| vorlon | mwhudson: autopkgtest retries needed with all-proposed=1? | 00:25 |
|---|---|---|
| vorlon | ah, no, the package is dep-wait | 00:26 |
| mwhudson | vorlon: yeah it's build-side | 01:18 |
| mwhudson | vorlon: i think someone must have uploaded things in a very careful order to debian while we weren't importing, or something like that | 01:19 |
| seb128 | cyphermox, hey, did you get anywhere with ffmpeg/armhf? | 07:18 |
| tjaalton | what's wrong with my sbuilder.. it refuses to install libudev-dev/libsystemd-dev | 07:32 |
| tjaalton | eoan | 07:33 |
| tjaalton | proposed enabled or not doesn't matter | 07:33 |
| seb128 | tjaalton, what's the error? | 07:36 |
| tjaalton | Depends: libsystemd-dev but it is not going to be installed | 07:44 |
| tjaalton | is libudev-dev gone? | 07:45 |
| tjaalton | libudev-dev : Depends: libudev1 (= 240-6ubuntu9) but 243~rc1-0ubuntu2 is to be installed | 07:46 |
| tjaalton | ah | 07:46 |
| tjaalton | proposed had that version for a while it seems | 07:47 |
| tjaalton | but no longer | 07:47 |
| seb128 | 243~rc1 had been deleted from proposed because not ready | 07:47 |
| seb128 | yeah | 07:47 |
| seb128 | has* | 07:47 |
| tjaalton | had to downgrade libudev1, libsystemd0 that's all | 07:52 |
| lennyb | Hi, what is a procedure to import pkg to Ubuntu bionic. This pkg exists in debian (https://packages.debian.org/search?searchon=sourcenames&keywords=networking-mlnx ) | 08:42 |
| rbasak | lennyb: that depends. https://wiki.ubuntu.com/StableReleaseUpdates or https://wiki.ubuntu.com/UbuntuBackports. I suggest you start with a PPA though. | 10:19 |
| rbasak | lennyb: looks like it was removed from Bionic because of an incompatiblity with neutron | 10:20 |
| rbasak | https://launchpad.net/ubuntu/+source/networking-mlnx/+publishinghistory | 10:20 |
| Locutusofborg | cpaelzer, hello, nspr sync? | 11:59 |
| cpaelzer | I haven't looked at nspr for a while Locutusofborg | 12:02 |
| Locutusofborg | I'm trying a sync on my ppa... lets see | 12:02 |
| Locutusofborg | I'm not sure if the fix was about a build failure or something else... debian might have fixed it differently | 12:03 |
| cpaelzer | yeah | 12:03 |
| cpaelzer | I'm not 100% sure that "Set LD_LIBRARY_PATH when running tests. Closes: #925790" covers what we had to do | 12:03 |
| cpaelzer | if it builds fine in your ppa then you should be good to go | 12:04 |
| cpaelzer | at least it is the same debian bug that our delta was created for | 12:04 |
| cpaelzer | and LD_LIBRARY_PATH should be cleaner than relying on --disable-new-dtags implicitly changing the rpaths | 12:05 |
| * cpaelzer would never have thought LD_LIBRARY_PATH could be the cleaner solution than anything else | 12:05 | |
| Locutusofborg | LOL | 12:10 |
| Locutusofborg | that one was good! | 12:10 |
| Locutusofborg | and it works LOL | 12:15 |
| lennyb | rbasak, thanks. We have some bug fixes that we would like to backport into Debian. how can I know on which debian version this failed pkg was based on? Are there any logs | 12:25 |
| rbasak | lennyb: examine debian/changelog, and you might find my writeup at http://www.justgohome.co.uk/blog/2015/01/ubuntu-package-versions.html helpful | 12:27 |
| === ricab is now known as ricab|lunch | ||
| GunnarHj | Hi Locutusofborg, do you possibly have time to sponsor bug #1731459? Only upstream patch. Pre-verified via PPA. | 14:10 |
| ubottu | bug 1731459 in sane-backends (Ubuntu Disco) "genesys_gl847 scanners produce a black band in scanned images on Ubuntu 17.10+, 18.04 LTS and 18.10 Cosmic cuttlefish" [Medium,In progress] https://launchpad.net/bugs/1731459 | 14:10 |
| === ricab|lunch is now known as ricab | ||
| === cpaelzer__ is now known as cpaelzer | ||
| Locutusofborg | GunnarHj, ack | 14:58 |
| GunnarHj | Locutusofborg: Thanks a lot! | 15:24 |
| blackboxsw | vorlon: thanks for the suggestion on cloud-init use of debconfig-loadtemplate for ubuntu-drivers work. The only wrinkle here is that cloud-init doesn't currently have a dependency on debconf-utils package. I'd like to avoid adding that new strict package dependency if we can to avoid bloat (57K) on most cloud-init deployments which don't try to configure ubuntu-drivers. one option is for cloud-init to | 17:02 |
| blackboxsw | install debconf-utils if we are already processing ubuntu-drivers in cloud-config. Are there any other alternatives you would suggest? | 17:02 |
| vorlon | blackboxsw: oh, sorry, I overlooked that the command was in the non-default debconf-utils. You can also use the X_LOADTEMPLATE debconf protocol extension to load it without having to install other packages | 17:07 |
| blackboxsw | ahh /me rtfms now thanks | 17:08 |
| vorlon | so then you just have to talk the debconf protocol, no big deal :P | 17:08 |
| blackboxsw | SMOP :) | 17:08 |
| blackboxsw | vorlon: looks like we'll have db_x_loadtemplatefile in /usr/share/debconf/confmodule so that should be a big help as it's delivered by debconf proper | 17:15 |
| blackboxsw | ok we can work with this | 17:16 |
| blackboxsw | thanks again | 17:16 |
| vorlon | blackboxsw: yeah, you're just then subjected to implementing in shell in order to use /usr/share/debconf/confmodule but otherwise ok :) | 17:16 |
| blackboxsw | vorlon: or maybe I could push a patch back to python3-debconf to support x_loadtemplatefile as it seems like that would be a trivial upstream commit as it seems the only command not supported by the python library | 17:22 |
| blackboxsw | maybe I'll try both routes, and cloud-init can adopt the python3-debconf library in the future | 17:22 |
| vorlon | ah, given that python3-debconf is already in the server task (which I hadn't realized), that seems like the better approach IMHO | 17:23 |
| vorlon | though we'd then have to SRU it back | 17:23 |
| blackboxsw | yeah and we'd need to MIR in xenial : / | 17:24 |
| blackboxsw | as it looks like it's only back to bionic | 17:24 |
| * blackboxsw checks state of /usr/share/debconf/confmodule in xenial too | 17:24 | |
| blackboxsw | db_x_loadtemplatefile exists in xenial /usr/share/debconf/confmodule | 17:25 |
| blackboxsw | hrm, ok so need to weigh using subp shell to source db_x_loadtemplatefile /usr/share/debconf/confmodule vesus just installing debconf-utils if ubuntu-drivers cloud-config is specified | 17:26 |
| blackboxsw | vs. rolling our own simple wire write of debconf cmds for X_LOADTEMPLATEFILE support | 17:26 |
| vorlon | blackboxsw: there's no expectation that we'll have these drivers on xenial fwiw | 19:02 |
| Odd_Bloke | Is there any way to change the make target that dh_auto_test will run, or should I just override it and run `make not_quite_test_but_so_close` myself? | 22:30 |
| Odd_Bloke | (`make test` runs both Py2 and Py3 tests, I'm looking at dropping the python2 binary packages produced, so don't want to install the Python 2 build deps to make the tests run.) | 22:31 |
| rbasak | https://sources.debian.org/src/debhelper/12.5.3/lib/Debian/Debhelper/Buildsystem/makefile.pm/#L167 | 22:54 |
| rbasak | Doesn't look like it - not for the "make" buildsystem implementation | 22:54 |
| rbasak | Assuming you're using the make buildsystem and not one of the Python ones | 22:56 |
| rbasak | So overriding it seems like the appropriate mechanism to use | 22:56 |
| mwhudson | Locutusofborg: um how many rebuilds is this golang-google stuff going to take? :) | 23:23 |
| mwhudson | Locutusofborg: fwiw it seems the grpc failure is real | 23:23 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!