:13 |
| ebroder | zyga: No, maco's right on that point in particular. OS X packages are big because they ship all their dependencies with them (i.e. a Qt app comes with a copy of Qt) | 17:14 |
| zyga | ebroder: that's true | 17:14 |
| zyga | ebroder: on osx you either use cocoa or bundle | 17:14 |
| zyga | that's a limitation | 17:14 |
| zyga | but again it's something they did to vastly simplify their platform | 17:14 |
| zyga | on the mobile you don't want 10s of toolkits | 17:14 |
| zyga | and version interedependenices between them and all apps, it's just a waste of qa and space | 17:15 |
| zyga | observe how despite this limiattion there has been an explosion of apps on both iOS and android | 17:15 |
| zyga | where both use identical (no deps) model | 17:15 |
| zyga | and it has promoted innovation and third party development with 1st party packaging | 17:16 |
| zyga | (where 1st party is the same 3rd party I mentioned, sorry about that sounding oodd) | 17:16 |
| ebroder | zyga: I'm not disagreeing with you in general | 17:17 |
| zyga | ebroder: in our world it could also work | 17:17 |
| zyga | ebroder: where we would extend current methods with a hybrid platform | 17:17 |
| zyga | you either use the old methods and packaging is a little more complex | 17:17 |
| zyga | or this is the "base platform" we ship and you have to bundle against that without deps | 17:17 |
| zyga | I thin that 2nd option is the one non-open aps would thrive on | 17:18 |
| zyga | from the user point of view _nothing_ changed | 17:18 |
| zyga | from developer's point of view it's a whole new world | 17:18 |
| zyga | no deps, no conflicts, no alternatives, no triggers, no scripts, atomiciy, need I conitnue? | 17:19 |
| zyga | all of that are improvements | 17:19 |
| zyga | perpahps the first comma should be replaces with => | 17:19 |
| ebroder | That I disagree with. I don't think we need to drop the concept of dependencies. If we do that, then an app developer can *only* use the libraries that we ship as part of the base system. We'd end up under pressure to include libraries and such just for 3rd party apps. | 17:20 |
| ebroder | Dependencies are good. And they can be largely be automatically calculated | 17:21 |
| ebroder | Even build-deps can be automatically calculated for many apps | 17:21 |
| zyga | ebroder: that's not the problem | 17:22 |
| zyga | ebroder: the problem is that we also ship _ONE_ version of a package | 17:22 |
| zyga | ebroder: and that just breaks things | 17:22 |
| zyga | if A nees foo 1 and B needs foo 2 | 17:22 |
| ebroder | Yeah, but there are *really* strong advantages, too | 17:22 |
| zyga | (and we don't ship libfoo*) | 17:23 |
| zyga | ebroder: I agree our model has upsides | 17:23 |
| cdbs | I can see many recent uploads stating: Fix build failure due to binutils-gold , and it turns out that the package doesn't even *use* ld.gold, nor does it have binutils-gold in its b-d. Is the uploader's fault here, or am I misunderstanding? | 17:23 |
| ebroder | Security support, for instance | 17:23 |
| cdbs | doko_: ^ | 17:23 |
| zyga | ebroder: that's true but only to a point | 17:23 |
| ebroder | zyga: I think that most people who ship bundled libraries don't even think about or are aware of security issues in those libraries, with the exception of the giant players like Google | 17:23 |
| zyga | ebroder: when the link between upstram and downstream is maitained by a distro | 17:23 |
| zyga | ebroder: then if the distro chokes with the amount the process stops working and quality falls | 17:24 |
| penguin42 | cdbs: I think there are moves to make sure everything works with gold | 17:24 |
| zyga | ebroder: in a hybrid model we would have market forces impacting quality | 17:24 |
| zyga | ebroder: and thus you would get crappy stuff _and_ good stuff | 17:24 |
| ebroder | zyga: I don't think market forces are good enough to ensure some factors of quality | 17:25 |
| zyga | ebroder: right now you just get average stuff, distro is not perfect - we cannot maintain the universe for examlpe | 17:25 |
| cdbs | penguin42: So, ubuntu aims to move to gold one day? | 17:25 |
| zyga | ebroder: sure they are, apple seems to do just fine, ;-) | 17:25 |
| zyga | ebroder: it's not a simple discussion, I don't mean to say that it would work better with code security | 17:25 |
| kklimonda | ebroder: and yet it works for windows ans osx | 17:25 |
| penguin42 | cdbs: I've heard it suggested for some platforms; not sure what the current story is | 17:25 |
| ebroder | kklimonda: No, people accept the way it's done on Windows and OS X. That's different | 17:25 |
| zyga | ebroder: but that it would change the dynamics in our favour | 17:25 |
| zyga | ebroder: all depends on how you set our goals | 17:26 |
| zyga | ebroder: if our goal is to have uber security then you have to just stick to main | 17:26 |
| zyga | ebroder: and current rules | 17:26 |
| zyga | ebroder: _and_ have enough mainitaners talking with upstream | 17:26 |
| zyga | ebroder: if our goal is to grow then this model will not work as well as a hybrid model | 17:26 |
| zyga | ebroder: people on those other platforms still have quality problems but most of that is IMHO a derivative of the base platform | 17:27 |
| ebroder | zyga: Anyway, I think you're in general oversimplifying the situation. But I need to step away for a bit to focus on this assembly I'm trying to understand :) | 17:27 |
| zyga | hehe | 17:27 |
| zyga | ebroder: I don't think that it's possible to understand and predict this without trying to do it really | 17:27 |
| === Keybuk_ is now known as Keybuk | ||
| doko_ | cdbs: gold has the same semantics as ld --no-add-needed, see http://wiki.debian.org/ToolChain/DSOLinking | 19:57 |
| zyga | doko_: --no-add-needed or --no-as-needed? | 20:17 |
| doko_ | zyga: --no-add-needed, please see the wiki page | 20:25 |
| zyga | doko_: Interesting read, thanks! | 20:34 |
| === warp11 is now known as warp10 | ||
| xnox | maco, ping =) | 21:20 |
| maco | what i do? | 21:20 |
| randomOfAmber | hello world :D | 21:20 |
| randomOfAmber | oops | 21:20 |
| xnox | maco, thanks a lot for the comment on my blog post "how to get a job in FOSS" | 21:20 |
| xnox | I can't believe the famous maco spared a comment =) | 21:21 |
| maco | ahahahahahaha | 21:21 |
| xnox | You rock ;-) | 21:21 |
| maco | you know mako and i are two different people right? | 21:21 |
| xnox | Yes. | 21:21 |
| xnox | You are famous enough for me ;-) | 21:22 |
| maco | thanks | 21:22 |
| xnox | It sucks that the path into FOSS is via unpaid internships based on all the comments I have gathered. | 21:23 |
| maco | unpaid internships are the norm in the US | 21:23 |
| maco | into ANY field, unpaid internships are just what students do | 21:23 |
| * micahg thought GSoC was paid | 21:23 | |
| xnox | My job post did generate invite to join a certain company in the UK which does Linux Support | 21:23 |
| maco | yeah GSoC is paid | 21:23 |
| xnox | micahg, GSoC was paid =) | 21:23 |
| ebroder | maco: I don't agree with that. There are some industries where unpaid internships are standard, but I don't think that's the case for software dev | 21:24 |
| ebroder | Though FOSS dev may be a subset where that is the norm | 21:24 |
| maco | micahg: thats why everyone wants to do gsoc | 21:24 |
| xnox | Igalia, Collabora, Codethink & Canonical - is all there is, it seems in the EU | 21:24 |
| maco | ebroder: depends if its at a tech company or doing tech for a nontech place i think | 21:25 |
| maco | ebroder: i got paid interning at a security company after 2nd year, but doing web dev for a newspaper during 2nd year i didnt | 21:25 |
| xnox | I had quite high paid engineering internships / full-time jobs. (E.g. I was paid as a student more than graduates) but it wasn't related to Electronics nor Software Development. Which is what my degree is in. | 21:26 |
| xnox | and what i want to do now. | 21:26 |
| maco | changing focus is tough :-/ | 21:27 |
| xnox | maco, Yeap.... | 21:27 |
| maco | i keep getting "hey wanna be a sysadmin?!" and im like "no, im a programmer" "but you have sysadmin experience" "because i wasnt a good enough programmer back then!" | 21:27 |
| xnox | micahg, /usr/lib/pkgconfig/libxul-embedding.pc is broken by the way =) in Natty / xul20 | 21:28 |
| micahg | xnox: yes, I forgot about that, you filed a bug, right? | 21:29 |
| xnox | micahg, han =) was too lazy will file now.... | 21:29 |
| micahg | xnox: ok, thanks | 21:29 |
| maco | xnox: did you just spell "nah" backwards? | 21:30 |
| xnox | maco =) yeap | 21:30 |
| * xnox is just back from ER and I'm on loads of painkillers. | 21:30 | |
| xnox | When I start my own company i will do paid FOSS internships. | 21:31 |
| ebroder | xnox: I remember seeing your blog post but can't seem to find it at the moment. Have you looked into, say, Red Hat? I would have thought they had offices in Europe | 21:32 |
| xnox | ebroder, I have applied to Red Hat. They are based in Czech Republic for software stuff. The interships are to do with Java and their java apps like JBoss | 21:33 |
| ebroder | Oh, huh. That's kind of unfortuate | 21:33 |
| xnox | which are proprietarish | 21:33 |
| xnox | I have also applied to Intel/Nokia < MeeGo. | 21:33 |
| * xnox there are no internships with Linaro..... | 21:34 | |
| maco | xnox: linaro is hiring for a few mentors though | 21:34 |
| maco | xnox: they need people who can show folks from big hardware companies the ropes when it comes to free software development, using launchpad & bzr, etc. | 21:35 |
| maco | (at least, when i talked to them at uds they were...) | 21:35 |
| xnox | Yeap I so job postings for mentors. But they don't seems to go away. Either they want a lot of people like that or they can't get them. | 21:36 |
| xnox | maco, I'm jealous about UDS =) I haven't been to any yet. | 21:36 |
| maco | xnox: i actually interviewed for one of those spots while at UDS, but im pretty solidly application-level so didnt get it. if you know hardware-level stuff though, you'll be in better shape | 21:37 |
| maco | i think theyre not going away because theyre not finding people who understand hardware AND are familiar with canonical's development tools...other than the folks they took from other canonical teams | 21:38 |
| xnox | they stuff they taught me in uni was - outdated, to say the least. I can play around with oscilliscope and do basic troubleshooting. But I haven't done debbugging on hardware with e.g. JTAG | 21:38 |
| xnox | hmmm.... | 21:39 |
| xnox | I did apply for internships with ARM. If I get that maybe I'll be able to go for Linaro next ;-) | 21:39 |
| xnox | maco, what's typical response time? Russians replied in 2 working days for all my applications. UK/EU seems to take 4 weeks + | 21:40 |
| maco | canonical's is supposed to be 3 weeks | 21:40 |
| maco | dunno about elsewhere. in the US, you often only get told "yes" and are simply ignored for "no" | 21:41 |
| xnox | maco, did you apply for the Gnome Internship - Outreach for Women? | 21:41 |
| maco | no | 21:41 |
| maco | im a kde user... | 21:41 |
| xnox | maco, same here in EU so far. with responses. It's like I tell them everything about myself and my life and they don't even bother to reply. | 21:41 |
| maco | tbh, ive not done job applications much before. i put my cv on monster and then i get contacted by others | 21:42 |
| xnox | maco, You could have been working on Freedesktop.org stuff =0) i have on and off relationship with kde ;-) | 21:42 |
| xnox | maco, good tactic =) I was picky so far about what I want to do and with who. But money is running out. | 21:43 |
| xnox | micahg, why is mozillateam not called ~ubuntu-mozilla like all the other ubuntu teams? | 21:44 |
| micahg | xnox: idk, maybe it was before there was a convention | 21:45 |
| maco | ubuntu-us-districtofcolumbia was ~dcteam originally | 21:45 |
| maco | (the docs say it should be ubuntu-us-{name of state} so we joked about ubuntu-us-werenotastateyetbutmaybesomeday and such) | 21:46 |
| xnox | maco, awesome love it =) | 21:46 |
| xnox | I'm the only Ubuntu Member from Latvia =) | 21:46 |
| xnox | There is one Debian Developer from Latvia though. | 21:47 |
| xnox | micahg, will there be a package xulrunner-2.0-dev-gtk3 ? | 21:50 |
| micahg | xnox: not likely | 21:50 |
| xnox | micahg, is it possible? | 21:50 |
| micahg | xnox: idk, why do you need it | 21:50 |
| micahg | xul20 is GTK2 only | 21:50 |
| xnox | micahg, xiphos is a gtk app that uses xulrunner. We are half way gtk3 ready | 21:51 |
| micahg | xnox: ok, but xulrunner itself isn't | 21:51 |
| xnox | and now I have realised that we might have to do webkit transition to build against gtk3. | 21:51 |
| * xnox oh the joy of doing a gtk3-webkit-gsettings release which will not be installable on pretty much any of the LTS releases. | 21:52 | |
| xnox | micahg, Bug #695166 I have subscribed chrisccoulson as well =) since it's his code. | 21:55 |
| ubottu | Launchpad bug 695166 in xulrunner-2.0 (Ubuntu) "/usr/lib/pkgconfig/libxul-embedding.pc is broken =)" [Undecided,New] https://launchpad.net/bugs/695166 | 21:55 |
| micahg | xnox: ok, if I don't fix it before next week, he'll probably take care of it | 21:55 |
| xnox | micahg, is it possible to use a recipe to test-build xulrunner? My machine cannot build xulrunner in any reasonable amount of days. | 22:00 |
| micahg | xnox: you can use a PPA | 22:00 |
| xnox | micahg, still have to generate source package =) well at least my machine should manage to do that. | 22:01 |
| micahg | xnox: yeah, that's not too bad if you're not unpacking it | 22:01 |
| === ogra is now known as Guest180 | ||
| === ResQue2 is now known as ResQue | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!