<Vaelund>nope, just temporary, didnt think of loadkeys, thanks
<tsyesika>civodul: hey, so you said last night that the usbkbd was loaded in the inird, just wondering if you know what my next step would be to disabling it and testing if that effects the keyboard
<tsyesika>I'm guessing i'm going to have to rebuild the package though i can't find a initrd package and i'm not sure how i'd then build an identical 0.9 image except with that package switched out (I can't see anything in the docs to build ISO's)
<davexunit>mark_weaver: adding a colon to the end seemed to work
<davexunit>which is great! but not a long-term solution.
<davexunit>I installed texlive-texmf directly into my profile, which takes ages to do because it's so big. I wonder if there's a better way...
<lfam>When I package a user-facing program, I like to email the maintainers to let them know, and I like to include the output of `guix graph` as well. Sometimes there are scary things in the dependency graph and they don't realize it. My question is, how can I characterize the output of `guix graph` for them?
<lfam>The default graph type "package" is described in the manual: "It shows the DAG of package objects, excluding implicit dependencies."
<lfam>Can I say, "This is the dependency graph of foo, excluding implicit dependencies like GCC and the tar used to unpack the source tarball."
<lfam>Line wrapping of emails isn't a huge issue on its own but it does indicate a general lack of care. Reading git logs of github hosted projects is usually a total waste of time. The commit messages are just one word, or just a merge message
<xd1le>wait maybe you were talking about something else... i agree with "reading
<xd1le>i was responding to "but almost nobody does."
<lfam>And I found that if you have a feature branch, it's impossible to submit github pull requests of subsets of the branch. You can only request a pull of the entire branch. It's such a crappy subset of the git tools
<lfam>A lot of people do in projects like this one :)
<xd1le>lfam: no i mean even on github hosted projects
<davexunit>lfam: yeah, github has destroyed patch quality.
<lfam>xd1le: Oh that's good. I haven't noticed it.
<xd1le>(including owners and stuff, i guess it just depends on which projects obv)
<rekado>github makes it easy to fork, but then you're on your own and have to deal with two clones, really: the local clone of your fork, and the fork on github, which will not be up-to-date with upstream.
<xd1le>lfam: what do you mean "subsets of the branch"?
<rekado>but you need a fork to submit patches ("pull requests")
<xd1le>rekado: yeah it's bc pull requests have to be done via the web interface
<rekado>I know people who don't understand git enough to add a remote for the upstream repo and instead refork the original repo from scratch.
<lfam>xd1le: I had a feature branch that should have been submitted as multiple commits. I couldn't figure out how to do that and I couldn't get an answer on #github. I ended up splitting the branch into multiple branches and requesting a pull of each branch.
<xd1le>rekado: yeah in fact i see that all the time
<fhmgufs>I think the worst thing is that you can't change things in your fork without addind them to your PR
<rekado>what I like about github is something to do with the fact that everybody uses it; it's easy to discover projects this way. But that's nothing to do with github services, really.
<lfam>Like, I recently had a feature branch to add the Khal program to Guix. I ended up working on related leaf programs in the branch. But there's no reason to pull in all those programs in one action if you don't want to. I could have submitted them as separate patch sets.
<xd1le>fhmgufs: that's because github "forks" are not even first class repos
<fhmgufs>lfam: the github principle is not good. If you *want* to contribute to a project, you should be able to join it's mailing list and use git yourself. And if you just want to show how much you contributed in the last year and how many many projects you contributed to, also if it just were typo fixes, you can create an own mailing list for that (I think nobody would join).
<rekado>oh, I think because we don't have a prosody package yet.
<rekado>I think that's the last thing I'm missing.
<lfam>Heh, I haven't made much progress on that recently.
<lfam>I just got some new hardware for a home server. I'm wondering about starting daemons at boot. Is there a simple method analogous to (package), or will it require something more... lengthy, like the other services I see in gnu/services?
<lfam>Yeah, I don't really know. I saw there was no source code and people complaining in the comments and closed the window. But there was discussion about how this was made possible by permissive open-source licenses.
<lfam>I think it is supposed to support the sales of some specific hardware