/srv/irclogs.ubuntu.com/2017/05/04/#ubuntu-devel.txt

=== led2 is now known as led1
=== klebers__ is now known as klebers
=== maclin1 is now known as maclin
xnoxI need networking help.... please12:50
xnoxubuntu@DEVAC02:~$ sudo ifup enc60012:50
xnoxRTNETLINK answers: File exists12:50
xnoxFailed to bring up enc600.12:50
xnoxubuntu@DEVAC02:~$ sudo ifdown enc60012:50
xnoxifdown: interface enc600 not configured12:50
xnox..12:50
xnoxwhat is wrong with me?12:50
ogra_xnox, two identical gateway entries (default routes) on different devices ??12:57
ogra_if it is urgent you should be able to just force-ignore the error btw: sudo ifup --ignore-errors enc60012:58
ahasenackxnox: is enc600 managed by network-manager perhaps?13:09
xnoxit's in /e/n/i13:10
ahasenackn/m could still think it owns it13:12
ahasenackunless this isn't a desktop13:12
ogra_any other devices in /e/n/i ?13:13
xnoxhm. there is network manager but it should no be doing anything, everything is done with ifupdown13:16
ogra_it should at least not touch anything thats in /e/n/i (unless that changed recently)13:16
ahasenackhi, can someone from ubuntu-core-dev please take a look at https://bugs.launchpad.net/debian/+source/krb5/+bug/1688121 and accept my zesty nomination?13:24
ubottuLaunchpad bug 1688121 in krb5 (Ubuntu) "KDC/kadmind explicit wildcard listener addresses do not use pktinfo" [Undecided,In progress]13:24
ahasenackit was spun off from https://bugs.launchpad.net/ubuntu/+source/krb5/+bug/168323713:25
ubottuLaunchpad bug 1683237 in krb5 (Ubuntu Zesty) "krb5-user: kinit fails for OTP user when using kdc discovery via DNS" [Undecided,In progress]13:25
ahasenack(see comment #11)13:25
ahasenack(in the latter)13:25
LocutusOfBorgahasenack, what do you need?13:44
LocutusOfBorgjust provide a patch for all applicable ubuntu series and subscribe sponsors?13:44
ahasenackLocutusOfBorg: I need the zesty task to be accepted13:44
LocutusOfBorgdoes it affect only zesty?13:44
ahasenackyes, only krb5 1.1513:45
LocutusOfBorgI did accept it13:45
ahasenackLocutusOfBorg: I want to mark the "ubuntu" task as fix committed, as it's uploaded into artful already13:45
ahasenackLocutusOfBorg: thanks!13:45
ahasenackLocutusOfBorg: actually, since it's in artful already, that's fix released then, right?13:46
ahasenackthx13:47
LocutusOfBorgI think it is fixed now (the LP tracker)13:47
LocutusOfBorgthe Debian task was pointing to upstream bug tracker, not the Debian one13:48
ahasenackops13:48
LocutusOfBorgso you want to upload 1.15-2 as 1.15.1ubuntu17.04.1whatever13:48
LocutusOfBorgand target zesty13:48
ahasenackright13:48
LocutusOfBorgfeel free to ping me when done13:48
LocutusOfBorg(and convert both bugs to SRU)13:49
ahasenackthere are two other fixes (two more bugs) I want to include13:49
LocutusOfBorghttps://wiki.ubuntu.com/StableReleaseUpdates#SRU_Bug_Template13:49
LocutusOfBorgahasenack, fixed in artful?13:49
ahasenackI'm writing the test cases for them13:49
ahasenackLocutusOfBorg: yes13:49
LocutusOfBorgI don't get it, between -1 and -2?13:49
ahasenackLocutusOfBorg: see this comment on the original bug: https://bugs.launchpad.net/ubuntu/+source/krb5/+bug/1683237/comments/1113:49
ubottuLaunchpad bug 1683237 in krb5 (Ubuntu Zesty) "krb5-user: kinit fails for OTP user when using kdc discovery via DNS" [Undecided,In progress]13:49
ahasenackLocutusOfBorg: yes, -2 was synced from debian fixing 3 bugs13:49
LocutusOfBorgah ok13:50
ahasenackLocutusOfBorg: https://bugs.launchpad.net/ubuntu/+source/krb5/+bug/1683237/comments/9 is the artful upload13:50
ahasenackLocutusOfBorg: I'm decomposing that original #1683237 into the 3 real bugs it fixes13:50
LocutusOfBorgok13:50
ahasenackto comply with the SRU format13:50
ahasenackand 3 separate test cases13:51
LocutusOfBorgjust if you change the code from -2, you need to first update artful/Debian13:51
LocutusOfBorgand then SRU to all applicable Ubuntu supported releases13:51
ahasenackno code change, the same 3 patches apply cleanly13:51
LocutusOfBorg(sorry if I'm saying something obvious)13:51
LocutusOfBorgI know they apply, because the codebase is the same as debian13:51
ahasenackyep :)13:51
LocutusOfBorgbut you might want to add some autopkgtests or similar13:51
LocutusOfBorgfeel free but prior add them in debian or artful13:52
ahasenackok13:52
LocutusOfBorgoh13:57
ahasenackLocutusOfBorg: hi again, could you do the same for https://bugs.launchpad.net/ubuntu/+source/krb5/+bug/1688310 ? Accept the zesty nomination?14:33
ubottuLaunchpad bug 1688310 in krb5 (Ubuntu) "KDC/kadmind may fail to start on IPv4-only systems" [Undecided,In progress]14:33
xnoxmdeslaur, also you have duplicate linuxi packages in that graph. are extra kernel source packages, result in extra security / cve work?14:39
mdeslaurxnox: the graphs haven't been updating in a while to properly handle new package names, etc.14:40
mdeslaurs/updating/updated/14:40
LocutusOfBorgahasenack, .14:40
=== jsing` is now known as jsing
* LocutusOfBorg leaves14:45
=== skorv is now known as Guest20674
naccrbalint: xnox: if either of you could take a look (not immediately urgent) at LP: #1576341, i'm particularly interested in feedback on the lvm2-lvmetad.socket and systemd-journal-auditd.socket issues (c#20, 1.c, 1.d)16:37
ubottuLaunchpad bug 1576341 in systemd (Ubuntu) "systemd in degraded state on startup in LXD containers" [High,Confirmed] https://launchpad.net/bugs/157634116:37
=== daniel1 is now known as Odd_Bloke
gpiccoliHi, have a question about networking discovery/naming on Ubuntu installer, hope this is the right place to ask. Sorry to bother!17:31
gpiccoliHow the naming of interfaces works on installer? I noticed it's not using predictable naming, but after the installation completes and we reboot, we have interfaces names as enP#p#s#f#17:32
gpiccoliDuring the installer, interfaces are like ethX17:32
gpiccoliWhy? I'm *not* using "net.ifnames=0" on installer, nor providing any other boot parameter17:32
gpiccoliAlso, didn't find any reasonable udev rules that would force ethX naming...17:33
gpiccolii'm confused hehehe17:33
gpiccoliAny guidance will be much appreciated17:33
=== BrAsS_mOnKeY is now known as g2
naccslangasek: does this seem sufficient (I just tested it in a LXD dist-upgrade and it did let dlm become removable and dlm-controld was correctly installed) for dropping this delta in src:dlm? Will also allow for dropping some delta in a few other srcpkgs: http://paste.ubuntu.com/24512380/18:08
slangaseknacc: LGTM18:34
naccslangasek: thanks18:35
=== JanC is now known as Guest75375
=== JanC_ is now known as JanC
jtaylorurg gcc 7 has a new gfortran soversion19:15
jtaylorthis will cause massive problems again when gfortran.so.3 disappears from distros ...19:15
=== santa is now known as Guest85647
=== Guest85647 is now known as santa_
gpiccoliinteresting fact is that on 17.04 it uses predictable namings20:15
juliankxnox: apt 1.4.2 is out. the locking got a bit cleaned up, and now works on non-bash shells. Will sync that tomorrow :)21:11
juliankI'm about ready to open the 1.5 APT branch for artful :D21:11
=== gsilvapt_ is now known as gsilva
=== gsilva is now known as gsilvapt
gsilvaptHello everyone. Apologies if this is not the correct place to ask for help but I could use some guidance to fix this bug: https://bugs.launchpad.net/ubuntu/+source/shadow/+bug/142780721:22
ubottuLaunchpad bug 1427807 in shadow (Ubuntu) "usermod's man refers to --*-sub-uids but accepts only --*-subuids" [Medium,Triaged]21:22
gsilvaptI'm not sure what source code I should be downloading because the LP repository forwards to a really old version of the package and the available tars are not bzr repositories21:23
naccgsilvapt: `pull-lp-source <srcpkgname>`21:27
naccgsilvapt: so in this case, `pull-lp-source shadow`21:27
naccgsilvapt: it will pull the development release by default21:27
naccgsilvapt: you will first need to see if it's fixed in artful21:28
gsilvaptthanks for the reply, nacc. I've checked the code available and it hasn't yet21:29
gsilvaptI'll give those commands a try21:29
naccgsilvapt: ok, then you'll need to fix artful first :)21:29
nacc!sru | gsilvapt: per this21:29
ubottugsilvapt: per this: Stable Release Update information is at http://wiki.ubuntu.com/StableReleaseUpdates21:29
naccgsilvapt: now, i'm guessing the manpages are unmodified by ubuntu, so really it's probably an upstream bug?21:30
=== gsilvapt is now known as gsilva
=== gsilva is now known as gsilvapt
gsilvaptit is, nacc21:42
gsilvaptthat launchpad command doesn't fetch the upstream?21:42
=== gsilvapt is now known as gsilva_
=== gsilva_ is now known as gsilvapt
=== gsilvapt is now known as gsilvapt_
=== gsilvapt_ is now known as gsilvapt
naccgsilvapt: right, so fix it upstream, probably first, then get it fixed in debian and the next merge will pick it up21:43
naccgsilvapt: honestly, though, manpage updates probably don't satisfy the SRU rules21:43
gsilvaptnacc, I believe we can patch for artful, right?21:44
gsilvaptnacc, can you give me any guide on how to patch this the right way? There are so many and I don't want to mess this up :p21:45
naccgsilvapt: sure, but i think it's better to fix upstream and in debian21:45
gsilvaptI'll have a look to see if I can find the upstream source code and/or the package in debian21:46
naccgsilvapt: `pull-debian-source shadow` will give you the debian source package21:46
naccgsilvapt: and to answer your earlier question, `pull-lp-source` is for pulling ubuntu source packages, not upstream21:46
naccslangasek: i think i'm uploading the last change (src:kopanocore, a new pkg) that has any revdep on src:php7.0 not from src:php7.0 itself (src:jsusfx does technically, but it's php7.0-cli | php-cli, so is fine to leave as a sync). After kopanocore goes through, i think php7.0 is removable22:08
naccwell, i know i'm uploading :) imo it's the last change22:08
gsilvaptThanks for the explanation, nacc. I'll give that a try now. Preparing a patch file for Debian is different than in Ubuntu, isn't it?22:16
naccgsilvapt: no, not really22:17
naccgsilvapt: and more than likely, the same patch might apply to both in this case22:17
gsilvaptOh, right, I need to patch both22:19
gsilvaptShould I use quilt for this, nacc ? That package has a couple of guides to help me out22:20
naccgsilvapt: yes, because you are changing the upstream source, you'll end up generating a quilt patch22:22
sarnoldthe 'native' package format these days uses quilt internally to manage debian/patches/ -- many packages still use e.g. cdbs or patchless packages, but enough use quilt these days that it's probably in your package too22:22
naccgsilvapt: after doing a `pull-debian-source shadow`, you can make changes to the source and then run `dpkg-source --commit`22:22
gsilvaptthanks, guys.22:22
gsilvaptAlthough I read so many things about this first, I don't know why but contributing is actually super intimidating...22:23
gsilvaptOh well, lets give this small fix a try :)22:23
naccgsilvapt: we're working making it less so (with git)22:23
naccgsilvapt: but not quite there yet22:23
gsilvaptI don't think bzr is an issue. The problem is more what to edit and getting used to all these terms. upstream, debian package vs ubuntu package, check for versions, don't edit previous versions unless it can go by the SRU rules, etc22:24
naccgsilvapt: yes, our tooling basically (eventually) will let you just give us the diff22:25
naccgsilvapt: and we'll figure out what to do with it22:25
naccgsilvapt: honestly, for a bug like this one, i'd focus on getting it fixed upstream22:26
naccgsilvapt: clearly no one really cares that much -- or upstream would have fixed it by now :)22:26
naccgsilvapt: and if fixed in upstream, debian will automatically get it eventually, and so will ubuntu22:27
gsilvaptYes, that's why this is *the* bug to start. Nobody has yet fixed this and this bug is 2 years old.22:27
gsilvaptI'll download the debian package, make the patch there and the Ubuntu will get their next code from debian. That's correct, right?22:27
naccgsilvapt: when someone merges it22:27
gsilvaptyes, I can propose that too after fixing, isn't that so?22:28
naccgsilvapt: not bzr or git merge, an ubuntu merge22:28
naccgsilvapt: where changes in ubuntu get replayed (sort of) on the latest verion in debian22:28
gsilvaptYes, I got that. But can I propose that merge or not really?22:29
naccgsilvapt: it looks to be a complicated merge22:31
naccgsilvapt: you can try, but i don't think something as core as shadow is a good one to start with22:31
gsilvaptwell, I'm starting to think it actually is not a good one22:31
gsilvaptJust seemed so because it is related with man pages22:32
naccgsilvapt: you don't need to do a merge to do a bugfix22:32
naccgsilvapt: they are separate tasks22:32
gsilvaptOk, lets review the steps then. I pulled the debian package and I can package this via quilt. Then should I do the same for the Ubuntu package or propose a merge?22:33
naccgsilvapt: for this case, i would do the same for the ubuntu package (or if you want you can just do ubuntu)22:35
naccgsilvapt: my point earlier was simply that it's 'better' to ensure a fix is in debian too22:35
gsilvaptI'm sorry if these are dumb questions but I'm not a developer (aspiring one) when all this source, upstream, ubuntu vs debian packages is confusing. Also, since the shadow package is not a LP repository, it is not helping getting comfortable through the guides22:36
naccgsilvapt: the guides are all old22:36
naccgsilvapt: just use pull-lp-source and pull-debian-source22:36
gsilvaptOk, I understood your point now. I'll try to follow http://packaging.ubuntu.com/html/patches-to-packages.html and see where that will take me.22:36
naccas i said, it's easier than the guide states22:37
naccliterally: 1) pull-{debian,lp}-source shadow22:37
nacc2) edit some upstream files22:37
nacc3) dpkg-source --commit22:37
gsilvaptThe 4th step should be report it in the LP bug page, shouldn't it?22:38
naccthat will prompt for a patch name and then will open the patch in the editor, add dep3 headers (see dep3 page in debian or on that page you just linked to) and then save the file22:38
nacc4) dep3changelog debian/patches/<patchname as given in 3)22:38
nacc5) dch --edit22:38
naccinsert the appropriate release and ensure changelog is what you want22:39
nacc6) dpkg-buildpackage -S -nc -d22:39
naccThis will result in a .dsc file in the parent directory (..)22:39
gsilvaptHum, you made it sound super simple :|22:39
naccdebdiff <old dsc file> <new dsc file> > yourpath.debdiff22:39
nacc*yourpatch22:39
naccattach debdiff to bug22:39
naccdebdiff creation should be 7) and attach should be 8)22:40
naccgsilvapt: alternatively, just fix it upstream and leave the bug open22:40
naccgsilvapt: again, not a high priority bug clearly :)22:40
gsilvaptnacc, could you explain what would I have to do if I wanted to just fix it upstream?22:41
naccgsilvapt: ideally you either build locally or use a PPA to test your change does hwat you want as well22:41
gsilvaptJust to make sense of the differences22:41
naccgsilvapt: upstream has no packages22:42
naccgsilvapt: find where the upstream maintains the code and send them a patch22:42
naccgsilvapt: http://pkg-shadow.alioth.debian.org/getinvolved.php22:42
naccapparently22:42
gsilvaptuff, the upstream package is something else even. Jeezzz xD22:44
naccactually https://alioth.debian.org/plugins/scmgit/cgi-bin/gitweb.cgi?p=pkg-shadow/shadow.git;a=summary22:44
naccgsilvapt: upstream is not a package22:44
gsilvapts/package/source22:44
naccgsilvapt: upstream is ... upstream, like a github project (or any other place that can generate source code)22:44
naccgsilvapt: yes, of course it is22:44
naccgsilvapt: how would it not be something else?22:44
gsilvaptthis should be linked in the LP bug page22:45
naccwhat should be linked?22:45
gsilvaptthe upstream url22:45
naccgsilvapt: why?22:46
naccgsilvapt: i don't know what you're referring to22:46
gsilvaptto ease the process of contributing. It is always recommended you fix the upstream source in these cases but I couldn't find any reference to where the upstream source was22:46
naccgsilvapt: as i said, we are working on making it easier22:47
naccgsilvapt: fixing upstream is an open source thing22:47
naccgsilvapt: you can find it by looking at where debian/ubuntu get the upstream orig tarball from (debian/watch)22:47
gsilvaptI know, I'm not pointing fingers! I'm just new around here and I'm trying to get my grips. Sorry for bothering you this much with a super single bug22:48
sarnoldsometimes there -is- no upstream22:48
sarnolddistros are the upstreams22:48
naccyes, that's true22:48
sarnoldthis happened to rxvt, jasper, etc22:48
sarnoldprobably hundreds more22:48
sarnoldwell we package lots, maybe thousands more? :) heh22:48
gsilvaptSo, okay... Lets recap. So I have the upstream source, I download that using git, make the changes, commit it to the repository and eventually the Ubuntu guys will fetch the upstream and package updated and fix this bug. Is that correct?22:49
naccgsilvapt: well you won't ahve commit rights22:49
gsilvaptOh, right. Forgot to check that first :D22:49
rbalintthe info on shadow is confusing22:49
rbalintupstream is https://github.com/shadow-maint/shadow22:49
naccrbalint: oh interesting22:49
gsilvaptI can fork and make a PR, or nobody checks that?22:50
rbalinti'm the new maintainer but could not fix all the docs yet:-(22:50
sarnoldyou can probably fork and make a pr and probably hallyn will see it22:50
sarnoldoh :)22:50
sarnoldor rbalint :)22:50
naccrbalint: fun! yeah, i was trawling from debian PTS22:50
rbalintgsilvapt: upstream is pretty active :-)22:50
gsilvaptAh, the maintainer is here! :D22:50
naccgsilvapt: it depends on the particular project, but in this case that's fine22:50
gsilvaptOk, so help me out rbalint, please. To do a small fix in the man pages, should I fork and PR the change to the upstream?22:51
gsilvaptyea, I guess so nacc22:51
sarnoldgsilvapt: .. and if you fix it upstream and stop there, it'll eventually make its way into new ubuntu releases in the future, maybe a year or two down the road. if you want it fixed in released releases then that'll take more steps22:51
rbalintgsilvapt: i would recommend pushing to upstream directly22:51
gsilvaptthat would be clone, right?22:52
rbalintgsilvapt: yes22:52
gsilvaptsarnold, well... In this case, I might do Ubuntu later on just to test and check the entire process.22:53
gsilvaptOkay, lets try to do that. Thank you very much for all your help, nacc, rbalint and sarnold22:53
naccgsilvapt: yw22:53
rbalintgsilvapt: yw :-)22:54
gsilvaptCrap, this has been fixed upstream22:54
sarnoldhooray!22:55
gsilvaptso much trouble for nothing22:55
gsilvaptLOL xD22:55
sarnoldthen you get to figure out the patch and do the cherrypick :D22:55
hallyneh hm?22:55
sarnoldhey hallyn :)22:55
gsilvaptwhat's that, sarnold?22:55
sarnoldhallyn: gsilvapt wanted to fix a typo in shadow manpages22:55
hallynplease do :)  PR against shadow-maint/shadow, i'll accept, and the debian maintainer will pull it in22:55
sarnoldhallyn: apparently it's already fixed upstream, hehe22:56
hallynoh22:56
gsilvapthallyn, I went to so much trouble to find out it has been fixed xD22:56
sarnoldgsilvapt: so, find the log for the file in question, take a look for the patch that fixes it, and isolate that one22:56
* hallyn steps back22:56
gsilvaptaccording to git log -p, this has been fixed in 2016. I need to find this entry in the changelog, right?22:58
sarnoldthe git commit is probably more useful; find that change22:59
gsilvaptI have the commit23:00
gsilvaptThis change is not in the changelog23:00
gsilvaptAh, wait, the entry came later23:00
gsilvaptWell, I have both now, sarnold23:01
hallyngsilvapt: that's my fault.  i was not keeping the changelog uptodate23:04
hallynbc, frankly, changelogs are redundant when there's a git log, but ....  ppl want it23:05
sarnoldhehe i knowthe feeling bug something to filter out the not-user-visible changes from user-visible changes is welcome23:05
sarnoldand higher-level overview of changes is nice too23:05
sarnoldforests vs trees23:05
sarnoldbut damn they make merging patches difficult when they include changelog changes23:06
gsilvaptwhat's next, sarnold?23:06
sarnoldnot 'i hate my life difficult' but 'meh gotta deal with it difficult' :)23:06
sarnoldgsilvapt: well... i'm about to push shadow updates out the door today, so you might find it easier to wait until i do so before grabbing the ubuntu packages. hehe.23:07
gsilvaptOk, I'll wait for you then :)23:08
gsilvaptShould I write something in the LP bug page in the meantime?23:08
sarnoldgood idea, package udates always happen faster with bugs23:08
gsilvaptOkay, I will say this has been fixed in commit XXX in the upstream source.23:16
gsilvaptYey, did something tonight aside from looking at code trees and tutorials on how to contribute.23:18
gsilvaptThanks guys!23:18
hallyn+1  thank you23:20
naccgsilvapt: so when you looked earlier, were you looking only at d/changelog to see if it was fixed, or did you look in the source pkg in artful?23:35
gsilvaptI checked source pkg in artful before and it hasn't been fixed23:36
gsilvaptI then checked upstream and there it has been23:36
naccgsilvapt: ah ok23:36
gsilvaptDid I mess something up?23:37

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!