IRC channel logs
2014-08-12.log
back to list of logs
<svetlana>how's things with the build system. and I wanted to ask, are there plans to put a correct version of one of the things into the 0.7 image? <mark_weaver>it's too late to change 0.7, it's already tagged and released. <mark_weaver>actually, allowing any version of guix to include the same version of itself internally will require some tricks. <mark_weaver>guix includes a cryptographic hash of the source code of every package it contains. <mark_weaver>putting the hash of guix-0.7.tar.gz into guix-0.7.tar.gz is not trivial :) <mark_weaver>since we can't easily include the hash of guix-0.7 in guix-0.7, the solution I have in mind is to compute the hash at build time. <mark_weaver>anyway, I hope to cook up a patch in a few days and post about it. <mark_weaver>as for the build system, it's unstuck and making progress again. <civodul>does anyone know of a wifi dongle that works with free software? <civodul>(other than the one that got RYF'd recently) <Steap>civodul: do you really need a dongle ? There are cards (for laptops) that work well too <davexunit>Steap: civodul uses a thinkpad which probably requires a hacked version of the proprietary BIOS in order to use a different PCI-e wireless chip. <Steap>civodul: Dell does that too ? <civodul>Steap: yeah, the Intel wifi thing in Dells <Steap>civodul: you can't replace it with a broadcom cared ? <Steap>maybe I'm getting mixed up with another manufacturer <Steap>they have some wifi cards that do not require a non-free firmware *Steap has an Intel card in his work Thinkpad :/ <Steap>Funny how we've got functional package managers, tools that automate installation of software with a precise configuration, but I still can't reproduce bugs from a machine to another <tadni`>mark_weaver: Hey! o/ In regards to my terminus-font package -- I checked similar bitmap font recipes in xorg.scm and everything besides pkg-config and maybe perl is a regular input. :^P <tadni_>mark_weaver: As in-everything should be a native-input? <mark_weaver>any input whose code will be executed at build time should be a native-input. <tadni_>Run-time, I assume is a regular input then? <mark_weaver>if you are cross compiling from one platform to another, the distinction is important. *tadni_ needs to check then. Though one would assume the template I borrowed from xorg.scm would be correct ... <mark_weaver>fwiw, I think you made a fair assumption to base your recipe on existing ones from xorg.scm. <mark_weaver>but we might as well try to get things right moving forward. <tadni_>mark_weaver: So should I just assume they are all native-inputs then, for now? They appear as-if they would be. <tadni_>Also, I have an an night hackathon the last friday of the month, I plan to work on wicd at ... so I'll ad correcting xorg.scm fonts to the list. <mark_weaver>I think that's a safe assumption. the output of a bitmap font recipe probably doesn't include any binary executable code, I would guess. <tadni_>Okay, I'll make the corrections to fonts.scm to terminus-font and submit it before I go to bed tonight. By the end of the month, I'll correct xorg.scm too if that doesn't happen on it's "own". <tadni_>mark_weaver: Thanks for the help!