<nckx>lispmacs: What did help me get my remote printing set-up working was (log-level 'debug) (access-log-level 'all) and tail -f /var/log/cups/error.log. Then you can at least figure out who's not talking to whom. <brettgilio>ScaredySquirrel: it takes a lot of effort to get gnome in a consistent and functional state <rekado>nckx: actually, neither of the two Overdrives appears to be reachable <nckx>rekado: Sergei is definitely up. I've restarted dmitri, but it's refusing SSH connections. <nckx>The dynamic IP is currently 109.128.138.53. <ScaredySquirrel>they have 10th generation Intel Core processors versus the 9th generations found in MacBook Pros <nckx>ScaredySquirrel: I didn't say which port… 😉 But thanks for testing. <nckx>Service ssh-daemon is not running. <nckx>Service ssh-daemon has been started. <nckx>rekado: I can reach them both from berlin. I noticed that the port numbers are swapped in the tunnels (i.e. 2222:dmitri.tobias.gr:5551 will take you to sergei instead) but that doesn't matter in practice. I'll swap them on my end. <nckx>Actually, that might mess up known_hosts, let's leave them in this state. <ScaredySquirrel>the mouse cursor just jumps around the screen and I lose cursor control unless I reset the touchpad with off then on <nckx>lispmacs: Did you restart cups? This sometimes happens when I update splix & CUPS gets out of sync (using an older store path). We'll probably need the exact 'debug error from error.log (see above) to get much further. <lispmacs>nckx: i did. I'll reconfigure with the debugging, reboot the whole machine, and try again <nckx>ScaredySquirrel: It would help if you gave the make & model of your touchpad/laptop. <nckx>lispmacs: OK, good luck. <nckx>ScaredySquirrel: For example, psmouse.synaptics_intertouch=1 on the kernel command line used to ‘fix’ some issues. *nckx disabled their touchpad because touchpads are the Devil's skin flap. <lispmacs>nckx: I am getting same behavior after reconfig and system reboot. Here are recent errors showing in the log: <lispmacs>I have to log out, so please send messages to the bot. <gnutec>nckx: Yeh! Best thing I did was buy a mouse to my notebook. <ScaredySquirrel>I have tried i8042.reset=1 for the Linux kernel arguments but it won't go away <gnutec>nckx: But I like the steam controller. <gnutec>ScaredySquirrel: Me? I only have problem with Wifi FN. <gnutec>ScaredySquirrel: But this problem happen only with Ubuntu and Trisquel. <leoprikler>(use-package-modules enlightenment) I would assume <bandali>leoprikler, do you know if simply adding a ;; comment to a package's definition will cause rebuilds? ***catonano_ is now known as catonano
<brettgilio>bandali: what is your question? I think it got buried. <bandali>brettgilio, i'd asked if adding ;; comments in a package's definition would count as difference and thus cause rebuilds? <bandali>not commenting anything out, but just adding a few short words here and there <brettgilio>In my thoughts Cuirass would check the commit, see the derivation matches, and then basically stop there not rebuilding anything. But maybe I'm wrong, bandali <bandali>thanks for your thoughts, brettgilio. i'd love to hear from nckx or other folks more familiar with the inner-workings of cuirass and/or guix itself <brettgilio>raghavgururajan: I am not a maintainer, but I have commit access. What is up? <raghav-gururajan>brettgilio Oh I see. Thanks. I have sent two new revised patches for gnome-characters and gnome-font-viewer. <brettgilio>raghav-gururajan: I am working through debbugs right now. I'll get to them shortly :) <brettgilio>raghav-gururajan: if I forget, give me a ping on here or at brettg@gnu.org tomorrow <brettgilio>I am on baby duty tonight, so there's a good chance that will take priority :) <brettgilio>raghav-gururajan: thanks for your effort and care! <brettgilio>Hi raghav-gururajan, I have responded to your gmome-characters email with some suggestions. <bandali>don't we already have a sway package? <ScaredySquirrel>I think the latest KDE fully supports it but the latest isn't in guix <ScaredySquirrel>I tried enlightenment_start like the enlightenment channel #e here says <brettgilio>I don't use GNOME but is GNOME with Wayland compositing done on Guix? <bandali>but as far as wayland goes, it should be okay (e.g. sway i just mentioned) <brettgilio>ScaredySquirrel: is there any open tickets on our issue tracker for it? <brettgilio>ScaredySquirrel: I think mine are rendering from google-noto <ScaredySquirrel>I have one problem: I want to change the font antialiasing to make it less blurry <lispmacs>what command has the terminal command "reset" in it? <ScaredySquirrel>I did rebalance parentheses because I saw how scheme applies functions <brettgilio>raghav-gururajan: i hope my email doesn't come off as harsh. I am always worried criticism I offer sounds mean lol <raghav-gururajan>brettgilio Oh not at all. They were really good suggestions. I am still learning :-) <raghav-gururajan>Folks! I am not able get the `.pre-inst-env guix lint package-name` to work. I keep getting no package found. I tried even after doing `git add` and `git commit`. <brettgilio>Can you put the whole output in a pastebin for us? raghav-gururajan <g_bor[m]>raghav-gururajan: only the lint is what's not working? What do youu get for ./pre-inst-env guix edit gnome-characters? <g_bor[m]>ok, I assume this is a modified checkout. Could you paste the diff? <brettgilio>raghav-gururajan: just got your email to the help list. I'm AFK now. So if nobody responds by the time I wake up I'll respond. <brettgilio>raghav-gururajan, rekado_ just pushed your gmome-characters package <talope>nckx: thanks for your help yesterday. My laptop is now hibernating. I ended up cloning guix and using channels.scm to direct to my clone rather than the environment approach you pointed me too -- basically because I couldn't get it to work with the nonguix channel I still have to use for my wifi card. loginctl has ample support for hibernation -- too bad the patch is not in, i'm going to refresh it and try to mainline it somehow. <fetsorn>Could anyone please help me write a system config with several user accounts? <rekado_>looks like GRUB in the installer switches to graphical mode, so I can’t actually install the new servers remotely via serial console :( <pkill9>how might i go about diagnosing an issue whereby when i wake the computer up from suspend, the laptop wakes up but the screen is blank, and switching virtual terminals does nothing <pkill9>note: im using the gnome desktop environment <leoprikler>I'm sure this issue has already been reported w.r.t. GDM. <leoprikler>I'm not sure how to fix this however, since you deliberately suspended your laptop (rather than inadvertently through GDM). <valignatev>Hello-hello! Maybe kinda dumb question, but is is there some way for cargo-build-system to figure out needed cargo-inputs automatically without specifying them manually? It looks like that it's possible considering that packages in crates-io.scm doesn't always have cargo-inputs explicitly specified <raghav-gururajan>leoprikler Thanks so much for your email. After `guix environment --pure guix -- ./bootstrap && ./configure && make`, should I stay in env or leave the env, for doing guix lint? <leoprikler>I don't think it matters as long as you prepend ./pre-inst-env <leoprikler>You can't do your git stuff inside the pure environment though, so if you want to reuse it, you have to spawn a second shell to do things in parallel. <raghav-gururajan>Ah, that is the issue I am facing. When I do `./pre-inst-env guix lint package-name`, I get no package found error. <leoprikler>(I prefer this approach, because then I don't accidentally use the normal guix, as it's not available in the pure environment.) <raghav-gururajan>Before I do `./pre-inst-env guix lint package-name`, is it enough to that insert the package def in gnome.scm file or should I do `git add` and `git commit` as well? <str1ngs>raghav-gururajan: you don't need to commit. just add it to gnome.scm <str1ngs>raghav-gururajan: no you don't need to add. <leoprikler>nope, you can test your packages before `git add`, `git commit` et al. <str1ngs>raghav-gururajan: though if you plan to have the changes for awhile. it does not hurt to commit to a branch. and rebase the branch from master as needed <raghav-gururajan>rg@secondary ~/guix [env]$ ./pre-inst-env guix lint gnome-characters ./pre-inst-env: line 62: exec: guix: cannot execute: Is a directory <str1ngs>raghav-gururajan: are you on foreign distro? <str1ngs>raghav-gururajan: try ./pre-inst-env /full/path/to/guix . does that fix the issue? <str1ngs>raghav-gururajan: also see with which -a guix if there is an issue with your PATH <raghav-gururajan>Doing a clean start. Cloned git repo. Doing `guix environment --pure guix --ad-hoc git nano`. <str1ngs>it's like a play by play of the GNU three finger salute! :P <raghav-gururajan>Did `nano gnu/packages/gnome.scm` and inserted package def for gnome-font-viewer. <raghav-gururajan>str1ngs I tried `./pre-inst-env /home/rg/guix guix lint gnome-font-viewer`. But got same error. :-( <raghav-gururajan>`which -a guix` gives command not found and that's understandable as I am in pure env. <raghav-gururajan>leoprikler Same error "./pre-inst-env: line 62: exec: guix: cannot execute: Is a directory". <valignatev>What "(properties '((hidden? . #t)))" means in crates-io declarations? I couldn't find anything useful in the manual <raghav-gururajan>Btw, I tried `./pre-inst-env guix build hello` outside env and it works. But outside env, If I do `./pre-inst-env guix build gnome-font-viewer`, I get no pakcage found. <str1ngs>raghav-gururajan: after configure you need to run `make or guix scripts/guix will not exist <raghav-gururajan>Okay. Some process is going on. Lot of file names with".o" extension. Something like CXX etc.. <str1ngs>raghav-gururajan: this might take some time, but you only need to do it once <raghav-gururajan>Also, If I leave the env or reboot my system, do I have to repeat any of the bootstrap, configure, make or git config steps?? <leoprikler>if your setup lives in /tmp you have to redo everything <leoprikler>otherwise you have to occasionally redo the bootstrap/configure if important files change, and regular make doesn't hurt <raghav-gururajan>ice-9/eval.scm:293:34: no binding `zip' to hide in module (gnu packages compression) <leoprikler>you probably have a syntax error somewhere which leads to weird error messages that can't be trusted <raghav-gururajan>Woah! I think the issue that I thought was with gnome is not true. The process called "guile" is hitting 200% CPU when I experience desktop stuttering in gnome. <kirisime>raghav-gururajan: Guile is your init, service and package manager, and build system. <rekado_>raghav-gururajan: I no longer have the x200s. <raghav-gururajan>rekado_ Ah I see. By any chance do you remember if that wifi chip you bought responded to kill-switch? Or do you use that chip on your current device? *raghav-gururajan wishes if gnome project releases lts versions. <rekado_>raghav-gururajan: it doesn’t fit into the new laptop that I have. <raghav-gururajan>rekado_ I was told that, that chip model does not turn-off when kill-switch is used, on x200 series devices. That's why wanted to ask you. <kirisime>raghav-gururajan: Which version of GNOME are you on currently? I might've had your issue and simply thought that the video I was watching was partly corrupt, but after upgrading to 3.32 the video in question seems to play fine. <raghav-gururajan>kirisime 3.32. But the issue started with 3.30. Things were fine before 3.30 upgrade. <kirisime>And I might've had my issues even in 3.28. Odd. <bandali>rekado_, do you if adding some ;; comments here in and there to a package definition will cause a rebuild by cuirass? not commenting anything out, just adding a TODO or something <snape>bandali: no it won't trigger any rebuild <snape>np :) Actually you can try to add a comment to any Guix package and you'll see that the inputs don't change and the commented package won't need any rebuild <snape>it's more a Guix thing than a Cuirass thing <snape>However, there'll be an Cuirass evaluation triggered, because there is one every x seconds if a new commit appeared <snape>but that evaluation won't return any new derivation to build (except for testing) <bandali>right, but i think that'll be inevitable: i'm adding a new package (emacs-next), but ludo prefers i don't change the definition for emacs for now if possible <bandali>so i'll just add a TODO comment for the change i'd like to see be done later at a more convenient time <kirisime>`guix lint' should probably check for commenting style and indentation if it's possible, though. <snape>yeah, adding a TODO comment is definitely safe ;) <mehlon>is there some way to search packages online? <sneek>mehlon, raghav-gururajan says: : flatpaks require FHS, which is not supported in guix yet. I have seen some conversations regarding a service for that. If you are looking for featured web browsers, you can use icecat and ungoogled-chromium. If substitutes are unavailable and you do not have high CPU power, you can wait for the build farm to finish building those substitues. <navik>Trying to boot from USB using the ftp.gnu.org-provided 1.0.1 x86_64 iso after dd from pureboot; it fails detecting the EFI and panicks out to recovery - is this expected? <navik>I'm currently checking the integrity of the stick, probably some of my own errors lurking. <navik>mehlon: obvioiusly not dd from pureboot, but it's pureboot that fails to detect the EFI of the USB <mehlon>is pureboot EFI compatible? I've never heard of it before <navik>ah, it ships with pureos on the librem systems from puri.sm <navik>rebranding of something else <mehlon>it might be that the GRUB bootloader included does not work with pureboot <navik>could be, should check with other device then <navik>I'm getting shit from fdisk on the USB; GPT PMBR size mismatch (2694403 != 30277631) <navik>lol, how did I even manage misspelling that. :) <mehlon>I can't find anything regarding grub on duckduckgo.. hmm <mehlon>based on what I've read so far pureboot is coreboot with additions; and the bootloader GRUB that guix uses does not work out of the box with it <mehlon>you'd need to make a custom image to support it <navik>mehlon: thanks, then I can stop bashing my head against the wall and start reading :) <navik>it did work on my other device also, so confirms the integrity of the stick <mehlon>what OS and computer model do you have? <mehlon>you'd need to find some generic way to run a custom distro, any distro, then apply that to guix <navik>mehlon: I'm on an archlinux system for creating bootable media, and have a librem laptop with coreboot as you say, which I intend running guix on <navik>It seems that archlinux installation ISO works fine, so I suppose using stuff from there would suffice <navik>at least, It found some EFI-entries, testing now <navik>mehlon: I stand corrected - there's obviously more to it. <valignatev>Can you some how give an entry that gets created during "reconfigure" a custom name? Something like guix system reconfigure --name "experiment with dangerous stuff". And then this "experiment with dangerous stuff" appears in grub <mehlon>I'd figured that coreboot cannot boot UEFI stuff <navik>mehlon: anyways, thanks for your guidance, I might post more thoughts here another day <mehlon>you should probably ask in a librem-specific channel <navik>I guess I'll have to look into the details of librem boot process. <g_bor[m]>mehlon: I thought that tianocore was a valid coreboot payload... <navik>please, enjoy a learning process displayed g_bor[m] :) <mehlon>running tianocore -> grub uefi might work... <guixy>I want to add a wrapper for the clojure package. <guixy>As I understand it, clojure is a java library and a java executable. <guixy>Is it better to add the wrapper to the clojure package itself, or to make a new package (clojure-executable maybe?) that wraps clojure? <guixy>I'm guessing the wrapper belongs in the clojure package itself. Like how plantuml has its own wrapper (and its own problems). <immodicus>Immodicus:I was trying to install guix on an external hard disk,out of pure curiosity, but suddenly it gives me the error "corrupt input while restoring the archive from # <closed: file 3bd97e0>". I used the graphical installer and automatic partitioning ...can anyone tell me how can i eventually resolve?can it be an installer bug? <guixy>That sounds deadly. How does your main disk look? <guixy>I think I encountered something like that once. I tried to make two disks in a virtual machine. One as the system root, and the other as /gnu/store. It didn't install back then. But this was in August. <g_bor[m]>guixy: there is a bug in the tracker about that. If you add a wrapper feel free to close it. <g_bor[m]>I would not add a separate package, most distro does have a wrapper in the package, so users will expect it when installing the packages <g_bor[m]>it's just something like java -jar clojure.jar ... I believe you can peek at the debian package also if needed. <mehlon>is your network connection unstable immodicus? <g_bor[m]>Immodicus: no, guixy was looking into the clojure thing. <guixy>Ok. I'll add a wrapper to the package itself. <guixy>Why doesn't clojure already have a wrapper? <Immodicus>can anyone please help me ? I don't understand why it gives me that error lol <g_bor[m]>guixy: because it does not come with on by itself. It is done ad-hoc on all distros. <mehlon>you can try repeating the install <mehlon>or, your installation medium itself might be corrupted.. there should be a way to do a hashsum check <guixy>Immodicus: I'm not sure why it gives you that error. <g_bor[m]>I was thinking about doing it, but I left is as an open easy bug, so when someone in the community shows up with interest, they have an easy and fast contibution to make :) <g_bor[m]>Does it happen on the system install page or earlier. <guixy>You could send a bug report. The mailing lists have often been helpful for me. <Immodicus>the error gives it when looking in repositories <guixy>When was the last time we updated the installer? <g_bor[m]>Also you can rule out the network problem, by installing the bare-bones manually, as the installer media includes the closure of that confgiuration. <Immodicus>so I try again but if I try to restart the installer, the installer gives an error <guixy>The error on restart could be a separate bug. I have seen something like that before. <Immodicus>guixy so ,in your opinion, have to send a bug report ? <g_bor[m]>guixy: yes, actually there was a proposal to do a force reboot on installer restart. I still think that would be the proper way. <guixy>I think it's a good idea to send a bug report. <g_bor[m]>The problem is related to that some mounted filesystems cannot be safely unmounted. <Immodicus>i can't use IRC ahahahah. sorry for the stupid errors <guixy>There's a guix matrix channel. You might find that easier to use. <valignatev>Hm, I just encountered another weird thing about cargo-build-system. There is already packaged rust-clap-2 in crates-io.scm that I'm using as a #:cargo-inputs for my package. But when I'm trying to build it, cargo-build-system complains that it can't find dependencies of clap-2. AFAIK these dependencies are packaged in guix already. <valignatev>If I'm starting to add these dependencies explicitly to already existing clap definition it seems to work, but I wonder why clap-2 was packaged without explicitly specifying cargo dependencies in the first place? Is there some flaky cargo-build-system feature that suppose to resolve cargo dependencies automatically as long as they are packaged but <valignatev>I also see that there's some overhauling work going on to improve cargo-build-system, so maybe it's better to just wait and see <civodul>rekado_: re mumi, looking at /proc/39559/fd, it seems that it's leaking socket file descriptors <civodul>so maybe it's the web server module that forgets to close client file desc. <bandali>civodul, hey, thanks for the feedback on my emacs-next patch. if you prefer to not change the package definition for emacs until next release, i can probably get away without touching it, but there'll be code duplication <civodul>bandali: it's not that i do not want, it's just that it would trigger a huge number of rebuilds <civodul>and thus it wouldn't be able to go to master <civodul>i guess it's ok to have some duplication here in the meantime <civodul>we just need to add a comment with an arrow pointing to the one thing that's different :-) <civodul>so that we can more easily merge both in the future <bandali>understood :) i'll probably add a ;; comment to emacs's definition with what needs to be changed later <g_bor[m]>civodul: I just had a look at mumi, It seems to me that guile (web server) is used without anything fancy happening... <g_bor[m]>Did anyone observe leaking fds in other guile web projects, or should the handlers be inspected? <guixy>guix lint says clojure can be upgraded. If I upgrade it, should I include it in the same patch as the wrapper? <g_bor[m]>guixy: I would keep them separate, and do the upgrade first. Wdyt? <guixy>I already have the wrapper working. <guixy>I'll see what I can do about upgrading it. If it needs additional inputs, I want to ignore it. <g_bor[m]>If you have the wrapper working it is fine even without the upgrade. <guixy>Should I put them in one patch series? <g_bor[m]>guixy: I believe you should, as they are ment to be appied on top of one another. It's easier to review it that way. <g_bor[m]>I go to have some sleep now. Good night!