IRC channel logs

2017-03-26.log

back to list of logs

***Piece_Maker is now known as Acou_Bass
<efraim>is marius here on irc?
***jonsger1 is now known as jonsger
<sphalerite>catonano: guix-dev would be the mailing list to write to about gsoc, right?
<sphalerite>s/dev/devel/
<catonano>sphalerite: I think there's a speciic email address or that but I wouldn't know what that is. If you write on de-guix you won't be wrong anyway
<catonano>on dev-guix
<sphalerite>dev-guix? Is that something different from guix-devel?
<catonano>sphalerite: no, sorry, guix-devel
<catonano>no dierence
<sphalerite>Great, thanks
<catonano>no difference
<catonano>np
<amz3>what's the relation between a package and a derivation?
<amz3>a) ^
<amz3>b) what are other items found in the store?
<catonano>amz3: I'm the last person who can talk about this. But you can think of a package deifinition as source code and a package derivation as compiled code.
<catonano>a derivation contains all the specific instrctions for the daemon to actually build the package and produce its binaries
<catonano>with all the complete and correct paths (to the dependencies) and with the actuual source base
<catonano>there's a diagram in Ludo's talks about this
<catonano>in which the passage from package to bag and ro bag to derivation is depicted
<catonano>from bag to derivation
<amz3>what is a bad?
<amz3>bag?
<sphalerite>Hm, I tried sending a message to the mailing list but it doesn't seem to have gone through
<catonano>amz3: a bag is one o the abstractions that guix uses
<catonano>it's an intermediate step between the package declaration and the compiled obect
<catonano>obJect
***jonsger1 is now known as jonsger
<amz3>ok tx catonano
<amz3>sphalerite: what's the subject of your mail?
<sphalerite>[GSoC] Newcome
<sphalerite>r
<amz3>sphalerite: some time the ml is slow, wait a bit, double check it's the correct to field and re-send your mail in 20 minutes or something
<efraim>got a rounding error on my aarch64 board building gsl
<efraim>-2.5579...e-13 observed vs 0 expected
<efraim>I didnt think aarch64 was so exotic, I think I'm going to consider again setting up test builds for some of the important yet smaller projects so I can pass build errors upstream
<htgoebel>Hi guix
<htgoebel>How can I "substitute*" a multi-line pattern?
<efraim>In my experience if you cant get \\\\\\n to register as a new line you might be stuck with a patch
<htgoebel>efraim: :-(
<htgoebel>efraim: I wanted to be more flexible then with a patch (which is pinned to a specific version)
<htgoebel>efaim: Will try hacking this myself
<efraim>i want to disable some of the mysql tests on aarch64, but the tests that fail are OOM tests and it runs out of memory :/
<lfam>efraim: I think we need to revert 36e17ab417078710939771d5ab19976eec257309 (gnu: audacity: Use gtk+@3.). At least on my system, the user interface is broken with GTK 3
<lfam>But, the new wxwidgets works fine
<efraim>If the UI is broken then that's no good
<lfam>efraim: What happens is that some elements of the UI disappear and never come back when I resize the window or roll the mouse over them, and others are truncated.
<lfam>But it works fine with that commit reverted
<lfam>Apparently audacity isn't ready for GTK 3: http://wiki.audacityteam.org/wiki/Linux_Issues#Hi-DPI
<efraim>Lots of fun
<lfam>efraim: I can do the reversion if you want. I'm testing the latest version of audacity now anyways
<efraim>Sure, go ahead. I'm still banging my head against all the test failures on aarxh64 with gsl
<lfam>efraim: Okay. Also, should I add a copyright line for you to the audacity module?
<efraim>Sure, why not
<efraim>ACTION heads off to bed
<civodul>ACTION gets a bad signature by Alex S. on commit 7398d96ee9141768de2a33df94109f7f10637d27
<civodul>lfam: same for you?
<lfam>civodul: It's good for me, with pimary key 3B12 9196 AE30 0C3C 0E90 A26F A715 5567 3271 9948
<civodul>hmm, indeed
<civodul>guile-git issue...
<marusich>Hello!