UbuntuIRC / 2021 /01 /11 /#ubuntu-devel.txt
niansa
Initial commit
4aa5fce
raw
history blame
6.3 kB
=== cpaelzer__ is now known as cpaelzer
[10:01] <cpaelzer> Hmm, was there any hickup that might have eaten my armhf autopkgtests recently?
[10:02] <cpaelzer> qemu depends on the "Test in progress" armhf runs for initramfs-tools libvirt and open-iscsi; But these tests are neither recently completed nor listed in the "running" page of autopkgtest
[10:02] <cpaelzer> so it seems they were lost; The solution is an easy retry but if there was a hickup removing them it would be good to know abut it.
[10:12] <cpaelzer> hmm - also https://autopkgtest.ubuntu.com/request.cgi?... seems to be not responding ...
[10:31] <Laney> cpaelzer: it's erroring: https://paste.ubuntu.com/p/6QqgCxXTnN/
[10:32] <Laney> dunno what this means right now :>
[10:43] <Laney> hmm no, maybe that's a red herring, I managed to submit some requests
[10:59] <cpaelzer> yeah Laney WFM as well now
[11:06] <Laney> weird though, it looks as though some requests never made it into the queue
[11:27] <rbasak> Can someone point me to an example of a manual source pocket copy into hirsute-proposed please? I'm trying to find an edge case example for some git-ubuntu development.
[11:27] <rbasak> eg. from a PPA
[11:27] <rbasak> Maybe even a binary copy.
[11:33] <rbasak> Aha, got one
[11:33] <rbasak> I ended up writing a search function
[11:40] <cpaelzer> Laney: and vice versa some of my error 504 requests got executed
[11:42] <Laney> there seems to have been some kind of disturbance in the force
[11:43] <Laney> I'm not sure why some of these requests went missing either, maybe rabbitmq had a sad period
[11:51] <Laney> I'm going to wait for the queue to drain, restart rabbitmq and get proposed-migration to re-request all its outstanding tests
[12:39] <apw> rbasak, any kernel i would think
[12:44] <rbasak> apw: yeah the first hit was linux-meta-azure 5.4.0.1037.35
[12:44] <rbasak> from ~canonical-kernel-team/+archive/ubuntu/ppa
[12:46] <apw> rbasak, and if you need one without binaries for comparison, the linux-signed-azure at a similar version would be that
[13:00] <rbasak> Ah, perfect. Thanks!
[13:03] <iourarum> Hi, quick question, I have installed intel-mkl and intel-mkl-full in Ubuntu with sudo apt get install. Is this supposed to install ifort / intel fortran as well?
[13:06] <tarzeau> iourarum: no, imho. mkl is just the libraries
[13:07] <tarzeau> iourarum: you'll have to buy ifc/ifort and icc
[13:08] <iourarum> tarzeau: thank you
[13:08] <tarzeau> which version of it do you use? and how's your benchmarks against amds compiler, gcc, clang?
[13:09] <tarzeau> iourarum: are you doing astronomy or particle physics? (not many people still use fortran)
[13:09] <tarzeau> and how's your experience with flang?
[13:09] <iourarum> tarzeau: I am doing computational chemistry
[13:09] <iourarum> tarzeau: I am actually trying to install cp2k for RESP charge derivation of some molecules :)
[13:10] <iourarum> tarzeau: I can't comment on compiler performance yet, not fortran at least
[13:10] <tarzeau> iourarum: interesting, any free software you use that's not packaged yet?
[13:12] <iourarum> Hmm, well it's nothing urgent really. I use vmd for visualisation but it's easy to install. I use gromacs for MD simulations but for these things you prefer to install from source I guess (to enable GPU acceleration). Other software I use I many times install using conda.
=== halvors1 is now known as halvors
[16:46] <ItzSwirlz> hey, so in upstream nemo a bug patch was made and was released. instead of patching nemo, instead can we package the new version for ubuntu only since debian wont do it anymore
[16:47] <ItzSwirlz> current nemo version is 4.4.2, the bug patch release is 4.4.3. is it better to sru that one patch or submit a pr to launchpad?
[16:49] <ItzSwirlz> it seems just sru the patch is the correct thing to do, yes?
=== ebarretto__ is now known as ebarretto
=== ebarretto__ is now known as ebarretto
=== ijohnson is now known as ijohnson|lunch
[18:04] <rbasak> ItzSwirlz: I'm not sure I follow your question exactly. But if you're asking whether to take an upstream version bump that is actually just the one patch different from a version in Ubuntu, or to add a quilt patch that is the same patch, then mostly from a policy perspective nobody cares since the functional effect is the same, and so it's up to the uploader for whatever's easiest proess-wise.
[18:04] <ItzSwirlz> well that'd be challenging becuase I'm not the uploader. I'm just trying to get sponsors to make SRU's really right now
[18:05] <ItzSwirlz> also nemo isn't really setup fully in launchpad, so an sru would be appropriate
[18:05] <rbasak> I'm not sure what you mean by that
[18:05] <rbasak> Do you mean git?
[18:05] <rbasak> git is secondary in Launchpad - it's an SRU either way
[18:05] <ItzSwirlz> yeah.
[18:05] <rbasak> And what is approved is an upload for which there is no real distinction as to whether git was used or not
[18:05] <ItzSwirlz> I think it's better to just add the patch and not basically make a new release
[18:07] <rbasak> Sure
[18:07] <rbasak> IMHO, both methods carry approximately the same level of but different pain
[18:07] <rbasak> From an SRU review perspective I'd accept either
[18:08] <rbasak> If you're not an uploader, then I'd negotiate with your potential sponsor, or if you don't have someone in mind, then just do it whichever way you prefer and I'd hope your sponsor will accept that.
[18:08] <ItzSwirlz> i'll get to work on it soon, school rn and i need to produce the bug first
[18:45] <fossfreedom> ItzSwirlz: I'm happy to be your sponsor for nemo related stuff. Just shout.
[18:45] <ItzSwirlz> alr
[19:39] <sergiodj> slyon: hey there! I think I need someone to approve my last upload of s390-tools so that it can be published and I can then upload s390-tools-signed
[19:40] <sergiodj> slyon: if you can't do this, do you know who can? perhaps some AA?
=== ijohnson|lunch is now known as ijohnson
[21:09] <rafaeldtinoco> xnox: I remember you were the person to go to on on s390-tools not long ago
[21:09] <rafaeldtinoco> sergiodj: ^
[21:31] <sergiodj> rafaeldtinoco: thanks!
[22:25] <banekondic996> Why do i have to use xrandr and to type 3 lines of code to get custom resolution of 1920x1080 in VM which is common? While in windows you can find modeline in adapter settings