| ScottL | persia, did I really do a horrible job with zynjacku? It certainly feels so | 00:03 |
|---|---|---|
| persia | Not at all. | 00:03 |
| persia | I reject about 80% of the packages I get on REVU, and I tend to only really review the ones that either have gotten prior reviews, or for which I've previously done an eyeball review and the uplaoder has fixed stuff. | 00:04 |
| persia | I'm really picky :) | 00:04 |
| persia | Since I hit your package without that layer of prior fixing, you got a *lot* of comments. | 00:05 |
| ScottL | i felt more or less comfortable with fabrice's comments (they seemed to be more about form) but your's is addressing some stuff that I have never heard of before | 00:06 |
| ScottL | which isn't really hard really considering i've only been trying to help for a couple of months and never used linix before Ubuntu | 00:06 |
| ScottL | don't take that as I'm giving up, 'cause I'm not :) I'm still having a good time learning and helping with what I can | 00:14 |
| persia | Do you have some time in about 15 minutes? I'd be happy to go over each item individually. | 00:17 |
| ScottL | we can try, i'm getting dinner and kids are running around so I might need to take a couple of minute breaks now and then | 00:18 |
| ScottL | i know your time is valuable right now and I probably will not get to working on it again until almost the end (or even after) February | 00:20 |
| ScottL | i need to write and record another seven songs before the end of the month ;P | 00:20 |
| ScottL | so basically we can do it at your leisure but no need to rush on my account | 00:21 |
| persia | heh. I'll go through them now, but won't expect you to fix them soon :) | 00:30 |
| persia | OK. So the description is just a matter of form. Some package managers use that cloze to present "user friendly" messages. | 00:31 |
| ScottL | works for me, i'll be sure to log this also | 00:31 |
| persia | If you don't add double-spaces, package managers are free to line-wrap in ways that may not make sense. | 00:31 |
| persia | This is important for lists, and unimportant for prose | 00:31 |
| persia | GTK/GTK+ is just because upstream complains when it's not used right. | 00:32 |
| persia | manpages may seem silly, especially for GUI stuff, but people really use them. | 00:32 |
| persia | Also, manpages end up building index files, and get mirrored to web pages, which means that the application is more discoverable. | 00:33 |
| persia | creating them is usually fairly easy: ask in #ubuntu-motu for help. | 00:33 |
| ScottL | I think i'll need it for the man pages :) | 00:33 |
| persia | The icon size and number of colors stuff is to support non XDG-compliant menu systems, which some people use for very lightweight environments. | 00:34 |
| persia | fluxbox is probably the most popular environment requiring this to be right in the audio community. | 00:34 |
| persia | fluxbox also requires that the menu files be just right, which documentation is in /usr/share/doc/menu after installing the menu package. | 00:34 |
| persia | I'm not sure why usr/bin/lv2rack ended up in zynjacku, but that's probably related to the install file stuff (which we'll come back to later) | 00:35 |
| persia | The dependency on python is fairly straightforward: it's possible (but very difficult) to create an Ubuntu install without python. | 00:36 |
| persia | "#! /usr/bin/python" or non-executable is to prevent users from shooting themselves in the foot (and they have a special knack for this) | 00:37 |
| persia | "#! (interpreter)" at the top of a script just tells the shell to run that script with that interpreter. | 00:37 |
| persia | Sometimes people don't insert the space, and most shells understand, but there *should* be a space. | 00:37 |
| persia | The wiki page explains why .la files are bad more than I could. Easiest way to deal with them is not to install them. | 00:38 |
| persia | (they also aren't correct, but we don't either care or have to care if we don't ship them) | 00:38 |
| persia | We generally put shared libraries in /usr/lib/${PACKAGE}/ or directly in /usr/lib because the library loading system knows to look there for stuff. Putting it somewhere else might work, but if so, it only works by accident. | 00:39 |
| persia | So putting it in the right place is another way to avoid toe removal. | 00:39 |
| ScottL | speaking of /user is it common for prefix=/usr or prefix=/usr/local? | 00:40 |
| ScottL | s/user/usr | 00:40 |
| persia | /usr | 00:40 |
| persia | /usr/local/ is for local stuff the sysadmins install directly. | 00:40 |
| ScottL | roger | 00:40 |
| persia | The distribution isn't supposed to mess with that | 00:40 |
| persia | (and sysadmins aren't supposed to mess with /usr) | 00:40 |
| persia | So, most upstream makefiles will be set up to install to /usr/local/ by default when building from source (this is correct) | 00:41 |
| persia | But when we package it, we install to /usr instead | 00:41 |
| persia | Having the library call exit() directly just ends the program without warning the user in any way. | 00:41 |
| persia | This can be confusing :) | 00:42 |
| persia | But that's hard to fix, and changes the behaviour and requirements for code that uses the library: it deserves an upstream bug, but there's nothing you need to do for it. | 00:42 |
| persia | (well, other than file the bug). | 00:43 |
| persia | Do you understand the bug? | 00:43 |
| ScottL | nedko has been hot and cold about interacting with me but i'll file the bug | 00:43 |
| persia | nedko is an odd, brilliant man. I think he has mixed feelings about all the rest of us, perhaps in part because we don't have his depth of comprehension :) | 00:44 |
| persia | But his code is *wonderful* and well worth trying to understand. | 00:44 |
| persia | He might even have a really good reason to do it this way, although it's usually not ideal. | 00:45 |
| ScottL | well, i've started understanding the makefile ;P it's a start | 00:45 |
| persia | heh :) | 00:45 |
| persia | The long description for lv2jack just doesn't make sense to anyone who isn't already steeped in the environment. | 00:45 |
| persia | It's fine for you and I, but my mother won't install that package because she won't understand that she wants it. | 00:46 |
| ScottL | yeah, i felt bad about that because i just stole it from his website and it's was rather terse | 00:46 |
| persia | Cribbing from the zynjacku and jackrack descriptions, I suspect you can come up with something better. | 00:46 |
| persia | (because it's a mix of the two, really) | 00:47 |
| persia | changelog comment is just a form thing. The syntax (LP: #nnnn) closes the bug, so you don7t need to explicitly declare that. | 00:47 |
| persia | You can get more on optional/extra from policy, but I think my comment (which I put in about 50% of packages I review) sums it up. | 00:48 |
| persia | Uploaders: only means anything in Debian. | 00:48 |
| persia | The LGPL-2.1 bit is because you listed the Files as LGPL-2+ and LGPL. The idea of that copyright file format is to make it machine-readable, so you really need to worry about making the strings match. | 00:49 |
| persia | There's two ways to solve it: either change the License: tags for the Files: stanzas, or for the License: stanza. | 00:49 |
| ScottL | what about files that did not define copyright? | 00:50 |
| persia | Putting in the short-form of the licenses and referencing /usr/share/common-licenses is intended to save space on user harddrives (nobody needs 2000 copies of the GPL) | 00:50 |
| persia | Files that don't define copyright, or have licensing can't be distributed. | 00:50 |
| ScottL | re: /usr/share/common-license that's what I did originally but dep5 seemed to imply that the _entire_ license should be there, which seemed silly to me | 00:51 |
| persia | It's times like that you should ask in here or in -motu for help :) | 00:52 |
| ScottL | i saw some .h files and some .c files without copyright http://revu.ubuntuwire.com/report.py/legal?upid=7861 | 00:52 |
| ScottL | i shall | 00:52 |
| ScottL | ask for help, that is | 00:52 |
| persia | double-check those: at least the few I examined were just licensecheck parse failures. | 00:52 |
| persia | "public domain" isn't valid in many countries (especially ones with civil law rather than common law). A good example of how to release rights to source in all jurisdictions is http://revu.ubuntuwire.com/revu1-incoming/ccscript-1001231639/ccscript-4.1.3/cmodel.sh | 00:53 |
| persia | That's another upstream bug | 00:53 |
| persia | Also, you didn't identify the public domain stuff in debian/copyright (needs it's own section) | 00:54 |
| persia | (it's not legal for me to use it, but it is legal for Ubuntu to distribute, just because of where I live and where the Ubuntu main mirrors are) | 00:55 |
| ScottL | i went off the legal link i posted earlier and didn't see any public domain stuff :( | 00:55 |
| persia | Yeah, that's only the output of licensecheck. You're supposed to read all the source files to check :) | 00:55 |
| persia | (which is different from reading all the source: you ought be able to tell at a glance if something has a license, etc.) | 00:55 |
| ScottL | well, next time i package i'll be choosing an app with fewer files then LOL | 00:56 |
| persia | When licensecheck works, it's lovely, but when it doesn't work, we fall back to human intelligence. | 00:56 |
| persia | We're just better at understanding prose than licensecheck. | 00:56 |
| persia | I usually do something like `less $(find . -print)` and press :n lots. | 00:56 |
| persia | I keep a separate editor window open to take notes | 00:57 |
| persia | The ,desktop file bits are just based on my past experience with .desktop files. Doesn't make that much difference, but can make things nicer. | 00:58 |
| persia | Same for the 8-bit xpm, and using a nicer graphic. | 00:58 |
| persia | And dh_install is tricky. | 00:58 |
| persia | So, the way dh_isntall works is as follows: | 00:58 |
| persia | The package builds, and installs everything to debian/tmp/... | 00:59 |
| persia | Then you have several .install files. | 00:59 |
| persia | Each consists of a set of lines, in either of two formats. | 00:59 |
| persia | A single pathname with no leading / will grab a file or a directory from debian/tmp/ and copy it to debian/${PACKAGE}/ | 00:59 |
| persia | For example if debian/lv2rack.install contained the line "usr/bin/lv2rack", that file would be put in the lv2rack directory. | 01:00 |
| ScottL | because no leading / | 01:00 |
| persia | A pathname followed by whitespace followed by another path will move files, using either files in the source directory or in debian/tmp as the source of the move, and debian/${PACKAGE} as the target. | 01:01 |
| persia | Right, so .../debian/tmp/usr/bin/lv2rack -> .../debian/lv2rack/usr/bin/lv2rack | 01:01 |
| persia | For example, if debian/zynjacku had the line "debian/zynjacku.xpm usr/share/pixmaps" this would copy that file to that directory under debian/zynjacku | 01:03 |
| persia | You can also add stuff from debian/tmp if you like, so you can do a two statement line like "usr/lib/zynjacku/*.so usr/lib/zynjacku" to install the .so file (assuming it got put in debian/tmp/usr/lib/zynjacku/ in the first place) | 01:04 |
| persia | Or if it got installed to the wrong place, "usr/share/misc/zynjacku*so usr/lib/zynjacku/" would put it in the right place. | 01:05 |
| persia | When the binary packages get built, the tree of files to install is taken from debian/${PACKAGE} | 01:06 |
| persia | this is probably why you end up with extra stuff in the zynjacku package and not enough in the lv2rack package (because of the DESTDIR setting) | 01:06 |
| persia | That all make sense? | 01:07 |
| ScottL | lol, well not ALL...at least not yet | 01:08 |
| persia | Heh. Questions? | 01:08 |
| persia | I think I covered all my comments. | 01:08 |
| ScottL | a side question that perhaps you will not be able answer about when I did get zynjacku and lv2rack isntalled | 01:09 |
| ScottL | i can run either from the terminal and it will start jack if jack isn't already running | 01:09 |
| persia | OK. | 01:09 |
| ScottL | but when i picked it from the menu both would not start because jack wasn't running | 01:09 |
| persia | Right. | 01:10 |
| persia | I've never understood that issue. | 01:10 |
| ScottL | if jack was already running both would start from the menu | 01:10 |
| persia | I ended up not maintaining a package in Debian bacause of it. | 01:10 |
| ScottL | is it something I bring up with nedko? | 01:10 |
| persia | JACK can't always be running : it blocks the sound card, which many people don7t like (including my mother). | 01:10 |
| persia | No need. We tell the users to use qjackctl | 01:11 |
| ScottL | other apps have the same issue i believe | 01:11 |
| persia | As hansfbaier gets his set of stuff in shape, we can recommend that instead (if it' ready) | 01:11 |
| persia | Yeah, everything has the same issue. | 01:11 |
| ScottL | what is hansfbaier working on? | 01:12 |
| persia | I think it has something to do with the way the GUI starts stuff. | 01:12 |
| persia | jackpanel | 01:12 |
| persia | and some other similar stuff. | 01:12 |
| ScottL | oh yeah, i read you two talking about that a few days ago | 01:12 |
| persia | Really cool UI, but unfortunately needs packaging help, and he's trying to do both upstream and packaging. | 01:12 |
| ScottL | jackpanel is a gui lv2 host? | 01:13 |
| persia | http://www.hans-baier.de/wordpress/jackpanel | 01:13 |
| ScottL | huh, it's not what I thought it was then, but it lets you do connections also right? | 01:15 |
| persia | That's still under development. | 01:15 |
| persia | But connections != hosts. | 01:15 |
| persia | But it means one doesn't have to run qjackctl & patchage | 01:15 |
| persia | That can all be in the panel, and one can then keep screen real-estate open for plugins, ardour, etc. | 01:16 |
| ScottL | i (think) i just came to grips with patchage and now it's going away! kidding | 01:16 |
| persia | I don't think it's going away. | 01:17 |
| persia | But I think most people don't really need all of it. | 01:17 |
| persia | And nedko has a prettier one available anyway. | 01:17 |
| ScottL | well it's somewhat true, i just set up a template for Ardour and patchage for my most common connections | 01:17 |
| persia | Right. | 01:17 |
| ScottL | which one of nedko's are you talking about? | 01:17 |
| ScottL | the prettier one | 01:18 |
| persia | My mother mostly just opens patchage to connect her mic & instrument inputs & hydrogen to muse, and then leaves it on a different desktop. | 01:18 |
| persia | She only uses it because the qjackctl connection tool confused her. | 01:18 |
| * persia hunts for a link | 01:18 | |
| ScottL | your mother uses ubuntu studio to record music? that's fantastic! | 01:18 |
| ScottL | do you record music as well? | 01:19 |
| ScottL | i wondered about your connection to ubuntu studio - well, everyone who frequents this channel actually | 01:19 |
| persia | No. I just fool with syntheised and mangled loops :) | 01:19 |
| persia | I used to spend lots of time playing with hydrogen and freqtweak, but much less now. | 01:20 |
| persia | I got involved with Studio because I kept patching packages that we shipped. | 01:20 |
| persia | My mother uses it because I work on it, and she likes the ideas behind Ubuntu (plus, it's good enough for her) | 01:20 |
| persia | http://ladish.org/wiki/lpatchage | 01:20 |
| ScottL | yes, i've seen that before but didn't really look at it closely to be honest, nedko has done quite a bit of programming | 01:22 |
| persia | Indeed. | 01:22 |
| persia | And *lots* of it work that directly improves the set of software available to our users. | 01:22 |
| ScottL | i'm so retro in some ways, maybe not 'retro' but simplistic, i find what works and tend not to deviate | 01:22 |
| persia | heh. | 01:23 |
| ScottL | i know I can make connections in ardour but i tend to still use qjackctl -> connect and just moved over to patchage | 01:23 |
| ScottL | because I already know how to do it and i don't want to waste time dorking with something that mgiht not work when i need it to | 01:23 |
| persia | I totally understand. I intentionally avoid making connections in muse because I like separation of concerns. | 01:23 |
| * persia should really review ardour again, but the MIDI bits never seem to quite be there | 01:24 | |
| ScottL | of course this mentality is why i also stayed with hardy and wanted to backport jack and ardour to it | 01:25 |
| ScottL | it seemed like audio tended to break each update (it probably didn't but it seemed it was) | 01:25 |
| ScottL | and i finally got my system where flash (with sound) and sound in general all worked and didn't want to have to dork with it anymore | 01:26 |
| ScottL | but karmic (in general) seems much more in line so I have high hopes for further releases of ubuntu studio and i'll probably upgrade each release | 01:27 |
| * ScottL needs to get daughter to do her spelling homework | 01:27 | |
| persia | ScottL: If you have time to test lucid along the way, that's the best way to make sure the upgrade would be clean. | 01:29 |
| persia | Of course, this requires an extra computer, or extra hard drive, or some such :) | 01:29 |
| ScottL | well, i currently have four partitions but i'll be eliminating two (hardy) soon :) but yes, I plan on getting rid of the karmic one and testing lucid in its place | 01:30 |
| ScottL | but that will have to wait until after February though | 01:31 |
| persia | Fair enough: you've all those songs to write :) | 01:34 |
| ScottL | that'll be easy, stock drum pattern, two passes of rhythm guitar, one bass and a solo (over most of it) and I'll be done but I want to | 01:37 |
| ScottL | do more vocal stuff which is not as easy for me as the previously listed formula | 01:37 |
| ScottL | but this is all an optional, voluntary project just to do it | 01:41 |
| ScottL | http://www.rpmchallenge.com/ is the project | 01:42 |
| ScottL | i'm even looking forward to next years challenge and transcribing some of my acoustic stuff, putting it into muse (or something similar) and having it done on piano...or tuba | 01:43 |
| TheMuso` | Why not use lilipond? | 01:43 |
| ScottL | hey ckontros, how are you doing, you're on kinda late tonight | 01:44 |
| ckontros | TheMuso`: Can you merge this and upload a new package? https://code.launchpad.net/~psyke83/ubuntustudio-look/UbuntuStudio/+merge/18821 | 01:44 |
| ScottL | TheMuso`, i can do that, i haven't done any music other than recording live instruments (except for hydrogen) so far | 01:44 |
| ckontros | ScottL: I'm well. Just tinkering a little. | 01:44 |
| TheMuso` | ckontros: Is there an official merge request on launchpad for that? | 01:44 |
| ScottL | i have until next february to get those things (midi, etc) under my belt :) | 01:45 |
| ckontros | TheMuso: Isn't that it? :) | 01:45 |
| * TheMuso` looks | 01:45 | |
| TheMuso` | ckontros: yes it is | 01:45 |
| * TheMuso` reviews | 01:46 | |
| ckontros | TheMuso: Last couple of dailies I've grabbed wouldnt boot in VM. Are yours working fine? | 01:47 |
| TheMuso` | ckontros: Haven't tested. | 01:47 |
| ckontros | Ok. Ill pull a fresh one tomorrow and test. | 01:48 |
| TheMuso` | ckontros: Merged, and uploading new package now. | 01:51 |
| ckontros | killer | 01:51 |
| ckontros | Did stochastic or anyone figure out a new wallpaper? | 01:51 |
| TheMuso` | not afaik | 01:53 |
| ckontros | So, I see a little about what's going on because of bug reports but can anyone fill me in? Anything going on like, pulse audio getting built with JACK support? :) | 01:58 |
| ScottL | well, the jack MIR is still going on and we (meaning stochastic) was supposed to get patches together for building a few apps with jack support | 02:00 |
| ckontros | Yeah. That's most of the bug-mail Ive got. Anything else? | 02:02 |
| ScottL | TheMuso`or stochastic (or the janitor) should have more depth on that than me :P | 02:03 |
| TheMuso` | Not shipping RT kernel on discs, because its 2.6.31 and is hard to maintain post release. So going preempt kernel for amd64, no flavour for i386 because the kernel team don't want any more than 2 i386 flavours/. | 02:03 |
| ckontros | TheMuso: So we're shipping stock -generic for i386? | 02:04 |
| TheMuso` | ckontros: Yeah seems so. | 02:04 |
| TheMuso` | alessio is maintaing RT in a PPA though which will be noted in release notes. | 02:05 |
| ckontros | Wait. Why the hell do they even care since we're maintaining it? | 02:05 |
| TheMuso` | ckontros: the maintenance issue is with SRUs post release | 02:06 |
| TheMuso` | and security | 02:06 |
| ckontros | Sure but the kernel team isn't responsible for -rt. Shouldnt matter. | 02:07 |
| TheMuso` | ckontros: nothing to do with the kernel team | 02:07 |
| TheMuso` | its called lack of man power. | 02:07 |
| ckontros | And hell, SRU's wouldnt be an issue if they got around to reviewing and approving them. | 02:07 |
| TheMuso` | If I had enough time to push RT SRUs, I would, but I don't. | 02:08 |
| ckontros | Ahh. You said "because the kernel team don't want any more than 2 i386 flavours" so, that's what I was going on. | 02:08 |
| TheMuso` | right I was talking about the separate rt package | 02:09 |
| ScottL | i'll be able to upgrade my recording box now, i'll "have" to...at least that's what i'll tell my wife :) | 02:10 |
| ckontros | ? Im confused. :) | 02:10 |
| ckontros | TheMuso: Is the kernel team stopping us from having -rt in the archive and on disk? | 02:10 |
| TheMuso` | ckontros: we are not shipping the rt kernel on the discs due to lack of manpower with maintenance. | 02:10 |
| TheMuso` | ckontros: no | 02:11 |
| ckontros | Ok. Gotcha. | 02:11 |
| ckontros | I'd *almost* shutdown the project to restructure. At this stage, shipping -generic feels, dirty. | 02:12 |
| ckontros | :P | 02:12 |
| ckontros | Like it shouldnt be. | 02:13 |
| ckontros | (shipping generic) | 02:13 |
| TheMuso` | I know what you mean | 02:13 |
| ScottL | can the ubuntu studio ppa already be in the repositories for ubuntu studio? off of a clean install i mean? | 02:14 |
| TheMuso` | ScottL: not easily. | 02:15 |
| ScottL | so all users who want to upgrade to -lowlatency (or -preempt or whatever) will have to do it manually :( | 02:16 |
| TheMuso` | It will be in the release notes. | 02:17 |
| ScottL | TheMuso`, is it worth it to start a release notes wiki page? | 02:25 |
| ScottL | rather than wait until the day before release? | 02:26 |
| ckontros | Sure. | 02:29 |
| TheMuso` | ScottL: yes | 02:34 |
| TheMuso` | Jack MIR has been given thumbs up. | 03:04 |
| ScottL | yay - that's a pretty big accomplishment :) | 03:15 |
| persia | \o/ | 03:26 |
| crimsun | and I've just uploaded pulse with jack enabled | 03:47 |
| crimsun | (along with other fixes recommended by Lennart) | 03:48 |
| crimsun | I think I owe mterry a few beers | 03:49 |
| TheMuso` | I think we all do. | 03:57 |
| TheMuso` | So pulse will be dep wait till the packages are promoted then. | 03:57 |
| crimsun | indeed | 03:58 |
| ScottL | so how will pulse and jack work now? doesn't pulse automagically stop when jack is started? | 04:17 |
| ScottL | currently: pulse stop when jack is started that is | 04:17 |
| TheMuso` | ScottL: Not quite. Hopefully we can get jack2 into lucid+1 and use dbus to tigh them together. | 04:19 |
| ScottL | so, until jack2 is ready then pulse and jack will continue to act the same as they do now? | 04:21 |
| ScottL | and is jack2 going into debian first? or will we have it first? | 04:22 |
| TheMuso` | ScottL: This is what needs to be worked out. I'd like it in Debian first. | 04:23 |
| TheMuso` | ScottL: The issue is whether we want jack2 to sit alongside of jack, or replace jack. | 04:23 |
| jussi01 | wow! great work lads! | 05:47 |
| crimsun | ...and alsa-plugins uploaded with jack enabled | 07:30 |
| jussi01 | crimsun: ++ | 07:30 |
| jussi01 | nice work | 07:30 |
| * abogani waves | 07:37 | |
| jussi01 | hi abogani" | 07:55 |
| abogani | jussi01: Hi! | 07:56 |
| stochastic | Can anyone with upload powers look at the patch for building xine against jack on Bug #152487 <- TheMuso` crimsun persia etc... before feature freeze becomes solid? | 09:42 |
| ubottu | Launchpad bug 152487 in xine-lib "Jack output for xine apps " [Low,Triaged] https://launchpad.net/bugs/152487 | 09:42 |
| * abogani listen Frankie goes to Hollywood... | 09:48 | |
| stochastic | also Bug #360590 | 10:36 |
| ubottu | Launchpad bug 360590 in ubuntustudio "Please compile portaudio with Jack support" [Medium,Confirmed] https://launchpad.net/bugs/360590 | 10:36 |
| ScottL | it also looks like libffado also made it into the main :) | 12:03 |
| crimsun | yep, due to it being one of j-a-c-k's build-deps | 12:23 |
| * crimsun fixes alsa-plugins | 12:23 | |
| ScottL_ | i added to the working release notes wiki page https://wiki.ubuntu.com/UbuntuStudio/WorkingReleaseNotes | 18:23 |
| ScottL_ | i started one for lucid and was able to add a lot more than I thought I would :) | 18:23 |
| persia | ScottL_: realtime won't be default for amd64 : preempt will be default. Realtime will be available for in the PPA for both i386 and mad64 | 18:28 |
| persia | Also, I'm not sure we applied to be part of the Long Term Release. | 18:29 |
| persia | So there may be no guarantee that our applications will get long-term support. | 18:29 |
| persia | (we could try ourselves, but I'm not confident we have enough developers) | 18:29 |
| ScottL_ | i will change the wiki then :) thanks for the checking and feedback | 18:29 |
| ScottL_ | quick fix - i'll reread back over it later (i had made the original during lunch but I'm back at work now) | 18:35 |
| ScottL_ | if there is anything that should be added, i'll be happy to make any additions, moderations or deletions if you just want to say something here about it (e.g. I know TheMuso` and persia are probably really, really busy right now) | 18:39 |
| * persia is kinda recovering from the pre-FF marathon, so not particularly busy, but similarly not particularly inclined to do much right now | 18:39 | |
| ScottL_ | persia: just to clarify -realtime vs -preempt, in synaptic the package says -rt but is that just semantics? it really is preempt kernel? | 19:22 |
| persia | No. | 19:23 |
| persia | The -rt kernel is different, but not up-to-date (I think) | 19:24 |
| persia | Oh, it is kinda up-to-date, but not perfectly. | 19:24 |
| persia | But that's not the -preempt kernel. | 19:24 |
| stochastic | hey fabrice_sp and persia is there any chance one of you could look at Bug #360590 and Bug #152487 ? | 19:28 |
| ubottu | Launchpad bug 360590 in ubuntustudio "Please compile portaudio with Jack support" [Medium,Confirmed] https://launchpad.net/bugs/360590 | 19:28 |
| ubottu | Launchpad bug 152487 in xine-lib "Jack output for xine apps " [Low,Triaged] https://launchpad.net/bugs/152487 | 19:28 |
| stochastic | patches are ready to be uploaded. Do we need FFE bugs as well now? | 19:28 |
| persia | stochastic: That requires main upload rights, which neither of us have. | 19:29 |
| stochastic | persia, where would be the best place to poke someone about this? | 19:30 |
| ScottL_ | dtchen might have those rights, no? | 19:30 |
| persia | Have you subscribed ubuntu-main-sponsors to the bugs? | 19:30 |
| stochastic | yes | 19:30 |
| persia | he does, as do others. | 19:30 |
| persia | But subscribing the sponsors is the best right now. | 19:30 |
| persia | Otherwise, wait, and either of the core-devs who are regularly in this channel will probably see it in backscroll. | 19:31 |
| stochastic | okay, I'll sit and wait then | 19:31 |
| persia | I think we all know we want it, so it's just a matter of time. | 19:32 |
| persia | If freeze exceptions are required, they ought breeze through. | 19:32 |
| TheMuso` | Building portaudio with jack support may need to be considered carefully. DOing so will put libjack onto the CD which may be a problem. | 21:06 |
| TheMuso` | Xine is not a problem however. | 21:06 |
| persia | Doesn't kubuntu pull xine as a phonon backend still? | 21:08 |
| ScottL_ | why would libjack on the cd be a problem? | 21:08 |
| TheMuso` | I don't know. | 21:08 |
| TheMuso` | ScottL_: Space for one. | 21:08 |
| ScottL_ | i'm guessing that libjack isn't necessarily that big, but that the cd is rather full already | 21:09 |
| persia | 187k, which adds up fast over time. | 21:09 |
| persia | Right. | 21:10 |
| persia | TheMuso`: libjack0 is already in Task: ubuntu-desktop | 21:10 |
| persia | It may be that it's already on the CD | 21:10 |
| TheMuso` | persia: hrm ok. Probably won't know till the next daily images. | 21:11 |
| persia | Gets pulled by libasound2-plugins at least | 21:11 |
| TheMuso` | Point. | 21:12 |
| TheMuso` | Although I was hoping that would be a separate package. | 21:12 |
| persia | Maybe later, but not as currently implemented. | 21:12 |
| TheMuso` | No need for jack anything to be on the CD by default, most people don't need it. | 21:12 |
| persia | For jackd & friends, I agree. WIth the way shared libraries work, the core lib might end up there. | 21:13 |
| TheMuso` | right | 21:15 |
| TheMuso` | In which case portaudio linking against jack won't be bad. | 21:15 |
| persia | Well, it's 187k, but we're probably going to end up with that there anyway. | 21:16 |
| TheMuso` | yeah | 21:17 |
| * TheMuso` checks latest daily images. | 21:17 | |
| TheMuso` | ok not there yet | 21:18 |
| persia | No, tomorrow. | 21:18 |
| persia | Or rather, later today :) | 21:18 |
| persia | Or ~11 hours from now, however one counts. | 21:18 |
| TheMuso` | yeah | 21:18 |
| TheMuso` | In which case I'll tend to those two bugs shortly. | 21:19 |
| ScottL_ | did anyone ever tell you guys that you rock? they should. you do! :) | 21:35 |
| TheMuso` | Thanks | 21:37 |
| TheMuso` | Right, portaudio debdiff uploaded, thanks stochastic. | 22:27 |
| TheMuso` | Yay can't build xine atm due to package installability breakage. :) | 22:46 |
| persia | heh | 22:49 |
| * TheMuso` refreshes his local mirror in case the problem is transient and fixed in the archive. | 22:54 | |
| TheMuso` | Otherwise, time to go digging. | 22:54 |
| TheMuso` | Yay ocaml breaks installability. :S | 23:08 |
| TheMuso` | Ah ok graphviz is affected by the ocaml transition. | 23:18 |
| * TheMuso` sighs. I really don't enjoy these wild goose chaces sometimes. | 23:48 | |
| TheMuso` | Oh yay. A launchpad/soyuz bug (i think) regarding not handling provides properly for package metadata. | 23:55 |
| persia | How so? | 23:56 |
| persia | Because virtual provides don't work by design. | 23:56 |
| TheMuso` | libgraphviz-dev in debian depends on ocaml-base-nox-3.11.2 but due to provides, ocaml-base-nox ets installed. | 23:56 |
| TheMuso` | ocaml-base-nox provides ocaml-base-nox-3.11.2 | 23:56 |
| persia | Which breaks the transition? | 23:56 |
| TheMuso` | which break libgraphviz-dev installability | 23:56 |
| TheMuso` | s/break/breaks/ | 23:57 |
| TheMuso` | so we need to carry another change ingraphviz to depend on a versioned ocaml-base-nox | 23:57 |
| TheMuso` | instead of a ocaml-base-nox-$bla dependency | 23:57 |
| TheMuso` | I think thats the best approach anyway. | 23:58 |
| TheMuso` | and all to rebuild xine... | 23:58 |
| persia | Oh, it's a versioned depends on ocaml-base-nox-3.11.2 ? | 23:59 |
| persia | Yeah, that's annoying. | 23:59 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!