[00:07] <nxvl> ScottK: http://tservice.net.ru/~s0mbre/blog/devel/networking/dns/2008_08_08.html
[00:08] <nxvl> ScottK: we are still fucked
[00:19] <wgrant> nxvl: Grmph.
[00:19] <nxvl> wgrant: exactly
[00:19] <nxvl> my paranoia reached an unthinked level
[00:20] <wgrant> Yes.
[00:20] <nxvl> i have just changes all my sources.list from URL's to IP's
[00:20] <nxvl> wgrant: btw, can you accept me on motu-swat please
[00:21] <wgrant> nxvl: Sure.
[00:22] <wgrant> Done.
[00:22] <emgent> uhm?
[00:23] <crimsun> it's still "broken"
[00:23] <nxvl> wgrant: thank you!
[00:23] <crimsun> as Dan mentioned in his talk, updates to packages have to be signed by a trusted party specifically for that package in a monotonically increasing version.
[00:24] <crimsun> and of course, distributed "securely" - whatever the heck that means.
[00:24] <sistpoty> well, the insecure nature would be obvious by a failed signature, wouldn't it?
[00:25] <wgrant> There are more avenues for exploitation than apt.
[00:25] <wgrant> If there's no way to fix BIND...
[00:25] <sistpoty> of course
[00:25] <emgent> nice.
[00:26] <emgent> very nice news. gh
[00:26] <sistpoty> just assume that the site of my bank would be diverted to another ip... I'd be doomed (well, i.e. I wouldn't not too much to get from my account *g*)(
[00:26] <wgrant> sistpoty: SSL saves there, as long as the CA's DNS isn't poisoned.
[00:27] <sistpoty> heh
[00:29] <soren> I really don't see what all the fuss is about. You've never been able to trust the response you got from DNS before anyway.
[00:30] <nxvl> soren: the problem is that now you have a LOT of script kiddies pretending to be crackers and trying to crack some dns just to say they have
[00:30] <soren> From my POV, DNS responses have gone from "untrustworthy" to "very untrustworthy".
[00:32] <soren> If you've been doing anything up until now that depended on trustworthiness of DNS replies, you've been doing it wrong :)
[00:33] <wgrant> Now anybody can do it reliably and easily.
[00:33] <soren> yes.
[00:35] <soren> The problem with the whole "oh, noes! We can't trust DNS anymore!" business is that it's nonsense. You couldn't trust it before either.
[00:35] <nxvl> soren: the actual worrying thing is that ppa don't use signed packages
[00:36] <nxvl> soren: so if you get poisoned, you'r fck'd
[00:36] <soren> nxvl: How would that help anything?
[00:36] <soren> Ah, well..
[00:36] <soren> Yes, it would help somewhat.
[00:36] <nxvl> soren: that if it's signed you can check the source of the package
[00:36] <sistpoty> nxvl: not really, I gues using ppa's in the first place is wrong ;)
[00:36] <soren> ...but any idiot or script kiddie can set up a ppa anyway.
[00:36] <nxvl> soren: that's why i don't use ppa's i don't trust
[00:37] <nxvl> just for testing
[00:37] <wgrant> soren: But I'm not going to add some scriptkiddie's PPA to my sources.list.
[00:38] <soren> wgrant: Right.
[00:38] <wgrant> Now no PPAs are trustworthy, even if they are owned by sane people.
[00:38] <soren> True, true.
[00:38] <wgrant> (they never really were before either, but particularly now)
[00:38] <sistpoty> hm... I wonder if revu should use ssl, and if so how to set it up right *g*
[00:38] <nxvl> even MY ppa is trustworthy
[00:39] <wgrant> Well, primary isn't much more trustworthy than a PPA right now, but that's another story.
[00:39] <soren> wgrant: How so?
[00:39] <sistpoty> heh
[00:39] <wgrant> Not sure it should be said.
[00:40] <soren> Of course it should.
[00:44]  * soren heads to bed
[00:44] <wgrant> Night soren.
[00:44] <soren> 'night, all.
[00:44] <nxvl> soren: night
[00:44] <soren> *whoosh
[01:02] <Brucevdk> Alright, I don't think anybody caught my question 5 hours ago. But I managed to create a super duper Debian package (wrote a Python distutils setup script and just let debhelper.mk python-distutils.mk handle the rest). \o/ Happy days.
[01:03] <Brucevdk> Now I can sleep. Zzz
[01:05] <crimsun> sistpoty: don't use a self-signed cert.
[01:05] <sistpoty> crimsun: heh... that's where my knowledge ends *g*
[01:06] <crimsun> sor.en's right, there's a lot of hype over something that has existed for decades.
[01:06] <sistpoty> anyway... got to go to bed now... gn8 everyone
[01:06] <wgrant> Night sistpoty.
[02:09] <effie_jayx> https://bugs.edge.launchpad.net/ubuntu/+source/cfv/+bug/256500
[02:09] <effie_jayx> does it make sense to change the suggests line then?
[02:09] <effie_jayx> or is it not enough reason to do it
[02:15] <_sebastian1> Andrew_Sayers: hi
[02:19] <_sebastian1> Hi
[02:23] <vorian> sebner: hello, connection problems?
[02:24] <vorian> bah, _sebastian1 even
[02:24] <vorian> sebner: sorry about that
[02:33] <emgent> vorian: hello
[02:33] <vorian> hiya emgent :)
[02:39] <_sebastian1> hi all
[02:41] <vorian> _sebastian1: hello again :)
[05:09] <infinitycircuit> how would i get a list of packages that have Build-Essential: yes?
[05:10] <StevenK> Build-Essential: yes isn't a tag.
[05:10] <StevenK> infinitycircuit: apt-cache show build-essential ?
[05:11] <infinitycircuit> in apt-cache dumpavail it does list it
[05:11] <nxvl> infinitycircuit: build-essential?
[05:11]  * StevenK blinks
[05:12] <StevenK> apt-cache dumpavail | grep-dctrl -sPackage -FBuild-Essential 'yes'
[05:12] <StevenK> That will do what you want
[05:12] <infinitycircuit> ah thank you
[05:54] <emgent> moin
[06:35] <andrew_sayers> _sebastian1: Hi.
[06:55] <_sebastian1> andrew_sayers: hi
[06:55] <andrew_sayers> I'm hoping you're the Sebastian I've been talking to for the past little while :)
[06:56] <_sebastian1> are you the andrew I wanted to meet :-D
[06:56] <andrew_sayers> :)
[08:26] <Iulian> G'morning
[12:18] <Whoopie> persia: Hi, time to discuss bug report 211252? what is the right way to apply patches? with CDBS?
[12:19] <persia> bug #211252
[12:20] <slytherin> Whoopie: cdbs has it's own patch system called cdbs-simple-patchsys. There is some documentation on wiki.ubuntu.com
[12:21] <slytherin> Whoopie: and obex-data-server is in main, so discuss on #ubuntu-devel first.
[12:21] <Whoopie> slytherin: yes, thanks. should I use the example from udev or cdbs?
[12:21] <persia> Whoopie: Yes.  Your debdiff was perfectly formatted, except for three things: 1) the fix needs to get into intrepid before it gets into hardy, 2) there's no value in also closing bug #227429, and 3) Full Names are appreciated.
[12:22] <sebner> persia: I sent a mail to him :)
[12:23] <Whoopie> slytherin: it's just to know which of the example is the prefered. plain patch or cdbs.
[12:23] <persia> slytherin: I'll admit to not looking at MoveToUniverse yet.  It's about third on my list, and ought get hit in the next bit (unless I get more strange results with the top two).
[12:23] <persia> sebner: Excellent to hear.
[12:23] <ion_> Hei hei, Jukka.
[12:23] <sebner> persia: we'll see what he answers ^^
[12:23] <slytherin> persia: no issues, I just upgraded to intrepid and will start looking at some of them.
[12:23] <ion_> Eh. Sorry.
[12:23] <persia> Whoopie: debdiff is generally preferred.  Whether this uses a patch system or not depends on the package being changed.
[12:24] <ion_> http://revu.ubuntuwire.com/details.py?package=compcache-setup anyone? :-)
[12:24] <slytherin> Whoopie: if you mean which patch system to use? the simplest is cdbs-simple-patchsys provided packages uses cdbs to build. The next best thing is dpatch.
[12:25] <sebner> or quilt :)
[12:25] <Whoopie> slytherin: ok
[12:25] <Whoopie> too much choices ;)
[12:25] <slytherin> sebner: we are talking about simple, not powerful. :-)
[12:25] <persia> ion_: As usual, I'm not going to give it a full review, but isn't -plow specifically designed to not block, and still use the debconf template?
[12:26] <sebner> Whoopie: I'm off the opinion that quilt is simpler than dpatch :P
[12:26] <persia> slytherin: Except the best thing is *always* to use the existing patch system.
[12:26] <sebner> argh
[12:26] <sebner> Whoopie: sry
[12:26] <sebner> slytherin: see above ^^
[12:26] <slytherin> persia: of course
[12:26] <slytherin> sebner: not for me.
[12:26] <Whoopie> for obex-data-server, there's none, so I'd need to choose.
[12:27] <ion_> persia: I want the message to printed for anyone reading messages printed by the scripts, but i don’t want it ever to block (no matter the debconf priority the user has).
[12:27] <persia> Whoopie: What does what-patch say?
[12:27] <sebner> slytherin: well, everyone has to decide on his own
[12:28] <Whoopie> persia: thanks!!! didn't know that command. it's cdbs
[12:29] <persia> Whoopie: Yep.  I've just also checked the the identity of the Debian maintainer, and it's the same as for bluez-utils, so it is likely they prefer the same patch system (when selecting with no hints, this can be a useful method to find the patch system most likely to be accepted).
[12:31] <slytherin> persia: now that openjdk is in main, we should prefer it over gcj for building packages, right?
[12:31] <ion_> persia: It’s basically the equivalent of the config script containing ‘echo "Restart compcache-setup to apply the change"’, but might as well use a debconf template for localizing that since the package uses po-debconf in any case.
[12:32] <Whoopie> persia: I attached the upstream patch for the nokia browsing issue to bug #211252. I think, Kenny's used patch is not complete.
[12:33] <persia> ion_: Right, but by using -plow you don't need the lintian override, and most users ought never see it anyway.
[12:35] <persia> slytherin: Best to follow https://lists.ubuntu.com/archives/ubuntu-devel-announce/2008-July/000460.html
[12:36] <Whoopie> persia: and another question regarding bluez-utils: why not syncing with debian to get latest version 3.36 instead of patching?
[12:38] <persia> Whoopie: Well, we've a bunch of patches.  That said, perhaps a merge is overdue :)  I'll take a look.
[12:39]  * slytherin wonders what is need of jorbis and jogg libraries in an ftp program.
[12:39] <persia> On the other hand, I still think your patch is correct for hardy-proposed (aside from the small notes above), as we wouldn't want that.
[12:39] <persia> s/that/a full merge as an SRU/
[12:42] <Whoopie> persia: are we in freeze? didn't follow the roadmap.
[12:43] <persia> Whoopie: We're in Debian Import Freeze, which means we need a good reason for an update
[12:45] <persia> http://packages.qa.debian.org/b/bluez-utils/news/20080720T163203Z.html has the changes for the latest upload to Debian: someone needs to investigate to verify that 1) the changes are an improvement and 2) it doesn't break anything.
[12:50] <persia> Hmm.  New upstream might be nice, but I'm not sure if any of the Debian-specific patches apply.
[12:50] <persia> slytherin, mario_limonciello: do either of you have an opinion on this?
[12:52] <slytherin> persia: it has been long since I stopped doing anything related to bluetooth. But as far as I know, the 4.x version of bluez stack is kind of rewrite.
[12:52] <persia> slytherin: This is 3.30 vs. 3.36.  Also, I see your name from only April :p
[12:54] <slytherin> persia: yes, that was the upload I did. I believe tollef is the right person to bug about this. In any case 3.36 is last release of 3.x series so whatever fixes we want to backport we need to do it before 4.x lands in Ubuntu.
[12:56] <persia> slytherin: I very much doubt 4.x will drop in Ubuntu until after intrepid.
[12:57] <slytherin> persia: why? I think it should be in intrepid.
[12:58] <persia> slytherin: Well, it would need someone to specifically try to get everything upgraded.  Why do you think we should use 4.x?
[12:58] <persia> Upstream says it's still experimental
[12:59] <tuxmaniac> persia: does that mean "sync requests" also need to undergo a certain procedeure? (wrt Debian import freeze) I thought it doesnt have to until FF. May be I am missing something here?
[12:59] <slytherin> persia: from what I have read in upstream news it is better implementation of all the services, specifically network and file transfer.
[12:59] <persia> slytherin: Hmm, but "experimental"...
[12:59] <slytherin> tuxmaniac: DIF means there is no auto sync. FF means sync needs exception.
[13:00] <slytherin> let ubuntu-mobile team decide, I guess they are the one who handle bluetooh
[13:00] <persia> tuxmaniac: sync requests need review now: before DIF, they only needed review if there was ubuntu variation: otherwise they happened automatically.
[13:01] <slytherin> persia: if there was variation it wouldn't be called sync, it is merge :-P
[13:02] <persia> slytherin: Not always: sometimes we sync over Ubuntu variation (when there is no longer a point in keeping it)
[13:02] <slytherin> ahh, right
[13:03] <persia> slytherin: A common example of this is when people backport bugfixes from Debian into Ubuntu during Ubuntu freezes.
[13:03] <slytherin> hmm
[13:09] <RainCT> heya
[13:09] <coolbhavi> heya RainCT
[13:14] <RainCT> persia: I may start the wiki page for new MOTUs which you requested yesterday now. Do you have any wishes on what should be there?
[13:16] <persia> RainCT: -v, -k, testing in Debian and forwarding patches, "give-back", +queue at least.
[13:17] <persia> (probably better as "Uploading a Merge", "Sponsoring someone else's work", "Pushing upstream", "Handling FTBFS Reports", "Checking package status".
[13:17] <persia> )
[13:19] <persia> RainCT: Do you remember anything else that is important *as* MOTU, but not typically needed/learned before becoming MOTU?
[13:19] <persia> (or anyone else: please chip in)
[13:24] <RainCT> persia: Right now, no. About the "Pushing upstream" point, what should it say (or rather: shouldn't you already be doing this before you become a MOTU)?
[13:25] <persia> RainCT: Hmm.  Maybe just something about looking at Debian bugs when sponsoring stuff?
[13:25] <persia> Essentially, I think that those with more experience are good about Debian collaboration, but I don't like to block new people just because they aren't yet, and I think most MOTU ought do that.
[13:26] <persia> As a result, we have a certain obligation to help with Debian collaboration for those seeking sponsorship until they are doing it on their own.
[13:26] <persia> (Mind you, if you strongly disagree with that, there's no point adding it to the page)
[13:31] <slytherin> does anyone know why seahorse doesn't act as passphrase agent in intrepid?
[13:31] <RainCT> So, something like "Check if the problem has been reported in Debian's BTS and link it to the LP bug, perhaps leaving a comment explaining the contributor how to do so. If it hasn't been reported there, but the problem is also in Debian, ask the contributor to forward the fix.", or you mean even a "Consider forwarding the fix yourself."?
[13:32] <persia> I even mean consider forwarding the fix oneself, linking it to the LP bug, and explaining to the contributor that they ought do that next time.
[13:33] <persia> I'm all for patch authors forwarding stuff, but believe there is a learning process that ought be supported.
[13:33] <RainCT> alright
[13:53] <slytherin> persia: should I log a catch all bug for MoveToUniverse or separate bugs for each package?
[13:54] <highvoltage> howdy motus
[13:54]  * wgrant wonders if anybody is looking at why libhibernate3-java doesn't build. I've been getting an awful lot of build failures because it is missing.
[13:56] <slytherin> wgrant: go ahead and bang your head on wall. :-P I did that for 2 weeks almost. There are hell lot of circular build dependencies in hibernate in jboss packages.
[13:57] <wgrant> I like to stay away from Java.
[13:57] <wgrant> I just happened to notice lots of packages failing while rebuilding multiverse today.
[13:58] <slytherin> wgrant: any report about these failures?
[13:58] <slytherin> I mean other than the usual FTBFS page
[13:59] <wgrant> http://builder.ubuntuwire.com:9998/dist/intrepid/arch/i386 has all so far, but it's a bit flaky at the moment due to the load produced by the rebuilding.
[13:59] <persia> slytherin: Seperate bug for each package.
[13:59] <slytherin> wgrant: Ok, I will check in an hour or two
[13:59] <slytherin> persia: ok
[14:13] <white> http://channel.debian.de/map.html
[14:16] <persia> Hmm.  Some of that looks like England.
[14:16] <persia> (But the rest looks rather nifty indeed)
[14:37] <persia> Is there a utility that will parse a .dsc file, and copy all referenced files (and the .dsc file) to a target directory?
[14:37] <laga> something like dget?
[14:37] <white> Any new MOTU/ubuntu enthusiast in melbourne?
[14:38] <ion_> persia: dcmd might work.
[14:38] <persia> laga: Sortof.  That takes a URL to a remote .dsc and moves it to the current directory, but I want something local.
[14:39] <tuxmaniac> Can someone please review bug 255224 and sponsor if found OK. Thanks in Advance. earlier posted it in #u-bugs by mistake
[14:39] <ion_> dcmd mv foo.changes bar/
[14:39] <RainCT> ion_: dcmd? is that new?
[14:39] <ion_> rainct: Dunno, just found it with apropos.
[14:40] <persia> ion_: I don't seem to have a dcmd, do you mean http://sourceforge.net/projects/dcmd/ ?
[14:40] <persia> ion_: Which package?
[14:40] <ion_> devscripts >= 2.10.20 apparently.
[14:40] <slytherin> persia: one down, jftp. :-)
[14:40] <wgrant> white: I think I'm about the only Ubuntu dev in Melbourne :(
[14:41] <persia> ion_: Nifty.  Thanks.
[14:43] <white> wgrant: i was just wondering about that email on debian-melb from Udo
[14:44] <white> wgrant: although it's still a few weeks, we might want to fix a date :)
[14:45] <wgrant> Indeed.
[14:45]  * wgrant -> bed
[14:52] <RainCT> https://wiki.ubuntu.com/MOTU/New - suggestions are welcome (if you can thing of a better name for the page, too)
[14:52] <RainCT> persia: ^
[14:53]  * persia grumbles at http://ftp.debian.org/debian/pool/main/d/dbus-java/ not containing an orig.tar.gz
[14:55] <slytherin> persia: that is because version number is wrong.
[14:56] <slytherin> persia: it got converted to native package
[14:57] <persia> slytherin: Did it now.  There's also no native package .gz there.
[14:58] <persia> By the way, the 2.5-3 .dsc references the same checksum for an orig.tar.gz as was used in the 2.5-1 upload.
[14:59] <persia> So the file is present at http://ftp.debian.org/debian/pool/contrib/d/dbus-java/ : it's just DAK being strange
[14:59] <slytherin> he he
[15:00] <persia> Still, it meant that dget didn't work, which is why I grumbled
[15:33] <Adri2000> any ubuntu studio people here?
[15:38] <persia> Adri2000: Why?
[15:41] <Adri2000> persia: I'm confused about the latest comment in bug #243722. ubuntu studio uses the ubuntu repositories just like ubuntu, doesn't it?
[15:43] <persia> Adri2000: Absolutely.  There are some differences (different kernel, some special settings, etc.), but it's really just a flavour.
[15:43] <RainCT> Adri2000: Is the new version already in -updates?
[15:44] <persia> Erm.  That is confusing.  Maybe a mirror that doesn't include -updates?
[15:44]  * persia investigates
[15:44] <Adri2000> it's only in -proposed for now
[15:44] <RainCT> Adri2000: To me it seems like that guy is just confused and doesn't know about -proposed
[15:44] <Adri2000> but the same person reported earlier that the fix worked in hardy, so I guess he know that
[15:44] <Adri2000> knows
[15:45] <Adri2000> or maybe he has -proposed enabled on his hardy machine without knowing it or something... I'll ask
[15:47] <persia> That would make sense: perhaps he's using -proposed on a -desktop install, but not on a -studio install.
[15:48] <Adri2000> probably
[15:50] <persia> Adri2000: I've just double-checked and confirmed that the ubuntustudio specific mirrors are CD mirrors, but that the sources.list ought be identical to any other flavour.
[15:51] <persia> s/CD/DVD/
[15:54] <Adri2000> persia: ok
[16:00] <RainCT> persia: have you placed links to MOTU/New somewhere or should I do so?
[16:00] <sebner> RainCT: cool page btw :)
[16:00] <RainCT> sebner: thanks :)
[16:00] <persia> RainCT: I was planning to just point new MOTU at the page on induction.
[16:02] <persia> RainCT: Also, thanks a lot for putting that together:  I think it's stuff we all mostly learn in the first week or two, but sometimes things get missed, so good to have organised for reference.
[16:03] <directhex> where would i file a request to have a package synced from debian that isn't in ubuntu?
[16:04] <RainCT> directhex: just against Ubuntu
[16:05] <directhex> seems it should be fine in universe for now, but i may not have considered all the possibilities
[16:05] <tuxmaniac> directhex: please rise a bug report in LP
[16:06] <tuxmaniac> directhex: for. e.g. like bug 253324
[16:07] <directhex> can someone with an intrepid build environment run a quick build test against http://ftp.de.debian.org/debian/pool/main/w/webkit-sharp/webkit-sharp_0.2-1.dsc ?
[16:07] <directhex> i only have a hardy environment on this machine
[16:10] <sebner> directhex: sure
[16:12] <tuxmaniac> directhex: it doesnt seem to build. Maybe I am doing something wrong. http://paste.ubuntu.com/36216/
[16:13] <tuxmaniac> thats output of pdebuild
[16:13] <slytherin> tuxmaniac: what did you do eaxctly?
[16:13] <sebner> directhex: working
[16:13] <tuxmaniac> pdebuild --debbuildopts -sa ?
[16:14] <directhex> tuxmaniac, intrepid?
[16:15] <tuxmaniac> directhex: yes.
[16:15] <sebner> tuxmaniac: did you do  a debuild in the source folder or just built the dsc with pbuilder?
[16:15] <tuxmaniac> sebner: debuild in the src folder
[16:15] <sebner> directhex: working with pbuilder building the dsc
[16:15] <sebner> tuxmaniac: no need to
[16:15] <slytherin> tuxmaniac: you simply need to build with pbuiler using .dsc
[16:16] <tuxmaniac> sebner: shouldnt both work?
[16:16] <directhex> tuxmaniac, when did you last update your pbuilder environment? cli.make is in cli-common-dev 0.5.7, from may 11th
[16:17] <tuxmaniac> directhex: today evening.
[16:17] <directhex> and possibly older too
[16:17] <sebner> tuxmaniac: sure but with debuild it's building on your machine and not in the clean chroot
[16:17] <tuxmaniac> sebner: i use pdebuild. which in turn uses the pbuilder environment right?
[16:18] <sebner> tuxmaniac: well, if debuild succeds ^^
[16:19] <geser> tuxmaniac: have you debhelper 7 installed?
[16:19] <slytherin> is javacio.us accessible to anyone? Can someone please search which jar file contains netscape.javascript.JSObject?
[16:20] <tuxmaniac> geser: no
[16:23] <directhex> meebey's a fan of dh7, it'll be a frequent build-depends of things i talk about
[16:24] <directhex> okay, sync request at 256641
[16:24] <sebner> directhex: I also once wanted to sync webkit-sharp. dunno why I didn't ^^
[16:25] <directhex> sebner, well, i need it in order to sync mono-tools
[16:25] <directhex> s/sync/merge
[16:25] <directhex> or maybe sync. most of the ubuntu1 changes are gecko-related
[16:26] <geser> slytherin: if my google foo is reliable than its in plugin.jar (> Java 1.4.2) or jaws.jar (Java 1.4.0/1.4.1) [http://www.rgagnon.com/javadetails/java-0172.html]
[16:27] <slytherin> geser: that is what I found from some forum. I just wanted to see if it was available anywhere else. Even rhino jar doesn't have it. :-(
[16:42] <slytherin> I have a question. I am trying to build worldwind with openjdk. But an applet include doesn't build. Is it ok to exclude the applet from build so that package can be built with openjdk and moved to universe.
[16:49]  * slytherin out for dinner
[16:51] <effie_jayx> persia,  justr an update on fonts I have them all done. all I have to do is package them
[16:57] <effie_jayx> if a new package closes several bugs how does one reflect that in the changelog
[16:58] <geser> effie_jayx: LP: #123, #456
[16:58] <effie_jayx> thanks
[16:58] <effie_jayx> all in parenthesis no?
[16:58] <geser> effie_jayx: like you want, parenthesis are not required for automatic bug closing
[16:59] <geser> you can also use several LP: #xxx entries, depending on your changelog entries
[17:11] <effie_jayx> I am building a fonts package.. there are changes in a binary file as well as in the source file
[17:11] <effie_jayx> I try to build the source package and it fails
[17:11] <effie_jayx> Irrepresentable Changes in the source
[17:11] <effie_jayx> I am using debuild -S -sa to build
[17:27] <geser> effie_jayx: you need to figure out how you can get the changes to the binary file(s) into plain text (and apply them during package build)
[17:29] <effie_jayx> geser,  mmkey, perhaps e fontdiff of some sort
[17:29] <jpds> effie_jayx:
[17:29] <jpds> effie_jayx: Err, uuencode and uudecode, can convert binary files to text and vice verse.
[17:29] <jpds> -a*
[17:30] <jpds> They're in the sharutils package.
[17:31] <slytherin> geser: Can you please comment on what I said above about worldwind?
[17:33] <geser> slytherin: the idea sounds sane as long as the applet is only an add-on to the package
[17:33] <effie_jayx> jpds,  ok, do I first dencode them into text files and then generate a diff from the text files ?
[17:33] <effie_jayx> then specify changes into the building process?
[17:34] <slytherin> geser: yes, it is not even part of core jar file, AFAIK
[17:35] <jpds> effie_jayx: Yes, see 'man uuencode'
[17:36] <jpds> effie_jayx: uuencode .ttf file. And on the prebuild stage for the package building "uudecode" the encoded file.
[17:45] <effie_jayx> jpds,  I am bit confused... I just ran uuencode with the ttf file.. Now I run uudecode with debuild?
[17:45] <jpds> effie_jayx: No, you place the encoded file somewhere and run uudecode in debian/rules.
[17:47] <effie_jayx> jpds,  right... the encoding seems stuck
[17:47] <effie_jayx> http://pastebin.ubuntu.com/36237/
[17:50] <jpds> effie_jayx: Try: uuencode Ubuntu-Title.ttf Ubuntu-Title.ttf
[17:54] <effie_jayx> jpds, get output
[17:55] <jpds> effie_jayx: To save the output - simply redirect it to a desired filename: "> Ubuntu-Title.ttf.uu"
[17:55] <emgent> heya
[17:55] <effie_jayx> jpds,  it takes a while aparently
[17:56] <jpds> Hello emgent
[17:56] <emgent> jpds: :)
[17:56] <slytherin> geser: need you advice. Can you please take a look at libjogl-java? Some part of code is under SGI Free Software License B which is supposed to be non-free. But the code is licensed with some exceptions which I guess should make it free.
[17:57] <jpds> effie_jayx: Depends on the filesize I guess.
[18:06] <RainCT> effie_jayx: https://wiki.ubuntu.com/PackagingGuide/Howtos/BinaryFilesInDiff
[18:14] <slytherin> is there any url where a list of Free and non-Free licenses according to DFSG is available?
[18:18] <geser> slytherin: can you point me to the exceptions? for libjogl-java
[18:19] <JasonSpradlin82> anybody here to answer newbie questions about packaging?
[18:19] <slytherin> geser: If you have source then you can find them in the LICENSE.txt file in jogl directory.
[18:21] <JasonSpradlin82> I was following the Packaging Guide and after getting pbuilder, it showed me an example of how to create a build environment
[18:21] <JasonSpradlin82> i ran the command, and created the build environment
[18:21] <geser> slytherin: you mean the paragraphs starting with "The JOGL source tree contains code
[18:21] <JasonSpradlin82> but what was the point of doing that, aside from creating a build environment that isn't affected by my computer settings, that can work as any hardy computer
[18:22] <JasonSpradlin82> i have this build environment there on my computer now, but what would a person normally do next?
[18:22] <slytherin> geser: yes
[18:26] <geser> slytherin: if I understand this correctly then libjogl-java still contains code licensed under SGI FreeB
[18:27] <slytherin> geser: re-licensed by Sun (that is what I believe)
[18:28] <jpds> JasonSpradlin82: Build packages with it.
[18:29] <geser> slytherin: I don't know, I'm not a licensing expert
[18:29] <JasonSpradlin82> ok... like how?
[18:29] <slytherin> geser: whom should we consult?
[18:30] <JasonSpradlin82> jpds: (forgive my ignorance, i am trying to fix that)
[18:30] <JasonSpradlin82> ((my ignorance))
[18:30] <jpds> JasonSpradlin82: Did you setup pbuilder?
[18:30] <geser> slytherin: ask an archive admin or the debian-legal ML
[18:30] <JasonSpradlin82> no, i just installed it using apt-get... the package guide simply said I would need it installed, if I wanted to start contributing
[18:31] <slytherin> geser: I think debian-legal ML would be right place.
[18:32] <jpds> JasonSpradlin82: Well, pbuilder builds packages in a chroot environment, it installs the necessary programs, libraries, etc - which a package needs in order to build.
[18:32] <jpds> JasonSpradlin82: This way, you don't have to install everything on your working system.
[18:33] <JasonSpradlin82> jpds: The problem with the packaging guide, methinks, is that it tells you how to do things, but assumes you already know why you are doing them... i've been using computers my whole life, was writing programs on my commodore 64 when i was 10, but i've never, ever worked on linux, and I really have no clue as to the way things are done... and the package guide seems to assume I have more common knowledge on these issues
[18:33] <JasonSpradlin82> jpds: yeah i kind of understand that... so can you give an example of when and how i might use it, while contributing to ubuntu?
[18:34] <jpds> JasonSpradlin82: To test the build process of your own packages?
[18:35] <JasonSpradlin82> jpds: yeah... i mean like... "you would take a package (or source code), and with programX, you would run it and it would be built into that build environment" or "it would be built according to the settings of that build environment" (i'm guessing here)
[18:37] <jpds> JasonSpradlin82: "man pbuilder" - explains the program a bit more.
[18:38] <JasonSpradlin82> jpds: would i be using pbuilder later to package a program, and then it go into that build environment to see what files are already on the system and see what files (dependencies) it would need to obtain, on a bare-bones ubuntu hardy install?
[18:38] <JasonSpradlin82> ok thanks
[18:40] <JasonSpradlin82> jpds: this is much more informative... real quick: is chroot as mentioned in this manual, the same as the base.tgz/build environment?
[18:40] <jpds> JasonSpradlin82: The base.tgz is uncompress to create the chroot.
[18:40] <JasonSpradlin82> ok cool...
[18:41] <JasonSpradlin82> and that would be a kind of virtual file system with base settings and any other changes i've affected to it using pbuilder, correct?
[18:41] <jpds> JasonSpradlin82: Have you seen the videos? https://wiki.ubuntu.com/MOTU/Videos
[18:41] <JasonSpradlin82> i was going to watch them earlier, but I don't have my sound set up yet
[18:41] <JasonSpradlin82> ((sheepish grin))
[18:43] <jpds> JasonSpradlin82: One has to get to grips with the debian/ directory: https://wiki.ubuntu.com/PackagingGuide/PackagingOverview - beforing starting with building the package ;-)
[18:45] <JasonSpradlin82> jpds: ah good... also, what is the upstream, actually?  The latest version undergoing debugging?
[18:46] <jpds> JasonSpradlin82: Upstream are the actual program coders/writers.
[18:46] <RainCT> JasonSpradlin82: upstream is from where we take the application (it can refer to the authors, or in some cases to Debian)
[18:47] <RainCT> JasonSpradlin82: The reason why pbuilder (or another system like it) should be used, beside avoiding your system to get full of unnecessary stuff, is that this way you ensure that the package declares al necessary dependencies, and that pbuilder allows you to have chroots for different Ubuntu/Debian versions (ie., you can build packages for Intrepid with it being on Hardy)
[18:51] <JasonSpradlin82> AHA!  "the original (sometimes called "upstream") tarball"
[18:52] <JasonSpradlin82> so when we update a debian package for ubuntu, are we only adding patches that will self install when unpackaged for installation, and leaving the original debian source untouched?
[18:52] <jpds> JasonSpradlin82: Untouched, but with patches if necessary.
[18:53] <JasonSpradlin82> awesome (that i get that)
[19:01] <JasonSpradlin82> ok wow... it should have been stated somewhere to me that the "basic overview guide" is much more helpful to the ignorant that the complete guide is....
[19:01] <JasonSpradlin82> it was extremely informative and cleared a lot of things up for me.
[19:01] <JasonSpradlin82> thanks
[19:01] <jpds> Hey, we're all here to help each other \o/
[19:03] <tuxmaniac> LucidFox: thanks for electric :-)
[19:07] <bdrung_> does someone know how to escape " in an $( ... ) command (in debian/rules)?
[19:08] <bdrung_> PLATFORM := $(python -c "from distutils.util import get_platform; print get_platform()") does not work
[19:08] <geser> use \"
[19:12] <bdrung_> geser: does not work
[19:13] <geser> bdrung_: try adding more \, e.g. \\"
[19:15] <bdrung_> geser: does not work
[19:17] <geser> hmm, then I've no idea
[19:18] <slytherin> bdrung_: why do you need to get platform that way?
[19:19] <bdrung_> slayton:  because it is used in a patch
[19:20] <geser> bdrung_: what error did you got?
[19:20] <slytherin> bdrung_: I mean isn't there any other way to retrive platform?
[19:20] <bdrung_> is maybe $(...) the wrong command
[19:23] <bdrung_> afk
[19:52] <slytherin> slomo: gst-plugins-bad FTBFS on intrepid. Are you already working on that?
[19:56] <bdrung_> geser, slytherin: i have to use $(shell ... ). in my case: PLATFORM := $(shell python -c "from distutils.util import get_platform; print get_platform()")
[19:57] <slytherin> bdrung_: yes, I completely overlooked shell part
[19:59] <bdrung_> slytherin: i use this way to get the platform name, because the directory name is generated using the same function. the debian maintainer used dpkg-architecture for getting the platform name, but this causes a ftbfs on i386
[19:59] <slytherin> hmm
[20:07] <slytherin> Can anyone please ack bug 254919 It will fix FTBFS for gst-plugins-bad on intrepid.
[20:24] <tuxmaniac> Can someone please ack bug 255224 after revieing the changes. its new upstream release
[21:04]  * RainCT hugs jpds because he is working on ubuntu-dev-tools
[21:04]  * jpds hugs RainCT.
[21:04] <RainCT> so guys, now you know what you have to do if you want a hug ;)
[21:16] <foxbuntu> would someone be able to revu my package, it would be greatly appreciated: http://revu.ubuntuwire.com/details.py?package=mythbuntu-log-grabber
[21:39] <nxvl> emgent: ping
[21:55] <Kopfgeldjaeger> foxbuntu: hi. you probably should have a look at https://wiki.ubuntu.com/DebianMaintainerField - and you can add yourself as XSBC-Original-Maintainer
[21:57] <Kopfgeldjaeger> and you could probably merge both changelog entries into one - if you want both names there sth. like [ Thomas Mashos and Nick Fox ] should be ok
[21:57] <foxbuntu> Kopfgeldjaeger, what do you mean? I was working on the package the with original author and just submitted it to revu becuase I completed it and had it ready to go
[21:58] <foxbuntu> Kopfgeldjaeger, I can do that
[22:22] <tgm4883_laptop> Is there someone that can clarify a packaging rule for me?  Should get-orig-source grab the source from the original authors site, or from a bzr branch that I placed the authors source on?
[22:26] <RainCT> tgm4883_laptop: usually from the author's site, but if you prefer to get it from a branch that's OK too
[22:26] <RainCT> you can pretty much let get-orig-source do whatever you want :)
[22:27] <tgm4883_laptop> yea, thats what I was afraid of
[23:50] <tgm4883_laptop> what would be the consequences for not having a get-orig-source?
[23:51] <tgm4883_laptop> RainCT, regarding that question you answered from me earlier.  Is it also OK to change that branch that i'm getting the source from if i'm not the orig author?
[23:52] <tgm4883_laptop> i'm asking for clarification, because the packaging guide seems to indicate otherwise
[23:52] <RAOF> tgm4883_laptop: Someone will likely complain at you a little, and then move on.
[23:53] <tgm4883_laptop> and i'm finding trouble getting the source from the projects homepage when I am trying to package this.  It keeps warning me about  dpkg-source: warning: ignoring deletion of file search.pl
[23:53] <RAOF> That's not a problem, as long as the resulting thing builds of course.
[23:53] <tgm4883_laptop> even though AFAIK, i'm not deleting anything.  Also, this error only occurs if I have the .orig.tar.gz in the dir above
[23:54] <RAOF> Presumably search.pl is being deleted on clean.
[23:54] <tgm4883_laptop> RAOF, ok, but I think that is what Is making all the files end up in the .diff
[23:54] <RAOF> No, it isn't.
[23:54] <tgm4883_laptop> which is what sistpoty said I needed to fix
[23:54] <tgm4883_laptop> h
[23:54] <tgm4883_laptop> oh
[23:54] <tgm4883_laptop> hmm
[23:54] <RAOF> All the files end up in the .diff?
[23:54] <tgm4883_laptop> yea
[23:54] <tgm4883_laptop> http://revu.ubuntuwire.com/details.py?package=mythstream-parser-youtube
[23:55] <tgm4883_laptop> in case you want to look ;)
[23:55] <RAOF> That probably means that your clean target doesn't work.
[23:55] <tgm4883_laptop> which is in my rules file?
[23:55] <RAOF> Yes.
[23:56] <tgm4883_laptop> yea, my rules file only contains 2 things
[23:56] <RAOF> Then it's quite possibly wrong :)
[23:56] <tgm4883_laptop> include /usr/share/cdbs/1/rules/debhelper.mk and get-orig-source
[23:57]  * tgm4883_laptop curses at the packaging guide
[23:57] <tgm4883_laptop> so would that be some dh_clean type thing?
[23:57] <tgm4883_laptop> just point me in the right direction and i'll go a learning
[23:57] <tgm4883_laptop> ;)
[23:58] <RAOF> Hm.  So you're using cdbs, which will be calling all the right things.
[23:58] <RAOF> Which would point to... how does the source actually get _built_?
[23:59] <tgm4883_laptop> built?  All I want to do is stick this parser into /usr/share/mythtv/mythstream/parsers/"name"
[23:59] <tgm4883_laptop> when you say built..
[23:59] <RAOF> Ah.  So, what you want the package to do is to copy foobar.script to /usr/share/mythtv/mythstream/parsers/foobar?
[23:59] <tgm4883_laptop> yes