=== medberry is now known as med_out === kentb is now known as kentb-out [06:59] good morning [08:41] hello, I uploaded a package to a ppa and it will be built in approx. 19 hours. I have a local copy of binary package though .. is there any way to make "apt-get install" use the local binary so I can test whether it works? [08:41] s/19/10/ [08:42] the package in question has dependecies and I'd like to see whether it install cleanly [08:42] al-maisan, sudo dpkg -i bla.deb? [08:42] * StevenK peers at the buildfarm [08:42] or just double clicking it? [08:42] dpkg -i won't install cleanly if the dependencies aren't installed [08:42] dholbach: I understand "dpkg -i bla.deb" will not take care of the dependencies [08:42] "sudo apt-get install -f" afterwards will [08:43] :) [08:43] al-maisan: Put the built package in a directory, run dpkg-scanpackages . /dev/null | tee Packages | gzip -c > Packages.gz [08:43] Then add deb file:/// . to your sources.list [08:43] It's a little nasty [08:44] aha .. that's interesting .. will try that [08:44] StevenK: thanks! [08:45] wgrant: Perhaps we want to gift i386 with an amd64 builder or two? 96 jobs versus 402 [08:47] The UK has woken up, so we may have a better option. [08:54] StevenK: All my usual volunteers are gone this time. [08:54] StevenK: So I don't know if any of the current amd64 virt builders are safe to put on i386. [08:54] (that sudo umask business) [08:55] But thallium is back now. === Quintasan_ is now known as Quintasan [10:25] bdrung: re your mail about the lintian build failure: could it be because of the umask change in oneiric (from 022 to 002) and that lintian needs to set the umask to 022 during the tests (as the tests expect that umask)? [10:26] geser: maybe. how can i set the umask in a makefile for all commands? [10:27] I thought niels said that the scripts set the umask themselves === al-maisan is now known as almaisan-away [10:29] bdrung: can you not just define it as a SHELL env? [10:29] bdrung: good question, don't know the answer yet [10:30] Laney: i see some umask(022) commands === almaisan-away is now known as al-maisan [11:30] hmm .. I got this failure in the packaged code: [11:30] failed to load external entity "/usr/lib/pymodules/python2.7/openquake/../docs/schema/nrml.xsd [11:30] sorry [11:30] ECHAN [11:30] please ignore the message above [13:40] hello [13:40] wheni do the command [13:40] dput reve myfile.changes [13:41] ita says that it can't find gpg signature [13:41] and that i need a .sig or .asc file [13:41] how can i do? [13:41] is the changes signed? [13:46] persia: howdy ho! do you still not read emails? [13:46] no [13:46] i made it with gambas 2 [13:47] highvoltage: I'm starting to get better, but not to the degree that I can be said to be a reliable user of email. [13:47] help [13:48] debsign changes-file [13:48] Raffa50: What command did you use to build the package? [13:48] debuild will do that for you [13:48] do oyu have a gpg key? [13:48] yes i have [13:49] and alll the fules [13:49] gambas 2 builded my deb [13:49] never used gambas, does debsign work? [13:51] hm its an ide? I'm not sure a ide created package would suitable for the archive [13:52] So, REVU doesn'T accept .deb files: it only accepts DEbian-format source packages. [13:52] (.dsc and the files referenced in the .dsc). [13:52] If gambas has a means to create a .deb file, that file is certain to be unsuitable for review by REVU. [13:52] i'm installig debsign [13:52] One needs to take the source one edited in gambas, and then package that (probably using a text editor, which gambas may provide) in the typical way, and ask for review of the source package. [13:53] Just signing the .deb won't make REVU accept it, although it may allow dput to upload it (but it will be silently ignored). [13:54] can't sign the .changes? [13:54] whats the error message? [13:55] Won't matter. REVU only accepts source packages. A .deb an the associated .changes is completely uninteresting to REVU. [13:55] if a changes is created probably also a dsc [13:56] ? [13:57] debsign [13:57] jtaylor: Nope. One can generate a .changes file with a .deb without any .dsc: one just has to construct the .deb using a different mechanism. I wouldn't be surprised if gambas2 included such a mechanism. [13:57] says that gpg key isn't aviable [13:57] ah k, didn't know that [13:57] gambas generated [13:57] .changes and .dsc [13:58] Raffa50: You need a key such that one of the identities precisely matches the Changed-By: string in the .changes file to sign it, unless you force it. Don't force it. [13:58] you may need to pass -k but its moot if you don't have a real source package [13:58] No. One should *never* pass -k unless one is sponsoring. [13:58] If one thinks one has to pass that, one is trying to hide a bug in the packaging. [13:58] no you need itt if you have more than one key [13:58] i shopuld do this? [13:59] debsign sly_0.0.4-1_i386.changes -k BB3E91C0 [14:00] sign file .dsc [14:00] jtaylor: If you have more than one key that asserts the same identity, debsign *should* pick one, based on your keyring order. I suppose that in the special case where you have multiple keys for a single identity, and want to use the non-primary key for that identity, you might use it, but it's probably better to set DEBSIGN_KEYID in your environment in that rare case. [14:00] gpg key isn't aviable [14:00] that's a changes file for a binary (+source) upload, you need a _source.changes file [14:00] Raffa50: Do you have more than one key with the same identity, and want to use your non-primary key for some reason? [14:00] no [14:01] anly one gpg key [14:01] If not, you probably want to fix your changes file such that the Changed-By: header matches an identity on your key. [14:01] ? [14:01] And, better, is to change whatever gambas is using to set Changed-By: to the right value. [14:01] i poast my .changes [14:02] Format: 1.8 Date: Tue, 26 Jul 2011 12:00:00 +0000 Source: sly Binary: sly Architecture: source all Version: 0.0.4-1 Distribution: unstable Urgency: low Maintainer: raffa50 Changed-By: raffa50 Description: sly - Whit Sly you can easy program without "spoil" you hans Changes: sly (0.0.4-1) unstable; urgency=low . * fixed glvars replaced with glvar. generating errors on if i [14:02] maybe i use pastie [14:03] http://pastebin.com/uR5YFEj7 [14:04] does your gpg key match raffa50 ? [14:05] no [14:05] Raffaele Aldrigo [14:05] then you need to fix gambas to use the proper id [14:05] ah [14:05] i try [14:05] but I recommend scraping gambas for packaging and doing it directly in the source [14:05] using regular text editor of your choice [14:05] the source is basic [14:06] how can i do? [14:06] That's OK. You'd use gambas to build the source in your packaging (probably in the debian/rules file) [14:06] i can't manually insert gpg key [14:06] in .changes? [14:06] with text editor? [14:06] sorry i'm italian [14:07] and only 18 years old [14:08] what generates the changes file should be correct, else you would have to repeat the editing every time [14:08] try setting DEBEMAIL and DEBFULLNAME maybe gambas is then smart enough [14:08] environment variables [14:10] Raffa50: No need to apologise: we all needed to learn, and those that learned have included a number of Italians, and a number of folk 18 years old (and both much younger and much older). [14:11] IT's just your turn :) [14:13] no don't work [14:13] i cahnged my name in gambas [14:13] but debsign can't find the key [14:13] Raffa50: force it with -kKEYID [14:13] before or after [14:13] paultag: No. See above :) [14:13] .changes? [14:13] Raffa50: where KEYID is the GPG short fingerprint [14:14] persia: ah, righto. sorry :) Showes me for just reading new messages :) [14:14] debsign ??? [14:14] Raffa50: Did you verify gambas has all of your name, email, and GPG uid comment? [14:14] debsign sly_0.0.5-1_i386.changes -k BB3E91C0 [14:15] no gpg uid conmment [14:15] debsign sly_0.0.5-1_i386.changes -k BB3E91C0 [14:15] debsign sly_0.0.5-1_i386.changes -k BB3E91C0 [14:15] correct? [14:15] If it does, it's worth filing a bug on gambas: we don't usually use that to create packages. [14:15] don't work [14:16] But if it can't construct correct .changes files, that just forces the use of things we consider to be poor habits. [14:16] -k before filename, and no space between -k and the keyid. [14:16] But, really, there's not much point, if your goal is to submit to REVU. [14:16] I promise that REVU won't accept a .deb, even with signed .changes. [14:17] ah [14:17] what happens when you run debuild in your source tree? [14:17] gambas copiled all [14:17] source is basic [14:17] jtaylor: IT's an IDE build: it does it in some special way. [14:17] it's in another dir [14:18] debsign -k BB3E91C0 sly_0.0.5-1_i386.changes [14:18] persia: yes but maybe its a ide that creates a correct makefile and debian dir :) [14:18] For all that I've critiqued quickly over time, I do believe it's the only IDE in the archive that generates (mostly) correct source packages by default. [14:18] don't work [14:18] wrong command [14:18] jtaylor: Maybe. I'd like to be wrong. [14:18] buhuhu [14:19] i only want to see my app on ubuntu software center [14:19] poor me [14:19] if that is your goal you must have a build system that does not rely on an IDE [14:20] no way [14:20] when i compile with gambas [14:20] s/must/should/ [14:20] i see trange commands [14:20] like [14:20] dpkbuild [14:22] aha [14:22] gambas says [14:22] gpg key isn't aviable [14:24] dpkg-deb: generazione del pacchetto "sly" in "../sly_0.0.6-1_all.deb". dh_testdir signfile sly_0.0.6-1.dsc gpg: skipped "Raffaele Aldrigo ": la chiave segreta non è disponibile gpg: [stdin]: clearsign failed: la chiave segreta non è disponibile dpkg-genchanges >../sly_0.0.6-1_i386.changes dpkg-buildpackage: binary and diff upload (original source NOT included) dpkg-genchanges: not including original sour [14:25] Raffa50: The email address matches the email on your key? [14:27] yes [14:27] Reloader90@gmail.com [14:27] but the nam ein the changes is different [14:27] i found [14:40] i exported my key [14:40] i have a .dsc file [14:40] .asc [14:40] scs [14:42] help [14:42] don't work [14:42] exporting your key does not help# [14:43] Whether the key is expoered or not doesn't change anything. [14:43] did you try editing the changes and see if that works? [14:43] gpg is going to use the key in your keyring, rather than any exported copy. [14:44] sed -i -e s/raffa50/Raffaele Aldrigo/ sly_0.0.6-1_i386.changes [14:44] ? [14:44] sed -i -e "s/raffa50/Raffaele Aldrigo/" sly_0.0.6-1_i386.changes [14:44] the files are on the desktop [14:44] ok i try [14:45] don't say nothing [14:45] and now? [14:46] try signing it again [14:46] with debsign [14:46] debsign [14:47] gpg: skipped Raffaele Aldrigo [14:47] gpg key not aviable [14:48] did the editing of the changes file work? or is there still raffa50 in it? [14:51] what happens when you do gpg --sign changes-file? [14:52] wait [14:52] it works [14:52] it ask me the passphrase [14:52] wait [14:52] gpg not aviable [14:53] in this session [14:53] sucesifful signed [14:53] works [14:53] uploading [14:53] thank you [14:54] uploaded [14:54] thank [14:54] love you [14:55] now is my software on the ubuntu center? [15:01] no, I assume you uploaded to revu, that is only a review platform for prospective packages [15:02] to get it into the archive it must be of good quality [15:03] using gambas to build the package is probably already bad, please read http://www.debian.org/doc/manuals/maint-guide/ for information on creating good pacakges [15:03] https://wiki.ubuntu.com/PackagingGuide === med_out is now known as medberry === andreas__ is now known as ahasenack === al-maisan is now known as almaisan-away === bambee_ is now known as bambee === andreas__ is now known as ahasenack === tuos is now known as tuomasjjrasanen [19:28] i uploaded my file [19:28] with dput [19:28] but i can't see it in ubuntu software cente [19:28] why? [19:29] I answered that already [19:33] no [19:36] Raffa50: see the last 4 lines from jtaylor on http://irclogs.ubuntu.com/2011/07/26/%23ubuntu-motu.html [19:36] ah I see now in my scroll-back that you left before jtaylor answered you [19:37] ah, I have leaves filtered ._. [19:38] sigh [19:38] jtaylor: me too but with a smart filter in weechat which displays the ones for people who wrote something a few minutes before they leave [19:42] that would be a good feature for the client I'm using [19:44] my app is an innovation [19:44] you can easy program [19:45] without touch code [19:45] it's like an uml [19:46] if its a good program you might find someone to package it for you [19:46] file a needs-packaging bug in ubuntu or an request for packaging in debian [19:48] Preferably the latter. [19:56] nothing? [19:56] so wath i did with dput revu ... [19:58] did you read https://wiki.ubuntu.com/MOTU/Packages/REVU? [19:59] can explain it in easy world [19:59] +i'm italian [20:01] i understand [20:01] i should wait [20:02] a package built solely with gambas is unlikely to be added to the archive [20:03] packages need manual labor to get them in good shape for distribution, thats why almost every package has a dedicated maintainer in debian [20:03] but i can make it only in basic [20:03] what should i do? [20:05] in that case either learn to do it properly or find someone willing to do the work for you [20:05] does your software have a website? [20:09] mmm, dh_python2 [20:20] * micahg hugs dtchen for going through open bugs w/patches [20:21] yep, been on an ~ubuntu-reviewers tear lately [20:32] * dtchen frowns at submittodebian: UnicodeEncodeError: 'ascii' codec can't encode character u'\xe9' in position 558: ordinal not in range(128) [20:32] oh well, another on my TODO === yofel_ is now known as yofel === Quintasan_ is now known as Quintasan === and is now known as and` === medberry is now known as med_out