IRC channel logs

2017-11-05.log

back to list of logs

<marusich>Hello, Guix!
<justin___>Hi all! I'm really excited about GuixSD, but am having a hard time getting it up and going as binary, USB, and also VM. Is there a anything like a "iso" image yet? Many thanks in advance.
<Digit>ACTION imagines once he starts using digital currencies properly, he'll want to do that through/on a machine running guix.... once guix devs are confident enough to drop alpha/beta from release names
<OriansJ>well confidence is the consequence of repeated success or ignorance. We try to avoid the latter.
<Digit>:)
<OriansJ>sadly we still need more people to do some crazy shit to find the pieces that need to be fixed and finish building the pieces still missing.
<TeMPOraL>what's the recommened Free Browser(tm) these days?
<TeMPOraL>(that's available under guix)
<rekado_>icecat or epiphany
<TeMPOraL>God, Guile needs a faster compiler...
<TeMPOraL>oh, conkeror is in guix \\o/
<slyfox>if guix would not delete portable bytecode files build times would be not that bad: http://git.savannah.gnu.org/cgit/guix.git/tree/gnu/packages/guile.scm#n237
<slyfox>i comment out those two lines locally to get saner build times back
<janneke>rekado_: having another stab at gcc-ddc using $ORIGIN and without nls/locales...
<rekado_>janneke: woo!
<rekado_>have you already confirmed that $ORIGIN does what we hope it would do?
<janneke>rekado_: we have been using $ORIGIN a lot with GUB, in principle it shoudl be fine
<janneke>i haven't checked with gcc (finding libgcc_s etc) specifically yet
<janneke>i did massage dkg/infinity0's build-path-prefix-map patch for guix and that brings the diff for bin/gcc down to 7 smallish chunks
<janneke>rekado_: also, OriansJ has very exciting news coming up. they're writing a transpiler in M1 that should be able to create a binary from a simplified mes.c
<rekado_>all great news
<rekado_>I followed the news on #bootstrappable
<rekado_>I’m currently in the land of boring code. SOAP and XML.
<rekado_>for mumi I won’t bother with a full SOAP library for Guile; it’s messy.
<janneke>are you working on soap bindings for guile?
<rekado_>I started but then remembered that I only need this for debbugs.
<rekado_>so I’m going to just implement the operations for debbugs that I need for mumi
<efraim>If you implement the guile bindings needed for debbugs then I'll have even fewer excuses to not write an ncurses interface for debbugs
<janneke>we need excuses not to write those, what's yours?
<janneke>;-)
<rekado_>I have no idea how to write a SOAP client, and documentation is sparse or too simple (because everybody just uses some library). In Emacs I did “(setq url-debug t)” and that gives me the plain XML that debbugs.el produces and sends off to the server.
<rekado_>that’s really helpful
<brendyn>monero has been hanging on Test 2 using 100% cpu for me. has anyone else had this issue building it? looks like hydra fails for yet other reasons
<civodul>Hello Guix!
<sneek>civodul, you have 1 message.
<sneek>civodul, janneke says: paste.lisp.net has been disabled, can we recommend another tor-friendly pastebin?
<mb[m]1>I think lfam suggested paste.debian.net.
<brendyn>Greetings civodul. Did you party hard over the weekend?
<civodul>heheh, almost :-)
***civodul changes topic to 'GNU Guix | https://gnu.org/s/guix/ | videos: https://gnu.org/s/guix/help/#talks | bugs: https://bugs.gnu.org/guix | patches: https://bugs.gnu.org/guix-patches | paste: https://paste.debian.net | log: https://gnunet.org/bot/log/guix | Guix in high-performance computing: https://guix-hpc.bordeaux.inria.fr/'
<civodul>mb[m]1, janneke: i've changed the paste in the title
<janneke>b
<OriansJ>and the first public source release of M2-Planet https://github.com/oriansj/M2-Planet
<janneke>OriansJ: congrats!
<efraim>brendyn: I have nothing useful to add other than I'm pretty sure the last time I built monero on aarch64 it built fine
<brendyn>It managed to finish building. It was doing this coretest without any output for a few hours
<efraim>I found again the softiron overdrive 1000, I don't think it has an HDMI port :(
<efraim>The 3000 has 2 PCI-E ports but is too much more
<mb[m]1>ACTION goes through some of the patches from this weekend
<benny>is there a guide on how to properly do a bug report? what all to include, whenever I do guix package -u I see two packages "will need to be upgraded" and then 2 lines later it says "nothing to do"
<db48x>benny: it'll be fine as long as you include enough information that someone else can reproduce the problem
<db48x>benny: in your case, that's probably just the version of guix (or guixsd) that you're using, and the packages in question
<vagrantc>my interest in guix was piqued again after the reproduible-builds summit ... really amazing distro you've got going on
<janneke>hi vagrantc!
<vagrantc>ACTION waves
<ilyaigpetrov>Hi. Please, bear with my silly question again. If guix is not a pure language, then someone may express his app update in impure language, so whole guixSD system update will be impure. So someone have to watch all apps update to be pure. Is this correct?
<civodul>hi ilyaigpetrov!
<civodul>ilyaigpetrov: Guix is not a language :-)
<ilyaigpetrov>right, guile
<civodul>in Guix package definitions are immutable
<civodul>but that's orthogonal to whether upgrades are "functional" or not
<civodul>in summary: don't worry :-)
<civodul>if in doubt, you can take a look at the various Guix and Nix papers
<ilyaigpetrov>what terms should I look there?
<ilyaigpetrov>is guix package manager abstracted from shepherd in such a way that I may use any other init system instead of it?
<janneke>ilyaigpetrov: the guix package manager can even be used on foreign distos such as debian
<janneke>GuixSD the distribution is tightly integrated with shepherd, using another init system would be a non-trivial operation
<efraim>I'm still getting a wierd Perl error during acl's tests on core updates on aarch64 :/
<jonsger[m]>civodul: guix challenge gives me just 68% reproducibility :(
<atw>What will the Guix/Guile presence be at the upcoming FOSDEM?
<efraim>I'll be there, just have to buy plane tickets
<civodul>ACTION will probably be there as well
<ng0>atw: no dev room this year, in terms of presence
<atw>OK, I think I still want to come...what talks etc are Guixs interested in?
***Salt_ is now known as Salt
<taohansen>what’s the status of PulseAudio on GuixSD?
<quiliro>hello guixters
<OriansJ>taohansen: it works
<quiliro>any Castillian (Spanish) speaker here
<quiliro>?
<quiliro>in Spain someone needs a speaker that aligns his views with FSF's
<taohansen>OriansJ: really? because when i start it up all that shows is a single Dummy Output
<OriansJ>taohansen: PulseAudio serves as a proxy to sound applications using existing kernel sound components like ALSA or OSS
<taohansen>i know the technical backend
<OriansJ>taohansen: so what seems to be problem?
<taohansen>PulseAudio doesn’t work on my system starting as an unprivileged user. no audio routing is present, nor are any sound devices revealed
<taohansen>my experience with Pulse is that dummy outputs are just that: nonfunctional devices with no alsa sound device behind them
<OriansJ>and you did the alsa sound configuration?
<taohansen>alsa works fine
<OriansJ>but getting pulse talking to alsa is a problem?
<taohansen>OriansJ: correct
***Igel_ is now known as Igel
<rekado_>hmm, the bioconductor importer needs updating
<rekado_>looks like they no longer keep the mirror on GitHub in sync
<rekado_>taohansen: I don’t do anything special on my system. Pulseaudio is started automatically when needed.
<rekado_>and it shows the expected output (no dummy output)
<taohansen>that’s super odd, i need to start it as a user first. i’ll try a reboot and see if that changes anything
<buenouanq>10 days until the EOMA68s ship
<buenouanq>how's uboot and arm support coming along?
<rekado_>buenouanq: are you sure they will be shipped in 10 days?
<rekado_>AFAIK the HDMI routing review is still ongoing.
<buenouanq>I thought I got an email saying that was finally fixed and they were moving on.
<buenouanq>their wobsite says the 15th
<rekado_>oh, that’s good.
<rekado_>I haven’t received any status update yet.
<OriansJ>better late than never I guess
<warreq>no status mail here either. But it's exciting news :) Hoping they will follow up with the rockchip-based design, once they've shipped the EOMA68s