[00:06] Pharaoh_Atem: sigh, http://trac.roundcube.net/ticket/1490544 [00:06] not packaged yet either, will need to figure that out tmrw [00:08] nacc: pictor, your patch is against a stale version; there's a 2.35-0ubuntu1 package that you're in the changelog of [00:08] slangasek: ha! ok ... let me check that, sorry [00:10] nacc: likewise, simplesamlphp is against a stale version [00:11] nacc: and I would dearly love for further debdiffs to be in a bug per package instead of in one bug report with 800 tasks... at some point LP is going to flip the table on us [00:12] slangasek: ok, i can do that from now on ... sorry [00:12] slangasek: so looks like roundcube 1.2 (current at beta) is teh only version that supports php7 ... similar issue as drupal7, afaict [00:14] kirkland: for the pictor upload (2.35-0ubuntu1) why did you keep php5 compat in the control file? there is no php5-imagick, e.g., in xenial [00:14] kirkland: are you ok if we drop that compat? [00:18] nacc: my view is that if it's in universe and it doesn't support php7, it's expendable and can be cut from release. I am not, however, the product manager for Ubuntu Server. From an archive/release POV, I'm perfectly willing to take a roundcube beta into the release instead, if a) someone is willing to do the work to prepare that package and make it work, b) upstream isn't going to yell at us for [00:18] including their beta, c) any requests for an update ... [00:18] ... from the beta to the final release post-16.04 follow the SRU process [00:22] slangasek: yep, understood. I think there are a class of applications that simply aren't going to be updated; but roundcube is making progress it seems like. I'll take a look at the packaging requirements tmrw [00:23] slangasek: updated simplesamlphp debdiff [00:23] slangasek: pictor may be fine as-is, depends on what kirkland wants to do [00:24] nacc: the | php5 stuff shouldn't break the package, if there were no changes to the actual contents [00:27] slangasek: it shows up in http://people.canonical.com/~ubuntu-archive/nbs.html [00:27] and i was aksed to fix that [00:27] nacc: well, when we remove php5 from the archive, it will not show up there anymore ;) [00:28] won't it? the pictor package will still depend on php | php5, php-imagick | php5-imagick ? [00:28] nacc: this is a limitation of nbs, it doesn't know how to unpick alternative dependencies [00:28] slangasek: ah ok, i wasn't sure, yeah, it's purely an alternative for that one [00:29] the others are real and fixed in today's set of patches [00:29] if that's the only remaining revdep of php5-imagick, and pictor depends php-imagick | php5-imagick, we can pull it already [00:29] yep, afaict that's it and how it is [00:30] nacc: php5-imagick removed from xenial === salem_ is now known as _salem [00:44] doko: python3.5 - why bump libmpdec_version instead of removing this check from the testsuite? why is python hard-coding a requirement of a matching minor version of libmpdec? [00:51] slangasek: thanks! [01:36] nacc: should simplesamlphp's bin/memcacheSync.php have been updated? [02:57] hi [02:57] how does something from debian packaging get into ubuntu [02:57] for example if a debian package is new [02:58] how would it get into 14.04? [03:02] karstensrage: You wouldn't, 14.04 is a stable release. [03:02] karstensrage: But we can get it into 16.04 [03:03] (And then you could perhaps request a backport) [03:04] how do i do that [03:04] has 16.04 synced to debian yet? [03:06] karstensrage: I don't know what package you're asking about... [03:10] can i pm? [03:11] Public conversations work fine. [03:11] Because of your asking, I requested a sync. :3 [03:12] * Unit193 waits for the day infinity kills him. :D [03:17] infinity, https://packages.debian.org/testing/libs/libufpidentity1 [03:17] there is another one coming too, but its stuck in mentors [03:18] hi Unit193 [03:19] karstensrage: Synced to 16.04 [03:19] it is? [03:19] karstensrage: When the mentors one clears Debian NEW, let me know and I'll grab that too. [03:19] alright? can i see that? [03:19] and how do i ask for a backport for 14.04? [03:20] thank you so much [03:20] https://launchpad.net/ubuntu/+source/identity4c/1.0-1 [03:20] karstensrage: To request backports, you need to talk to the backports folks. [03:21] karstensrage: https://wiki.ubuntu.com/UbuntuBackports [03:22] ok thank you [03:29] infinity, how long do i have for clearing Debian New [03:29] karstensrage: Well, we release 16.04 in under a month. So, that long. [03:29] hmmm [03:30] ill do my best [03:42] oh infinity did that pull in the devel package too? [03:42] dev package? [03:44] oh yes [03:44] i see thank you [03:56] whats a good candidate for the reasoning part? [03:56] just want new functionality on 14.04? [03:56] and possibly 12.04 since there are still machines out there that havent updated? [05:25] hey [05:25] Has OTA 10 been released yet by Canonical? [06:14] nacc: so, the simplesamlphp diff switches a dep to php-mhash, which doesn't exist? [06:14] nacc: but neither does php5-mhash [06:17] slangasek: mhash is builtin to all SAPIs (for both 5 and 7.0) [06:17] So that's a useless dep. [06:19] ok, so it was probably a legacy provides [09:13] hi folks, do you have any news about glibc/binutils/powerpc breakage? [09:14] ginggs, I found your chat, but I didn't see any answer :( === lynxman_ is now known as lynxman [10:03] LocutusOfBorg: What glibc/binutils/powerpc breakage? [10:04] infinity, nobody can install fpc on powerpc [10:04] simple as that [10:04] And this is glibc's fault somehow? [10:04] ginggs, did some debugging [10:04] the postinst script has a command that triggers a SEGFAULT [10:05] and downgrading glibc fixes the issue [10:05] let me find the bits, just a second [10:05] [17:20:25] doko, infinity, LocutusOfBorg: I've confirmed the [10:05] problem with installing fpc 3 on powerpc is related to glibc 2.23. [10:05] After downgrading libc6*, gcc* and locales, fp-compiler installs just [10:05] fine. Any ideas? [10:06] "I get the same access violation just running 'fpcmkcfg-3.0.0'. [10:06] And even just asking for help segfaults." [10:06] Hrm. No ideas, but I can look into it later. [10:07] you can just call that "fpcmkcfg-3.0.0 -h" command [10:07] Right. I'll give it a stab later. It's almost certainly fpc's fault, not glibc's, but I'll need to dig and sort out why to prove that theory. :P [10:07] the problem is: fpc doesn't build on powerpc anymore because it depends on itself [10:07] zigo: python-googleapi has a few incompatibilities with the python-oauth2client now in xenial. [10:07] (And it needs fixing either way) [10:08] https://launchpad.net/ubuntu/+source/fpc [10:08] Unit193: Oh... :/ [10:08] LocutusOfBorg: Yeah, I know, I have the build log open. [10:08] thanks! [10:08] LocutusOfBorg: Like I said, I'll look later. [10:08] zigo: New upstream fixes it at least.. [10:08] LocutusOfBorg: I'd like to understand why it broke before I just go and rebootstrap it. [10:08] I'm happy to fix fpc sure, but I think I won't be able to do it alone, right? [10:08] Cause if it's using internal symbols or something, it should stop. [10:08] thanks a lot, funny enough it was working with the older glibc :) [10:08] thanks for giving a shot! [10:09] infinity: thanks. FWIW with the new glibc, I was able to install fpc 2.6.4 from wily but building 3.0.0 failed with the same access violation at 99% :( [10:09] Unit193: I'm not the maintainer of it, so maybe you could ping Laszlo Boszormenyi, he's usually very responsive. [10:09] (and maintains a lot of stuff there...) [10:09] that fpc has important fPIC fixes on arm* I need them, because they are a big regression compared with 2.6.4, and fixes hedgewars [10:09] Unfortunately and more importantly to me, broke gcalcli too. ;/ [10:10] LocutusOfBorg: Ahh, well, wouldn't want hedgewars to be broken. ;) [10:10] yeah, nothing is more imporant in life than games :) [10:10] LocutusOfBorg: I'm about to take a long weekend, but I might poke fpc when bored. [10:10] I vaguely recall the last time I rummaged around in fpc's guts. [10:10] thanks! and happy Easter, I'll be taking a rest too :) [10:11] s/rest/helpwifeincookingalotofstuffforeaster/ [12:30] doko: no they aren't stuck anymore, thanks [12:42] Anyone know where I can find langpack generation logs? [12:42] or another way that I can get a list of packages that have their translations put in there [12:50] aha I found mapping.txt [14:05] infinity: since beta2 is out no concerns from me in unblocking that mysql-5.7 out of proposed [14:05] i can sort out fallout in the myth* stuff now [14:20] superm1: Nice. We also need to fix some test failures with dbconfig-common for it (and redmine, but I'm guessing those can also be traced to dbconfig) [14:30] doko: julia has been FTBFS on armhf for several weeks, I've just tried downgrading to gcc-5 5.3.1-3ubuntu1 (where I know I definitely built before) and it builds fine. What can I do further? [14:31] infinity, slangasek ping [14:32] Program received signal SIGPIPE, Broken pipe. [14:32] 0x00007f82e0535417 in __libc_writev (fd=50, vector= [14:32] at ../sysdeps/unix/sysv/linux/writev.c:49 [14:32] the x crash [14:33] compiz didn't crash it crashes later because of x [14:33] Skuggen: http://paste.ubuntu.com/15496370/ [14:33] Skuggen: that's with the stuff in proposed, not sure where that header is supposed to be coming from [14:34] Hm. Could you check what's in include/mysql/mysql? There's some mess with the header layout the devs haven't gotten around to clean up yet (though I don't think it should have been moved) [14:34] sorry :) [14:35] wrong signal [14:35] that's earlier :p [14:35] Skuggen: looks to be in that directory [14:35] http://paste.ubuntu.com/15496401/ [14:37] superm1: But it used to work in 5.6? From what I can see in the source the file is in the same place for 5.6.28 [14:37] Skuggen: never caused build failures previously [14:37] so maybe the path declared in /usr/include/mysql/mysql/client_plugin.h changed? [14:42] Skuggen: from what i see in 5.5 (on 14.04 box) client_plugin.h wasn't included in /usr/include/mysql/mysql.h, so that's the part that's newer [14:42] don't have a 5.6 handy ATM [14:43] Ah, no it's not included in 5.6 [14:44] https://github.com/ltangvald/mysql-5.6/blob/master/include/mysql.h [14:46] If you add include/mysql on the include path, does it work? [14:47] yes that works [14:48] the part that broke for me was this is how a configure script was testing things [14:51] Yeah, it's kind of a mess the way the includes are structured right now :| [15:10] slangasek: ah sorry, thanks for catching that! [15:14] slangasek: we can probably do a best effort here, if you're ok with an addition fix for simplesamlphp to suggest `phpenmod -s cli memcache`, iiuc [15:14] Pharaoh_Atem: --^ is that rigth? [15:27] Pharaoh_Atem: and independently, how well do you know horde? :) [15:28] Pharaoh_Atem: nm on the latter [16:17] slangasek: i think we should ignore pbgenomicconsensus amd64. retries don't help but local adt-runs in xenial-proposed chroots always pass for me [16:17] slangasek: then there's python-pbh5tools; pretty much same deal [16:30] slangasek: re: LP: 1562060, i tested that the roundcube 1.2~beta installs and seems to run (I don't have an IMAP server setup so can't test it further) [16:30] Launchpad bug 1562060 in roundcube (Ubuntu) "Please update to roundcube 1.2" [Undecided,New] https://launchpad.net/bugs/1562060 [16:49] infinity, how come it looks everything has been published but when i search for the package here https://launchpad.net/ubuntu/xenial it doesnt show up? [16:50] infinity, also do i have to woo some sponsor for a backport bug or are they just picked up eventually? [17:21] karstensrage: hrm? https://launchpad.net/ubuntu/xenial/+search?text=libufpidentity1 [17:23] slangasek: fyi, I think with LP: #1562090 just filed, the horde stuff stuck in proposed should pass its tests [17:23] Launchpad bug 1562090 in php-horde-passwd (Ubuntu) "Update to PHP7.0 dependencies" [Undecided,New] https://launchpad.net/bugs/1562090 [17:23] i'll keep working through the rest of horde [17:34] nacc: on the former, that looks right to me [17:34] Pharaoh_Atem: ok, will send a patch, thanks [17:35] on the latter, I have little clue about Horde :/ [18:54] cjwatson: fyi (I don't need anything right now), bug #1562118 [18:54] bug 1562118 in debmirror (Ubuntu) ".temp/dists/.../Packages goes out of date" [Undecided,New] https://launchpad.net/bugs/1562118 [20:15] hikiko: did you get a full backtrace under gdb of the crash? 'bt full'. also are you sure it was the SIGPIPE that killed the process, and that X didn't handle that SIGPIPE later? [20:19] barry: pbgenomicconsensus, you probably can't reproduce the failure because you're testing with different versions of packages than what the autopkgtest infrastructure does. I'm guessing you're doing a run with -proposed versions of packages, rather than using python-pysam 0.7.7-1ubuntu1 like the log shows? So, it's possible to skip this test failure for python-setuptools since we know python-set [20:19] uptools isn't the cause, but please follow up on ... [20:19] ... python-pysam because this indicates real bugginess currently in xenial [20:34] slangasek: looks like moodle is similar to drupal and others ... working on it now, had to fix the watch file too :/ [21:07] barry: there, pbgenomicconsensus no longer a failed test ;) [21:07] nacc: hehe, enjoy... [21:08] nacc: LP: #1562090 uploaded, which horde tests do we want retried once that's published? [21:08] Launchpad bug 1562090 in php-horde-passwd (Ubuntu) "Update to PHP7.0 dependencies" [Undecided,Fix committed] https://launchpad.net/bugs/1562090 [21:09] slangasek: sorry, we'll need that one and the others i submitted today, wnat me to give you the list? I think i subscribed you to all of htem? [21:09] turba is the big one, it seems like (affecting a bunch of other packages that is) [21:10] nacc: if you subscribed me then the list is in my inbox. ok, from your earlier comment I understood that 1562090 was the only one needed for unblocking, but yes I had seen the turba failures and wondered how -passwd was going to unblock that. I'll have a look, though I'm actually off today so no promises that I'll get through the stack before Monday [21:10] slangasek: it's ok, i just wanted to let you know they were taken care of :) [21:11] slangasek: i'd rather you didn't work today and relaxed! [21:13] nacc: I relax better with fewer incoming tasks in my inbox ;-) [21:14] slangasek: uhh, unlikely! although on my last check, we were down to 350 packages! these bigger frameworks slow me down, obviously, as pulling in a new version is not the easiest thing [21:17] slangasek: magic! [21:18] barry: python-pbh5tools still broken though, also points to pysam, so retrying with -proposed pysam there as well. So maybe this unblocks it without hinting, but then python-pysam still needs attention to get it migrated [21:20] slangasek: ack === jgrimm is now known as jgrimm-afk [22:53] Hi! [22:53] I made a bugreport about E173 not being able to connect in Xenial. [22:53] Sorry. [22:54] In Wily. [22:54] It works on Xenial. [22:54] Can someone please push an update of NetworkManager and Policykit from Xenial to Wily? [22:55] The problem is at the window to enter the pin code. [22:55] It just won't accept it even thou it's entirely correct. [23:45] jdstrand: yeah, hoping to have time to sort that out pretty soon