IRC channel logs

2019-02-11.log

back to list of logs

<nand1>looks like civodul's FOSDEM talk has been uploaded
<nand1> https://fosdem.org/2019/schedule/event/gnu_guix_new_approach_to_software_distribution/
<Formbi>hello
<Formbi>how to start a vm-image?
<Formbi>I created one and qemu says «qemu-system-x86_64: -net user: Block node is read-only»
<Formbi>ah, nevermind
<Formbi>I found it
<ZombieChicken>Hello. I was wondering if the documentation for setting up nontrivial filesystem layouts had been improved or not. Last I saw it was almost useless.
<Levy[m]>how was it last time?
<ZombieChicken>seemed like a bunch of disjointed data scattered around the website as I recall. Kind of hoping there is an actual guide or something now
<Tv`>is there an easy way to get the config.scm that is used by the default installation (what the vm image download is)
<atw`>section 3.7 of the manual references https://git.savannah.gnu.org/cgit/guix.git/tree/gnu/system/install.scm?h=v0.16.0#n327
***atw` is now known as atw
<mange>I think the VM image download is different. Looking at the Makefile I think it's https://git.savannah.gnu.org/cgit/guix.git/tree/gnu/system/examples/vm-image.tmpl
<Tv`>ahhh i postponed reading about channels
<Tv`>well it ain't https://git.savannah.gnu.org/cgit/guix.git/tree/gnu/system/examples/vm-image.tmpl based on /etc/issue contents
<Tv`>and the other one pretends to be in Paris, so that doesn't sound right either
<Tv`>:-/
<atw>ooh, VM image, missed that part
<mange>Tv`: what do you mean by /etc/issue contents?
<mange>I just downloaded the VM image and the issue is "This is the GNU system. Welcome." Looking through the source, that's exactly what I would expect from that vm-image.tmpl.
<Tv`>mange: the link has a longer message
<Tv`>though now you have me doubting whether I saw the longer message somewhere
<mange>Tv`: What was the message you saw? Have you reconfigured the VM image at all?
<Tv`>mange: i get just the first line
<Tv`>and all i've really done is "guix package -i tmux"
<Tv`>oh huh the rest displays after login, but it's not in /etc
<mange>Yeah, it's the pam motd, not /etc/issue.
<mange>You should see the message I quoted (with a preceding newline) in /etc/issue.
<Tv`>yeah, that's tere
<Tv`>oh yeah i see the motd plugin
<Tv`>in pam. okay so https://git.savannah.gnu.org/cgit/guix.git/tree/gnu/system/examples/vm-image.tmpl it is
<Tv`>thanks!
<mange>No worries!
<Tv`>hrm so i started `guix system container config.scm` based on that file, and it's building _a lot_
<Tv`>i would have expected my store to already contain everything
<mange>Is the version of guix the same?
<Tv`>it's the vm image that i'm running that in
<Tv`>and as far as i understand, i upgraded the vm image before starting this
<Tv`>guix pull etc
<mange>If you ran guix pull then you've update guix. That may have changed the package definitions, which requires things to be built (or their substitutes fetched).
<mange>*updated
<Tv`>i think i'm missing something from my understanding; i would have expected guix pull and whatever that other command was that i've already forgotten to have brought my /gnu/store to contain "the latest stuff" already
<Tv`>and if the config.scm is the same, it shouldn't ask for more
<mange>"guix pull" is like "apt update". It doesn't actually update your system/profile, it just updates the package definitions. Then you have to "guix system reconfigure" or "guix package -u" to upgrade your system/profile respectively.
<Tv`>guix package -u is the one i meant and forgot
<Tv`>couldn't have done reconfigure before, because i didn't have a config file to use
<mange>Then "guix package -u" only updates your user profile, not your system. If you're doing "guix system container" then it's likely that doesn't share many packages in common with your profile.
<fps>hi again. i got no answer previously. maybe now someone knows. guix pull fails, i guess while building python: http://paste.debian.net/1066645/
<fps>i tried to guix pull occasionally over the last weeks and it always failed..
<mange>fps: That's... strange. I wouldn't expect guix pull to need python. Can you confirm that the exact command you are running is "guix pull"? Can you also tell me what version of guix you're running by running "guix describe"?
<fps>mange: guix describe: http://paste.debian.net/1066858/
<fps>and yes, the command i ran was "guix pull"
<fps>guix pull rebuilds all kinds of stuff including gcc, python, yada yada
<Tv`>mange: ah ok so i'm paying for the delta of the vmimage vs latest of the system packages
<Tv`>mange: btw on a freshly booted vm image, guix describe failed
<fps>marusich: one thought was that there was a change to a different substitute url in the last few weeks/months and maybe my guix version is too old to use it..
<fps>marusich: oops, sorry, that was for mange
<mange>It wouldn't be too old, just misconfigured. I would still hope that it would be able to build everything it needs locally, but it probably does make sense to get it to fetch substitutes. Check out "(guix) Substitutes" in the manual.
<fps>mange: this is a guixsd system where i never touched the configuration manually. i will check the manual though for the new susbstitute url
<mange>An older installation won't have the new substitute server's key authorised, and also won't be configured to ask it for substitutes. See https://www.gnu.org/software/guix/manual/en/html_node/Substitute-Server-Authorization.html for details of how to authorise https://ci.guix.info/ as a substitute server.
<Tv`>similarly, my guixsd vm (fresh from the 0.16 image) is wanting to build qemu on its own (and the poor vm is running out of ram)
<Tv`>actually --dry-run claims nothing but downloads? that'd be good news.. but the guix system container from earlier definitely tried to build qemu
<Tv`>by --dry-run i mean guix build qemu --dry-run
<Tv`>perhaps there's two different "qemu" things, or something
<mange>Tv`: I don't know the details of the vm-image, but you might have to authorise/add the substitute servers, too.
<Tv`>mange: is there a way to see what's authorized? /etc/guix/acl has something in it already
<fps>what's the $prefix the manual mentions on a guixsd system?
<Tv`> https://www.gnu.org/software/guix/manual/en/html_node/Substitute-Server-Authorization.html#Substitute-Server-Authorization speaks of *prefix*/share but guixsd doesn't have the traditional /usr/share path at all
*Tv` glares at fps and adjusts tin foil hat
<fps>:)
<fps>i'll just do a find ;)
<mange>Tv`: I think that's how you see what's authorised, but I don't know where to find the keys for each of the servers to check what you have authorised. :/
<Tv`>fps: /run/current-system/profile
<fps>it's probably somewhere in the store..
<fps>ok
<mange>Ah ha! Here are the keys: http://git.savannah.gnu.org/cgit/guix.git/tree/etc/substitutes
<mange>Except that the ci.guix.info one doesn't seem to have any content...?
<Tv`>well i ran the recommended guix archive --authorize and /etc/guix/acl didn't noticably change
<mange>Ah, it's just the berlin.guixsd.org key. My local repo shows it as a symlink.
<Tv`>mange: it's a symlink to berlin
<Tv`>mange: the "context" is the target name
<Tv`>that git web frontend is a bit silly
<mange>So, Tv`, if you've authorised ci.guix.info you should be able to run "guix pull --substitute-urls=https://ci.guix.info" to definitely use substitutes.
<Tv`>mange: it seems like the vm image had ci.guix.info preauthorized, as far as i can tell
<fps>mange: after authorizing it it seems to use ci.guix.info without explicit --substitute-urls..
<fps>mange: guix pull worked now.. i still would have expected it to work without that though..
<Tv`>mine still wants to build a bunch, including the qemu that fails.. it's downloading some but apparently can't find substitutes for all
<Tv`>that guix pull is currently running, we'll see what happens after it's done (slow internet here)
<fps>btw: does every terminal in guixsd totally break on reverse history searches for anyone else. too?
<fps>[break as in similar to when you setup your $PS1 wronly and redraws are all over the place]
<mange>fps: I would have expected guix pull to work without substitutes, too. Substitutes just helped us to work around the fact that it wasn't building on your machine. I don't know why it wasn't building on your machine.
<fps>oh and i think it happens *after* running some guix commands. my hinch would be the progress bar implementation screwing up the terminal
<fps>mange: yeah.
<mange>Tv`: It can take the build machines a while to catch up when packages definitions are changed. Sometimes when you "guix pull" then try to reconfigure it will mean that you end up building some stuff yourself. The easiest solution is to just wait. Alternatively, you can pick an older commit and "guix pull --commit=" to it (in the hopes that the build machines have already built those packages).
<fps>hmm, interesting. during "guix package -u" bash segfaulted
<fps>and during "sudo -E guix system reconfigure /etc/config.scm" make segfaulted..
<fps>something is really odd on this box
<fps>is there a way to check the integrity of all installed packages? i mean short of guix publish'ing them and checking all hashes manually? :)
<Tv`>fps: there was a flag to guix gc
<fps>Tv`: ah thanks, --verify it seems
<Tv`>ruh roh guix gc just claimed to have freed 1.3 TB on a 20GB drive
<fps>sure, it's not 1.3GB?
<Tv`>"guix gc: freed 1,356.43359 MiBs"
<Tv`>and i'm dumb
<fps>:)
<Tv`>and/or this vm is using too tiny of a font
<Tv`>that's my cue to go sleep and let the builders catch up
<fps>guix gc --verify=contents
<fps>found nothing odd it seems.. hmm
<fps>it seems python segfaulted, too, during the guix pull without subs from ci.guix.info
<mange>fps: Reboot and hope for the best?
<fps>mange: won't work, but let me try it anyways ;)
<fps>nope :(
<fps>[ 61.837211] traps: bash[14656] general protection ip:7ffff7e921e0 sp:7fffffff8018 error:0 in libc-2.28.so[7ffff7e32000+13f000]
<civodul>Hello Guix!
<marusich>Hello, civodul
<civodul>hey marusich
<civodul>how's everything?
<marusich>It's alright. I have been trying for over 2 hours to get 'guix copy' to work, though, which has been a bit...frustrating
<marusich>If you're curious, I'd love a second opinion. I sent an email to help-guix@ just now about it. Not sure what the problem is :(
<marusich>Perhaps I'm overlooking something simple.
<marusich>I also still need to write up the summary of what we discussed in the session at Guix Days regarding porting mes to ARM. I was thinking of making it a small blog post, like Efraim's post.
<marusich>Anyway. I have to get some sleep or I'll be a zombie at work tomorrow. If you have any thoughts about "guix copy", please let me know! Goodnight o/
<efraim>I'm not convinced the guile-2.2 from make-bootstrap actually works as expected, two different versions segfaulted for me
<janneke>efraim: we're not using that, we're still on guile-2.0 for bootstrap-guile
<efraim>janneke: I know, I had almost everything set up for porting to powerpc-linux, so I figured I'd give it a shot
<janneke>i switched to guile-2.2 on wip-bootstrap -- but i think we agreed to make an effort to go back to guile-2.0
<janneke>ah, OK
<efraim>I might try downgrading it
<efraim>for anyone curious, here's the strace output: https://bpaste.net/raw/eed62039735d
<civodul>marusich: oh, 'guix copy' can be frustrating when it comes to error reporting...
<roptat>hi guix!
<civodul>hey roptat!
<civodul>efraim: that's on PPC right? could you try to run a non-stripped variant and get a backtrace?
***notnotdan` is now known as notnotdan
<eubarbosa>cool, building's progress "phase" is neat
<civodul>better than a lonely spinner :-)
<eubarbosa>hehe
<efraim>civodul: yeah, that's on powerpc
<efraim>i'll have to do some more testing with it and the non-stripped one
<efraim>bad news on the bug seeming more legit though, when I downgraded guile in make-bootstrap to 2.0.14, that one seems to work just fine
<raghavgururajan>I know this a far-fetched question, but just wanna ask. Any ETA on GuixSD v1.0 ?
<civodul>raghavgururajan: early March is the goal
<civodul>efraim:
<civodul>efraim: interesting thing is that this 2.2 bootstrap Guile also crashes on GNU/Hurd
<civodul>so maybe there's a real issue after all :-)
<civodul>well, for the cross-compiled one, because otherwise it works fine AFAIK
<efraim>Who was working on power8 support? I don't want to step on any toes if I go about testing bootstrapping guix
<efraim>I still need to find that gcc or gentoo architecture flags page to make sure I have any flags I need for ppc
<jonsger>efraim: me and Tobias Platen, but more on Power9
<efraim>jonsger: I assume its worth skipping POWER8
<efraim> https://wiki.gentoo.org/wiki/Safe_CFLAGS
<jonsger>efraim: yes, as all those Talos systems are POWER9
<g_bor>hello guix
<pkill9>hey
<lsl88>hi guix! sorry if i asked this here before, but i am trying to upload my gpg public key to mit gpg without success. I get a 502 or a 503. They key is fine.
<lsl88>this has been happening for days
<efraim>lsl88: iirc the mit server is down
<lsl88>efraim: yes, getting 503 several times and 502 kind of redirecting me to a simmilar message was "ok, i guess this is not my fault
<g_bor>lsl88: you can use the key server pool, that is mentioned in the docs.
<g_bor>it's at pool.sks-keyservers.net
<lsl88>g_bor: let me check the hacking file, but i guess i tried to use that one because it was the one mentioned.
<efraim>looking at https://gcc.gnu.org/onlinedocs/gcc/Floating-Types.html it looks like '-mfloat128' is for ppc32 and ppc64*
<htgoebel>I'm seeking somewhat up-to-date package definitions for WIP gnunet. (guix contains a quite old release only.) Any hints?
***rekado_ is now known as rekado
<rekado>htgoebel: we hoped that the GNUnet folks would make a new release.
<rekado>htgoebel: we were repeatedly told that a new release is just around the corner … but so far nothing new has been released.
<rekado>Hey Guixers, I’d be happy about comments on this: https://git.elephly.net/software/guile-studio.git
<rekado>It’s Emacs with some defaults for Guile + Geiser, CUA mode, tool-bar buttons for evaluating Guile buffers and looking up documentation, + support for the picture language.
<roptat>rekado, is the overdrive at my place being used by berlin now?
<rekado>no, not yet
<rekado>still need the firewall update
<roptat>ok
<civodul>¡hola lsl88!
<demotri>Will hydra.gnu.org evaluate a new master (with the merged staging)? I would expect some more undetected failures.
<zimoun>rekado: how to install the Guile Studio?
<rekado>zimoun: the README.org now says how: https://git.elephly.net/?p=software/guile-studio.git;a=blob;f=README.org;hb=HEAD
<rekado>(before installing it, though, please lower your expectations. There isn’t much to be seen yet.)
<htgoebel>rekado: This is why I asked for some WIP gnunet.
<htgoebel>I found the package description to be part of the gnunet.git (in the contrib folder)
<zimoun>rekado: ahah! you've just updated the README :-) Well, it is interresting
<civodul>hey zimoun!
<zimoun>Hi Ludo! I feel like a teenager to be back on IRC ;-)
<dustyweb>hello
<dustyweb>rekado: oh!!
<dustyweb>that's great to see your initiative on this :)
<dongcarl>Hey all, I'm back. Wondering how to make a graph for a package extending all the way down to bootstrap binaries?
<raghavgururajan>Urgent doubt! If I run "guix system reconfigure" after running "guix gc", will the packages removed by " guix gc" get downloaded and installed again by "guix system reconfigure"??
<raghavgururajan>Test
<civodul>raghavgururajan: things will be built/downloaded as needed
<civodul>dongcarl: you can use "guix graph -t bag", see https://www.gnu.org/software/guix/manual/en/html_node/Invoking-guix-graph.html
<raghavgururajan>civodul: So reconfigure won't reverse gc?
<raghavgururajan>Also, how to delete system generations?
<civodul>currently deleting system generations has to be done by hand: removing some of the /var/guix/profiles/system-*-link symlinks
<raghavgururajan>Okay.
<raghavgururajan>Thank you!
<raghavgururajan>Mainly, how to prevent "guix gc" from removing certain packages?
<civodul>it won't remove things that are "reachable", that is, packages that are in a profile, etc.
<raghavgururajan>In that I tried to install "grub" package make it available in profile. But phasing returned with error code 1?
<civodul>phasing?
<raghavgururajan>guix gc removed "grub.cfg" file from /gnu/store. I don't want that to happen.
<civodul>why?
<civodul>the general idea is that you don't need to worry about it
<civodul>just run "guix gc" when you need more space, that's it
<raghavgururajan>Yeah, "guix package -i grub" ran for a while, when it reached "building...grub.drv... 'check' phase". Then " failed to complete phase with error code 1".
<apteryx>raghavgururajan: so the grub test suite failed for some reason
<raghavgururajan>Because that's the grub.cfg file I am using to boot the system.
<civodul>raghavgururajan: no it's not; the grub.cfg file that's actually used cannot be gc'd
<raghavgururajan>When guix gc removed grub.cfg, my system couldn't boot from libreboot. Had to reinstall guixsd.
<raghavgururajan>It did
<raghavgururajan>I tested again
<apteryx>raghavgururajan: maybe you had deleted manually some system generations?
<apteryx>(the one actually in use)
<civodul>also /boot/grub/grub.cfg is not in the store, so "guix gc" won't touch it
<raghavgururajan>I ran "ls" command to list that grub.cfg at /gnu/store before running "guix gc". It listed the file. But after guix gc, ls command didn't list that files.
<raghavgururajan>*file
<civodul>the grub.cfg file in master doesn't matter
<raghavgururajan>I didn't install boot loader
<civodul>the one that GRUB uses is /boot/grub/grub.cfg
<raghavgururajan>I am using full disk encryption.
<raghavgururajan>so no /boot directory at all.
<raghavgururajan>I have been using the grub.cfg in gnu/store to boot my system from linreboot.
<civodul>i used full-disk encryption and i do have /boot, which GRUB stage1 decrypts
<civodul>but maybe there's something special with the Libreboot setup?
<raghavgururajan>Whoah! did you pass --no-bootloader while guix system init???
<rekado>there’s nothing special happening when using Libreboot. You just let the on-chip GRUB decrypt the disk and pass execution on to the on-disk GRUB. You still need to install GRUB.
<raghavgururajan>How?? For full disk encryption, there is only one partition for me. If I don't pass --no-bootloader, I am getting error that no efi partition found at the end of guix system init command.
<rekado>you probably don’t want to install in EFI mode then.
<rekado>what does your configuration file’s bootloader field look like?
<raghavgururajan>I tried bios too. error is no bios partition found.
<raghavgururajan>It looks same as in the guide.
<raghavgururajan>2nd example in the guide.
<raghavgururajan>desktop.scm
<raghavgururajan>I would paste the config if I am on desktop or laptop. I am just typing from phone. If you couldn't find, I cam try pasting here.
<raghavgururajan>let me know
<dongcarl>turns out it was a mistake running `guix graph -t bag` on `bitcoin-core`...
<dongcarl>Took 23.5 minutes to complete
<dongcarl>and `mupdf` is just failing to start
<rekado>it would be great if we could add a dynamic dependency browser that would display package connections at a given focus, allowing a user to move back and forth on graph edges.
<raghavgururajan>rekado: is there a way to install grub-efi, without making FAT fs partition, that mentioned in the guide?
<raghavgururajan>Like just one root partition with ext fs, grub will be installed at /boot/efi in that root partition?
<raghavgururajan>Test
<jonsger>oh https://www.pantherx.org/ seems to be an distribution of Guix
<mikadozero>Is anyone working on getting Guix SD to work on the free and open source RISC-V instruction set architecture?
<apteryx>mikadozero: I don't know of someone working on it
<dongcarl>Wondering how the graph for `bag-emerged` was generated here: https://www.gnu.org/software/guix/manual/en/html_node/Invoking-guix-graph.html
<dongcarl>When I invoke the same command the output PDF isn't as orderly
<marusich>jonsger, interesting! Thanks for sharing.
<nckx>mikadozero: I was (though just on qemu) but took a few months off last year.
<mikadozero>nckx, thank for working on it.
<mikadozero>nckx, what would you say the status is for getting Guix SD on RISC-V?
<Elon_Satoshi>my computer has locked up while building qemu-minimal-2.10.2
<lfam>Elon_Satoshi: Completely unresponsive? What did it print before it stopped responding?
<lfam>And what kind of CPU do you have? How much RAM? What kind of storage device?
<lfam>It's possible that it's just working very slowly
<Elon_Satoshi>it says it's in the middle of a 'check' phase
<Elon_Satoshi>It's not just working very slowly because I can't switch to another tty or turn on capslock or ctrl-C
<Elon_Satoshi>Can I resume the installation after shutting down and rebooting?
<lfam>It will start from the beginning
<lfam>I have successfully built GRUB yesterday on a 10-year old 64-bit Intel CPU platform
<lfam>Without knowing anything about your computer, my guess is that it's swapping
<lfam>On a spinning disk for me, this tends to cause the computer to be almost completely unresponsive until the work is done
<lfam>This is why I asked about your computer
<Elon_Satoshi>My computer is a lenovo thinpad t400
<lfam>Okay, that's similar to the one I mentioned (x200s). It will simply take a long time to do certain things that are computationally expensive or require a lot of memory
<lfam>Mine has 3 GB of RAM and that is enough for me to not use swap
<Elon_Satoshi>Oh? Mine had like 7 point something gigs ram so I added 8gb swap
<lfam>Wow, I doubt it was swapping then
<lfam>Well, maybe... Anyways, I would try again and just be patient. Maybe go out for a bit :)
<lfam>Ideally you wouldn't need to build this program but it seems that binaries are not available at the moment
<Elon_Satoshi>nobody thinks compiling everything is a good idea ._.
<Elon_Satoshi>lfam: Are the binaries PGP signed?
<lfam>Elon_Satoshi: No, they are signed with a simpler method that is built into Guix
<Elon_Satoshi>Oh
<lfam>When I said that binaries are not available, I meant for qemu-minimal-2.10.2, not for everything
<Elon_Satoshi>Well I'm installing guixSD with --no-substitutes
<Elon_Satoshi>I want to have a system that's fully built locally so I can guix challenge it all
<lfam>Okay
<lfam>It will take some time :)
<Elon_Satoshi>yes
<mikadozero>lfam, I like the idea of --no-substitutes was something that helped make Guix attractive to me.
<Elon_Satoshi>I hope the fact that I hit ctrl c, tried to switch to another tty, and hit capslock a few times doesn't come back to bite me too hard when this thing is less busy
<Elon_Satoshi>And if it's still locked up at qemu-minimal-2.10.2 I think I'll just reboot and install the thing with binaries :/
<lfam>It would be helpful to do `guix build qemu-minimal-2.10.2` and observe the output of the build process so that you have a better idea of what it's doing
<Elon_Satoshi>Will do, if my computer ever stops being frozen
<lfam>Yes :/
<Elon_Satoshi>Why is guix a "transactional" package manager?
<lfam>Elon_Satoshi: Do you mean "why would somebody want that" or "how does Guix implement transactions"?
<Elon_Satoshi>The latter
<Elon_Satoshi>Actually both
*Elon_Satoshi wants to learn about guix
<lfam>"Traditional" distros like Debian and Red Hat update the system by changing the installed software "in place", where the software is deployed. If the process is interrupted for some reason, the system is left in an inconsistent state that may or may not work. The package management tools on those distros cannot recover from such a state and the systems are basically broken at that point
<Elon_Satoshi>Ah. But Guix installs software somewhere else, in its own environment, and then simply changes the symlinks
<lfam>Guix (and Nix) prepare the updated software in another part of the system. They create a set of symlinks to the updated versions in what you could consider a "staging" area. In the final step, a single symlink is updated, and this single symlink is what makes the updated system available for use. We call these sets of symlinks "profiles"
<lfam>Yes, exactly :)
<lfam>By reducing the change to a single step, it's possible to reason about the state of the system in a transactional way. Either it was updated, or not. There is no possibility of an in-between state
<lfam>And then that one symlink's history can be tracked, and this gives us the ability to keep a history of previous states and change between them freely
<Elon_Satoshi>When I hear the word transaction, I think money. What does transaction mean in this context?
<Elon_Satoshi>Ooh
<lfam>It means the change was either performed or not. Similar to how your bank would like to know whether or not the money is in your account, and not allow the money to exist in two different accounts at once
<Elon_Satoshi>Ah
<lfam>When the money is transferred between accounts, you want to model the transfer as a single thing
<emacsen>Elon_Satoshi, no, transactional has specific meaning in eg database world. It's like the term atoic
<lfam>The recommended answer to this question is reasonably concise: https://softwareengineering.stackexchange.com/questions/289139/what-is-a-transaction
<lfam>Anyways, the point for Guix and Nix is that we aim to eliminate a common failure mode of older GNU / Linux distros
<mikadozero>I am getting an error when building qemu-minimal-2-10-2 during the check phase. The log the error mentions for tests/qom-test: FAIL can not allocate memory broken pipe. The system has 1GB of memory.
<Elon_Satoshi>qemu minimal again? :o
<Elon_Satoshi>mikadozero: I am having a problem that is very possibly related. My computer has frozen up during the check phase of qemu-minimal-2.10.2. My system has 7.something GB of RAM and 8GB swap
<mikadozero>I created a /swapfile in an encrypeted luks partition following the installation guide. What should I have in my config.scm to make sure it is using the swapfile?
<Elon_Satoshi>mikadozero: (swap-devices '("/swapfile"))
<Elon_Satoshi>at least I think
<Elon_Satoshi>I'm still in the process of installing a system using a config.scm that has that line, so I'll let you know if that works
<mikadozero>The size of the swapfile is 10GB.
<Elon_Satoshi>Neat
<lfam>In order to use the swapfile during installation I believe you need to set it up by hand
<Elon_Satoshi>yes
<Elon_Satoshi>just type swapon /swapfile
<Elon_Satoshi>Make sure you did mkswap /swapfile or it's not a swap file, it's an empty file that isn't set up for swap and won't work now or during the installation
<mikadozero>lfam, I did set up the swapfile by hand during the installation following the instructions in the install guide.
<Elon_Satoshi>mikadozero: what's the output of `free`?
<mikadozero>Elon_Satoshi, swap total 0B
<Elon_Satoshi>mikadozero: try running `swapon /swapfile`
<Elon_Satoshi>if it fails let us know what the error is, if it runs successfully run `free` again and see if there's any swapo
<Elon_Satoshi>swap*
<Elon_Satoshi>hey adamnr
<mikadozero>Elon_Satoshi, did swapon /swapfile now free says swap total 9Gi free 9Gi
<apteryx>mikadozero: you are probably better off waiting for substitutes to be available or configuring a beefier "guix offload" machine
*mikadozero trying reconfigure again now that swap is on
<Elon_Satoshi>mikadozero: careful
<Elon_Satoshi>mikadozero: your computer might get frozen up like mine
<Elon_Satoshi>and i'm on almost 16gb total memory
<Elon_Satoshi>apteryx: if we are to wait for substitutes we have to edit config.scm so that it doesn't install qemu-minimal-2.10.20, how do we do that
<Elon_Satoshi>is it part of base-packages or desktop-services?
<lfam>Elon_Satoshi: It's used to run the test suite of GRUB
<Elon_Satoshi>oh crap does that mean the only way to avoid qemu-minimal is to not install grub?
<lfam>Not exactly but it's starting to get complicated. Basically, you would need to create a new GRUB package called, for example, grub-no-tests, and use that instead of grub
<Elon_Satoshi>oh
<lfam>I'm looking in to the issues with qemu-minimal-2.10 now
<Elon_Satoshi>what about using an earlier version of grub, that uses an earlier version of qemu-minimal for which there is a substitute, or even will compile without locking up a thinkpad?
<Elon_Satoshi>okay
<Elon_Satoshi>if my computer unfreezes maybe i can share my qemu binary with those in need
<lfam>There are binaries available on the <https://mirror.hydra.gnu.org> build farm
<mikadozero>lfam, I am not using the flag --no-substitutes right now but it is still building qemu-simple,
<Elon_Satoshi>mikadozero: close everything important and wait a few days
<efraim>guix-publish on bayfront seems to have died
<mikadozero>Elon_Satoshi, the only thing running on that computer is Guix reconfiguring itself.
<kmicu>jonsger: 😮
<mikadozero>Elon_Satoshi, the reconfigure finished for me now that I have the swap on.
<lfam>Urgh, trying to build runc the source download fails like this:
<lfam>Throw to key `gnutls-error' with args `(#<gnutls-error-enum Resource temporarily unavailable, try again.> read_from_session_record_port)'.
<Elon_Satoshi>mikadozero: wow
<lfam>It works to use curl or wget outside of Guix
<Elon_Satoshi>my computer's still locked up
<mbakke>lfam: I wonder if the Guix snapshot needs to be updated to include e4ee84202633636b4c8cef4a332f0c74912a3b23.
<lfam>mbakke: Hm...
<lfam>It appears that not all connections to Github fail. We can build Git checkouts, but downloading tarballs is not working
*lfam tries it
<lfam>mbakke: In what context would I need to use this new Guix? The guix-daemon?
<mikadozero>Elon_Satoshi, thanks swapon worked and (swap-devices '("/swapfile")) also works.
<mbakke>lfam: Yes, I suspect the daemon needs a Guix with that commit. Not sure how, though.
<lfam>mbakke: I will just try updating my root profile and going from there
<lfam>It's on a foreign distro
<lfam>And yes, the server is offering TLS 1.3
<mbakke>I wonder why Berlin is not evaluating core-updates: https://berlin.guixsd.org/jobset/core-updates-core-updates
<lfam>In the meantime I can cheat with `guix download`
<Elon_Satoshi>you cheater!
<lfam>It's important to know how to cheat ;)
<Elon_Satoshi>i used cheat engine to cheat once
<Levy[m]>and now here you are
<Formbi>could there be a package with only info manual for Guix?
<Formbi>installing the whole Guix package (I mean «guix package -i guix») is a bit problematic
<rekado>when you have “guix” as a command you also have the info manual
<rekado>you don’t need to install it again to get access to the manual
<lfam>Sigh, the 'coalesce-sources' Guix test failed
<Formbi>rekado: well, not really
<Formbi> https://termbin.com/dybj
<rekado>how did you install Guix?
<rekado>you may need to set an environment variable.
<Formbi>btw I use Arch
<Formbi>/usr/local/share/info:/usr/share/info:/home/formbi/.guix-profile/share/info:/usr/local/share/info:/usr/share/info:/home/formbi/.guix-profile/share/info:/usr/local/share/info:/usr/share/info:/home/formbi/.guix-profile/share/info
<Formbi>there aren't any info files in my profile
<rekado>what does “which guix” tell you?
<rekado>I still don’t know how you installed Guix
<rekado>when it’s installed the usual way (binary installation method, guix pull) it’s unrelated to your default profile.
<rekado>it’s installed at ~/.config/guix/current/
<rekado>that’s where you should find info files.
<Formbi>first I used the aur package, but I removed the remnants of it (/usr/bin/guix and so on) and also used guix pull many many times
<rekado>you should have ~/.config/guix/current/share/info/ then
<Formbi>ah, now I see it
<Formbi>but Guix didn't suggest me to add this to INFOPATH
<Formbi>and actually why this isn't in ~/.guix-profile/share/info ?
<Elon_Satoshi>mikadozero: have you finished installing qemu-minimal yet?
<rekado>Formbi: it’s not in .guix-profile because “guix pull” doesn’t use .guix-profile.
<rekado>“guix pull” has it’s own profile.
<rekado>that’s on purpose.
<Formbi>mhm
<Formbi>thanks
<rekado>this allows you to roll back Guix or use an older version of Guix without having it impact your installed software.
<rekado>Guix should have suggested adding this to INFOPATH.
<Elon_Satoshi>oh yeah you did
<mikadozero>Elon_Satoshi, yes see above at 14:16
<Formbi>how to use EXWM on GuixSD?
<Formbi>nevermind
<lfam>Hi civodul!
<civodul>hello lfam!
<civodul>wazup? :-)
<lfam>Chillin
<lfam>;)
<lfam>I sent a patch to update the Guix snapshot because tarball downloads are failing from Github due to failed TLS 1.3 negotiation
<lfam>I tested it on my machine. Is this sort of thing safe to go ahead and push?
<civodul>oh i thought we had fixed it (or at least worked around it)
*civodul looks
<civodul>hmm is it on guix-devel?
<lfam>civodul: I hid it in the bug tracker: <https://bugs.gnu.org/34444>
<lfam>I think the issue is that the change did not yet end up somewhere that guix-daemon could use it, although I don't fully understand
<civodul>lfam: oh indeed, good catch!
<civodul>so yes, we need to update guix-daemon to get the fix
<civodul>er, the 'guix' package
<lfam>Right
*Elon_Satoshi gives up on using --no-substitutes, reboots and runs guix pull && logout