| === Elimin8r is now known as Elimin8er | ||
| === _salem is now known as salem_ | ||
| === salem_ is now known as _salem | ||
| doko | xnox, so for #793222 I think I should at least break libboost-date-time1.55.0 unconditionally, not sure for the other libs. could you have a look at https://codesearch.debian.net/results/catch.*%3A%3Afailure%20package%3Aboost1.55/page_0 ? | 09:41 |
|---|---|---|
| xpheres | hello | 10:47 |
| xpheres | if anyone has an ubuntu touch device I will be happy if you write in reviews that my app actually works: https://uappexplorer.com/app/analyticaltranslatordemo.xpheresdev | 10:48 |
| Laney | doko: trying a build with new dh-python | 11:45 |
| Laney | but pyside does some renaming action itself | 11:46 |
| doko | Laney, see my fixes for newt and comedilib | 11:47 |
| doko | or let dh-python do the work | 11:48 |
| Laney | looks like it just renames stuff so that dh-python will pick it up | 11:54 |
| Laney | will check on the test build later | 11:54 |
| doko | thanks | 11:57 |
| xnox | doko: what do you mean break boost1.55? boost 1.55 should remain on 98 abi, otherwise it will be nuts.... | 13:42 |
| xnox | doko: boost1.58 is with C++11 abi and should be peachy, no? | 13:43 |
| doko | xnox, please read the bug report, that's about not being able to catch an exception | 13:44 |
| doko | so with libstdc++6 5.2.1-12, this may get broken in boost1.55 | 13:45 |
| xnox | doko: i'd rather have broken parts of boost1.55 and deal with that. i only see like 20 packages using boost-date-time | 13:48 |
| doko | xnox, so I'll add the breaks? | 13:48 |
| xnox | yes. | 13:48 |
| doko | but just for the datetime package? | 13:49 |
| xnox | doko: yes. | 13:51 |
| xnox | doko: i don't see for other libraries to be affected. | 13:52 |
| xnox | from codesearch. | 13:52 |
| xnox | (there is like tests and examples) | 13:52 |
| sforshee | hallyn: overlay is the upstream fstype, overlayfs is the legacy fstype in wily. As far as mount options go, the same code is used for both fstypes - if you supply upperdir then you also need workdir. I don't know how it worked in the past though. | 15:08 |
| hallyn | sforshee: as i said above, 'overlay' is not supporting unpriv mounts now, 'overlayfs' still does. | 16:08 |
| hallyn | 'overlay' used to, so it's a functionality regression | 16:08 |
| hallyn | sorry, heading out right now | 16:08 |
| teward | rbasak: infinity: so, i discovered something on the Apache package, they actually install to /usr/share and then have a postinst script that copies files around | 17:10 |
| teward | (it's their solution to the /usr/share/... being written to by package managers problem...) | 17:10 |
| teward | it's an interesting solution... | 17:10 |
| teward | (for the default site) | 17:11 |
| teward | ... and it's a solution i could adapt for nginx... with some work and testing, of course | 17:14 |
| infinity | teward: Pretty sure I suggested that very same thing for nginx. | 18:58 |
| teward | infinity: i didn't see that, sorry, but i wanted to check the logic :) | 18:59 |
| * teward yawns | 18:59 | |
| teward | the side effect of bad-sleeping during the week: things're missed | 18:59 |
| infinity | teward: With a caveat that it should only do so on new installs, and only if /var/www/html is empty (so you don't stomp over another webserver or user content) | 18:59 |
| teward | right | 18:59 |
| teward | infinity: there's checks for that too | 18:59 |
| teward | in the apache implementation | 18:59 |
| infinity | I would assume so, yes. | 18:59 |
| teward | but my current focus is a trusty bug that should've been sqashed and i was too lazy to squish | 18:59 |
| infinity | So, it probably should just do the same thing and be done with it. | 18:59 |
| teward | right, that's my thoughts | 19:00 |
| teward | can never have enough testing though | 19:00 |
| ari-tczew | cjwatson: I've got an odd problem with builders. The package subtitleeditor has been built fine on PPA, but in wily-proposed got FTBFS. Could you check that? | 22:27 |
| ari-tczew | PPA: https://launchpad.net/~ari-tczew/+archive/ubuntu/testing/+build/7725896 | 22:27 |
| ari-tczew | wily: https://launchpad.net/ubuntu/+source/subtitleeditor/0.52.1-1 | 22:27 |
| slangasek | ari-tczew: check the versions of the build-dependencies that were installed in the ppa, vs. the ones used in the archive? | 22:37 |
| ari-tczew | slangasek: there is no build-depends installed in ppa | 22:43 |
| slangasek | ari-tczew: that is categorically false | 22:55 |
| slangasek | ari-tczew: what I mean is that you have build logs, which show you which versions of the build dependencies were installed and used for each build | 22:55 |
| ari-tczew | slangasek: why versions should be different? PPA takes B-D from -proposed, doesn't it? | 23:02 |
| slangasek | ari-tczew: a) that's a configurable setting, so not necessarily; b) the ppa build and the archive build didn't happen simultaneously so the contents may have changed in between | 23:05 |
| ari-tczew | slangasek: the first what I pointed out is a count of build-depends. ppa gets only 13 packages, archive build gets 22, but I guess it doesn't matter | 23:09 |
| ari-tczew | slangasek: can be it a problem related to gcc 5 ? | 23:11 |
| slangasek | gcc5 has not landed in the archive | 23:11 |
| teward | slangasek: that's still in testing first, right, before it lands? What was the timeline on gcc5 landing, out of curiosity? (I probably need to make sure nginx builds there) | 23:15 |
| slangasek | teward: https://lists.ubuntu.com/archives/ubuntu-devel-announce/2015-July/001143.html | 23:16 |
| infinity | ari-tczew: Your PPA isn't configured with -proposed enabled, so it's not surprising that the results differ from an archive build. | 23:51 |
| ari-tczew | infinity: I had no idea there is a way to configure it... thanks | 23:56 |
| ari-tczew | Primary dependency added: Primary Archive for Ubuntu - PROPOSED | 23:56 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!