<EternalZenith>Although it can take a little while, especially your first time <EternalZenith>I'll be away for about an hour, I'll try and do what I can if you need help when I get back <cornburglar>I had it installed ~6 months ago. I'm switching back from OpenBSD <ngz>I have a trivial question. What would be a simple way to copy the contents of a directory DIR, which contains files and sub-directories, to a directory OUT? <ngz>For some reason, I only find convoluted ways to do that. <ngz>Yes, that, but in Guix lingo. <ngz>My convoluted way looks like (for-each (lambda (f) (unless (member f '("." "..")) (copy-recursively f OUT))) (scandir DIR)) <ngz>I was hoping for something more elegant. <ngz>I could throw in a `match-lambda' ... <EternalZenith>Is there a resource that shows which guix procedures translate to common shell builtins/utilities? <moaz>that's kind of how guix works I think <moaz>you would install grub and then grub-install [device] <moaz>where [device] is the appropriate device as shown by lsblk <luciddreamzzz>there was possible error in config i used (uuid="abcd-1234") 'fat32 <luciddreamzzz>it gets all the way through no errors but never installs grub because it i dunno <luciddreamzzz>is that why it couldnt find it wants 'fat32 instead im thinking <luciddreamzzz>i spose i could just setup substitute server of my own and publish em <luciddreamzzz>why doesnt it calculate hash on substitutes in store and not re-download them moaz <luciddreamzzz>i suppose the build environment and dependancies come into play when doing that... <moaz>I'm not sure I understand the question tbh <moaz>i think im too sleepy to understand whats being said <luciddreamzzz>saw nix when it was new but never gave it a try so im learning <luciddreamzzz>i want btrfs and i am wondering if it would be better to use ext4 /boot <moaz>I use xfs for everything <moaz>well to be fair I dont use guixsd <moaz>that said, I'm not sure why only those two FSes would wörk <moaz>sorry im not a powr user :-( <moaz>because I have redshift running <luciddreamzzz>hmm not shure what deal was but now super grub2 boot disk detects guixsd <luciddreamzzz>yep so guessing i can do grub-install after this crap finishes <luciddreamzzz>maybe i did and forgot but efi must have been there whole time ***rekado_ is now known as rekado
<efraim>i got distracted from core-updates by qt failing on core-updates, and now I'm working on a new qt-updates branch <g_bor>I was looking at strip-jar-timestamps in ant-build-system, and realized that it uses jar to extract the files. <g_bor>In the java bootstrap we don't have that yet. <g_bor>I don't think any problem would arise if we changed this function to use zip instead. <g_bor>That way we could use the code in the java bootstrap unchanged. <janneke>g_bor: sounds like a good way to break the bootstrap loop <rekado>g_bor: “jar” has some special behaviour when a manifest is involved (e.g. creating it when it is not provided). If that’s no concorn here then using zip would be fine. <rekado>looks like it’s only used to unpack <janneke>indeed, g_bor: what are your concerns? <g_bor>rekado: ok, thanks, just wanted to know if you already have some exeprience with this. <g_bor>I will prepare a patch then. <g_bor>did you also notice that debbugs breaks lines? <g_bor>It cause me some headaches yesterday... <g_bor>I had a patch sent inline in an e-mail, and the context contained a long line. <g_bor>That was one line in my mail client, but it was wrapped on debbugs. <g_bor>I downloaded the mbox from there, and used git am on that. <g_bor>It complained that the patch is corrupt, and indeed it was, as the wrapping created a line starting with no space + or -. <g_bor>rekado: I intend to target these java bootstrap related patches to staging, after core-updates merge. Is that ok? <g_bor>In the meanwhile, can I help on core-updates? Is there anything that is priority before the merge? <rekado>g_bor: targeting staging is fine. core-updates really should be merged soon. <rekado>we’re a quite some time behind schedule. <rekado>what everyone could do is to reconfigure their own systems on core-updates. <rekado>ajjlmau: it spawns a process and raises an error if the process fails. <rekado>we are using it instead of the (zero? (system* …)) idiom. <ajjlmau>How can I find the definition of it? <rekado>ajjlmau: it’s defined in (guix build utils) <ajjlmau>When I try to install it, it says "no code for module (screenfetch)" <rekado>There is no module (screenfetch). Your module is called (gnu packages screenfetch). <rekado>(I recommend putting it into an existing module instead of creating a new one.) <rekado>ajjlmau: to anticipate future errors: your builder needs to end on #t and (use-modules (guix build utils)) because “invoke” is defined in that module. <ajjlmau>I have set GUIX_PACKAGE_PATH to the path of screenfetch.scm and I'm trying to install it with guix package -i screenfetch <rekado>the path to Guile modules must match the module name. <rekado>for a module named (gnu packages screenfetch) you need to place it in $GUIX_PACKAGE_PATH/gnu/packages/screenfetch.scm <janneke>i get an error compiling gnu/tests/install.scm <janneke>ERROR: In procedure string-drop: Value out of range 0 to 27: 38 <ajjlmau>rekado: I changed builder to `(#:builder (invoke "chmod +x screenfetch-dev") #t) <ajjlmau>Now I'm getting "Invalid keyword: #t" <g_bor>rekado: Ok, I will reconfigure on core-updates. <janneke>g_bor: could you tell me if this works for you? <janneke>rm -f gnu/tests/install.go; srcdir=. /gnu/store/2w5kxhmkq4zkamsgx1b40vr3ii1z4dxb-profile/bin/guile -L "." --no-auto-compile -s build-aux/compile-all.scm gnu/tests/install.scm <janneke>...eh, choose your own /gnu/store/...bin/guile of course <rekado>ajjlmau: wrap this all in (begin …) <Laalf>does anyone run reverse prime on guixsd? <pkill9>does anyone know if a tool exists that reads a configure file to find a list of all dependencies required? <ng0>and sometimes the output of ./configure --help <janneke>let's hear it for mixing code and data, yay! <efraim>Reposurgeon is dead to me, broken on core-updates as-is and ported to go in the newest version <efraim>On the plus side my qt-updates branch might be ready for hydra <EternalZenith>According to the manual, the socket for users should be found in /run/user/<UID>/shepherd/socket, but that doesn't exist <EternalZenith>And again, as soon as I ask a question I figure out the issue... <EternalZenith>What is the recommended way to start shepherd on login? I'm trying to replace my startup scripts with services <EternalZenith>efraim: How are you starting shepherd in the first place, though? I don't want to have to run "shepherd" each time I log in, and there are things I would like to have started whenever I log on, even if I don't have X started <pkill9>i know some services allow you to specify a user to run them as <pkill9>although that one in particular failed to work with pulseaudio last time i tried <EternalZenith>pkill9: I've seen that, but I'm trying to use shepherd services to run everything I have in the background <pkill9>that's what i want to do EternalZenith, unfortunately the only way i know of is running shepherd as a user <efraim>perhaps something with .profile and making sure it's only running once, but i'm really not sure, we don't have anything automatic like systemd user services built in <pkill9>perhaps there could be a shepherd service for shepherd, lol <EternalZenith>I was thinking about tossing "shepherd &" in .zprofile, but I'm not sure if it's honored when logging in through SLiM <EternalZenith>I'm planning on trying to figure out how to use startx on GuixSD, but I have a bunch of other things I want to figure out <efraim>startx isn't easy, some other people have looked into it in the help-guix maillinglist <pkill9>to get the startx command, just add `xorg-xinit` package to your profile (system or user profile i assume is fine) <EternalZenith>pkill9: I have all that, but startx on here has a lot of issues when using it normally <EternalZenith>I've seen there is a service provided to create a config you should use, but I'm not exactly sure how to configure it <emacsomancer>I haven't had the chance to dig into it, but shouldn't shepherd allow for (guile) configuration of services? ***jonsger1 is now known as jonsger
***jonsger1 is now known as jonsger
***jonsger1 is now known as jonsger
<cbaines>Does it normally error, when trying to report errors? I'm seeing: ERROR: In procedure scm-error: <cbaines>I've wrapped the whole start gexp in a (catch #t ...) expression now, so at least it doesn't crash when I start the service now *janneke starts two more world rebuilds -> zZzz