IRC channel logs


back to list of logs

<Gamayun>Hmm... I'm trying to offload building to another machine. GUILE_LOAD_PATH seems to be set, and the guix modules seem to be in the load path, but whenever I run 'guix offload test', I get the error message about guile modules not found. Any ideas?
<Gamayun>^ rebooting fixed it...
<luciddreamzzz>I defined my own list but I get a traceback... Any tips on debug this or no
<luciddreamzzz>could it maybe have to do with %base-services and %desktop-services both being listed
<luciddreamzzz>well in this case %my-services and the desktop ones...
<luciddreamzzz>i will try modifying desktop list but I see nowhere explicitly saying desktop includes base, I guess I can assume that hmm...
<luciddreamzzz>ok yeah never mind graph clearly shows bas services i try that
<lfam>luciddreamzzz: What does the traceback say? Did you make sure to add '(services %my-services)'?
<luciddreamzzz>i was going to re-run to paste traceback... I don't know how to read it
<luciddreamzzz>but then i thought it might be what I mentioned
<lfam>Also, I think the value of substitute-urls should be like this: (substitute-urls '(""))
<lfam>I didn't see %desktop-services in your paste
<luciddreamzzz>yeah its not giveing traceback that fixed i guess
<luciddreamzzz>that is where i copy paste from
<lfam>Yeah, that should work. Also make sure to authorize the key of
<luciddreamzzz>yep i did that
<luciddreamzzz>i think i see what you are saying..
<luciddreamzzz>I wonder if since i am defining the list the syntax is a bit different?
<luciddreamzzz>thank you lfam. and now the servers are playing nicely too :)
<lfam>Great :)
<luciddreamzzz>Installation finished. No error reported.
<luciddreamzzz>after i ask i remember thinking %desktop-services and %base-services could not be used together, that they must bump into each other somehow
<lfam>%desktop-services includes %base-services, so you don't need to use both of them
<luciddreamzzz>ok cool I wasn't picking up on that at first :)
***jonsger1 is now known as jonsger
<g_bor>hello guix!
<g_bor>sneek: botsnack!
<g_bor>sneek: later tell roptat: Thanks, it seems that the .so is generated from the .jsa, so fixing that will solve this.
<sneek>Got it.
<roptat>hi guix!
<sneek>roptat, you have 1 message.
<sneek>roptat, g_bor says: Thanks, it seems that the .so is generated from the .jsa, so fixing that will solve this.
<roptat>g_bor: cool :)
<g_bor>roptat: I was also thinking about getting rid of strip-jar-timestamps and strip-zip-timestamps...
<g_bor>They are similar in idea to strip-nondeterminism in debian, and a preferable solution would be to make the generation tool obey source-date-epoch.
<roptat>sure, but how hard would that be?
<g_bor>actually I don't know, but I don't think it is too difficult, once we find were this is done...
<g_bor>It would be nice, as we could try to get these patches upstream, benefiting everyone.
<tune>anyone know why recently my music playing in mpd gets quieter when I open something else that will have sounds? e.g. a video in mpv, or a webpage with a video on it
<tune>I'm wondering if it's some weird pulse behavior. I don't like it
<tune>it happens even if the video is paused and not making actual sound, probably hasn't been happening for more than a week
<kmicu>nckx: nah, we can throw tables but we can also throw flowers on the stage or lambdas to Guix folks. It was a sign of appreciation.
<kmicu>(Also GRUB finally shows up so ‘Houston, we have a progress’.)
<XWill1>I reconfigured my system with xf86-input-wacom but i found that the wacom config file and modulepath are not in the xserver.conf, what is the right way to see the install log of this package?
<jlicht>hey guix!
<roptat>XWill1: how did you add xf86-input-wacom?
<XWill1>Hi roptat, i add it to my config.scm and then run guix system reconfigure ...
<roptat>have you seen this: ?
<roptat>there is a #:modules argument to xorg-configuration-file, that should be what you want I think
<XWill1>no... i'll read that
<XWill1>ok! thank you, i had no clue where to start
<roptat>in the example about the keymap, instead of #:extra-config (list ...) you want to use #:modules (list xf86-input-wacom %default-xorg-modules)
<roptat>hum, #:modules (cons xf86-input-wacom %default-xorg-modules)
<roptat>(cons, not list, sorry)
<roptat>and of course you don't need to define bepo-evdev
<XWill1>Thank you, i'll try this right now
<roptat>and you don't need the package in your packages declaration
<XWill1>ok i put ;;
<kristofer>good morning! is returning 504 currently
<roptat>kristofer: that's because hydra is down for maintenance
<roptat>if you want substitutes, you'll have to use berlin for now (--substitute-urls='')
<ethanrjones97[m]>Has anyone been successful in using gnunet, bittorrent, IPFS, etc. for substitutes yet?
<roptat>ethanrjones97[m]: I'm working on implementing substitutes via bittorrent
<roptat>but you still need a substitute server to know what to download
<ethanrjones97[m]>ropat: very cool!
<ethanrjones97[m]>would DHT not work?
<roptat>you still need to know what you're looking for
<roptat>DHT indexes content by their infohash, which is derived from the content itself. Unless someone knows what that content is and tells you what the infohash is, you can't know that value in advance
<roptat>so you need a substitute server to build the substitute, compute the infohash and give you the info
<ethanrjones97[m]>ropat: Oh yeah, true.
<ethanrjones97[m]>ropat regardless it sounds like it would free-up a lot of bandwidth.
<roptat>yes, especially for big packages
<roptat>so right now I'm implementing a guile library for bittorrent, and I can download the ubuntu iso (that's the most standard thing I could find)
<nckx>kmicu: Ah! Great news.
<roptat>now I need to implement the dht, encryption and actually listen for incoming connections
<roptat>then some glue code to add that to guix itself
<jlicht>roptat: wonderful! Are you calling out to an existing C-based torrent library?
<roptat>jlicht: no, it's all guile code
<roptat>but maybe I should use an existing C library (I tried to look at libtorrent, but it's C++ and it has a big dependency graph)
<ethanrjones97[m]>ropat: all guile? very interesting
<ethanrjones97[m]>can I can a link to the repo?
<ethanrjones97[m]>*can I have
<roptat>I'm limited by my sha1 implementation: it takes 1s to compute the sha1 of a 512KB block
<jlicht>is there no C library to do that for you, at least?
<jlicht>thanks for the link by the way
<roptat>gnutls can compute a sha1, yes
<roptat>also it's a bit of a personal challenge, but I've already cheatted for sha1, so I'll use gnutls' implementation
<ethanrjones97[m]>ropat: very cool, I'll have to dig into the code. I'm not too good with guile yet, but I'm hoping to learn
<roptat>ethanrjones97[m]: it's probably not best quality code ;)
<jlicht>roptat: existing code is almost always better than perfect code that does not exist yet
<nckx>The first rule of free software: always apologise about how bad your code is.
<efraim>The second is ask for patches
<g_bor>hello guix!
<g_bor>I have this old bug report,
<g_bor>I've just came across this again.
<g_bor>I believe this might also be security problem, so I would be glad if someone could have a look.
<roptat>what are these keys used for?
<roptat>and how is it used in guix?
<g_bor>Unfortunately I don't really know. As far as I know shishi is a kerberos 5 implementation...
<g_bor>shishi.keys seems to be a kerberos server key.
<g_bor>What I feel most problematic about this, is that the same key can get to a lot of different places with the substitutes of this package, which obviously is not intended.
<roptat>you're right of course
<roptat>what does debian do?
<roptat>they don't seem to install that file. Can we add a phase, or is there an option, to remove that file from our package?
<kristofer>can guix system container have their own network interfaces like a vm does?
<roptat>kristofer: yes, but manually
<g_bor>roptat: I will have a closer look at that.
<roptat>kristofer: you have to find a PID from the herd inside the container, then run something like "ip link add type veth peer ns <pid>" which adds a veth pair, one in the current namespace and one in the namespace of the container
<roptat>then you have to connect that veth to your actual network device somehow and assign ip adresses manually too
<kristofer>roptat: the container in question should have a static ip
<kristofer>roptat: I'm trying to decide which is better. I think a container would be more suitable on my somewhat antiquated hardware
<roptat>you can assign IPs from inside the container
<roptat>or outside the container: "ip netns exec <pid> ip a add veth0" or something
<roptat>the veth connects the container to the host, then it's up to you to decide how it's connected to the internet
<roptat>you can enable ip forwarding and use iptables to masquerade internal IPs, or you can use a bridge with brctl
<pkill9>so guix containers are safe to use as sandboxes?
<nckx>Is it just me or is issues.g.i more down than up? (504)
<nckx>snape: Handy site. Thanks!
<snape>nckx: now it's up :/
<snape>but when I sent the link it was down
<nckx>Another similar site: ‘The remote name could not be resolved: ''.’ But that sounds bogus. All NSes return the right IP.
<nckx>Now it loads but ‘recent activity’ is from 1 October.
***daviid is now known as Guest19351
***Guest19351 is now known as daviid
<g_bor>hello guix!
<g_bor>something stange happened on my guixsd.
<g_bor>It seems that libgcrypt-1.8.3 with the hash starting hc5... is not in my store any more. Guix ERROR-s with 'In procedure dynamic-link' no such file.
<g_bor>I've checked my store for a working guix-command, but unfortunately all I have depends on this.
<g_bor>Any idea?
<nckx>g_bor: What I would do (because I like using a hammer for absolutely anything) is re-mount the store link an existing libgcrypt-1.8.3 directory to the missing hc5... one.
<nckx>If all you need is to get a working guix back.
<nckx>I've done that once.
<g_bor>nckx: nice, thanks, I will try that.
<g_bor>One more thing, do you have any idea hot to find out what happened?
<nckx>I have /gnu/store/hc5cak3fj0dijbm86kpz2asl7ld4gf8y-libgcrypt-1.8.3 if you'd really need it.
<nckx>g_bor: No, I'm as surprised as you.
<g_bor>I have another version, so I will try that :)
<nckx>We're talking about a properly installed (i.e. not local git ./pre-inst-env'd) guix, right?
<g_bor>yes, absolutely, but a pre-inst-env one is also not working, same error.
<g_bor>This is somewhere so early, that issuing guix without any arguments fail.
<g_bor>I have ran gui gc earlier today, but not problem showed then.
<nckx>Then I really have no idea how the reference scanner allowed libgcrypt to be GC'd. It's either something so trivial that it didn't occur to ask, or a very grave bug. I'd guess the former for now :-)
<g_bor>Then did a pre-inst-env guix build --rounds=2 first, which failed because of a non-reproducible package.
<g_bor>After that guix did not work...
<g_bor>Actually I've tried each and every guix-command in my store, and all of them needs this version...
<g_bor>I will give a try to the remount trick :)
<nckx>g_bor: Mutating the store is a big no-no to some, but it's often the easiest way to get out of such a deep mess 'if you know what you're doing'. And don't we all think we do?
*nckx → AFK, good luck.
<g_bor>The funny thing is that actually everything is working, but guix.
<mbakke>g_bor: Do you have another machine around so you can `guix copy` it?
<mbakke>Do you have the .drv so you can rebuild it?
<mbakke>g_bor: Is `guix` itself broken, or just `./pre-inst-env`?
<nckx>mbakke: How would I find out which .drv would build the missing item in this case? (I mean, I'd use grep, but there's prolly a cleaner way.)
<g_bor>mbakke, actually it seems my system is more than broken than I thought...
<mbakke>nckx: `guix gc --derivers /gnu/store/abc123-foo`
<g_bor>emacs misses
<nckx>mbakke: Thanks. I wouldn't have looked under gc.
<g_bor>I have a really old vm image somewhere, I will have a look.
<mbakke>g_bor: Would be good to know how these obviously live store items were deleted. Did you run `guix gc` while /home or /var was not mounted or something like that?
<g_bor>I am digging the logs now, maybe I can find something interesting.
<g_bor>I tried what nxkc said, using that trick I get a partially working guix.
<g_bor>For example lint works.
<g_bor>But pull does not.
<ethanrjones97[m]>Any tips on how to fix that libgcrypt hash mismatch on a install?
<g_bor>I would be not too said, if I had to ditch this vw image, but really interested in what happened :)
<XWill1>What is the right way to find the path of an installed package from the config.scm (ex: /gnu/store/...-kmod/bin for modprobe)
<g_bor>wow, I am sleepy, my letters got confused :)
<g_bor>good night
<nckx>Good night.
<XWill1>good night
<nckx>(XWill1: You're not being ignored; I also wonder what the right way would be.)
<pkill9>XWill1: I use `realpath $(which [command])`
<pkill9>e.g. `realpath $(which modprobe)`
<XWill1>thank you, it's fine, i know it can take some time to get an answer
<nckx>pkill9: Right, but that's a bit of an indirect hack (it's what I'd do, though).
*nckx shrugs.
<XWill1>thank you pkill9, it works, but i'd like a guile/guix command, i do not understand gexp, i tried #$kmod with use-modules (gnu package linux)
<pkill9>do you mean how to refer to the package inside config.scm?
<XWill1>yes nchx, because the path could change after a guix system reconfigure
<XWill1>yes pkill(
<XWill1>yes pkill9
***NB0X-Matt-CA is now known as nonlinear
<pkill9>XWill1: I'm not sure what you want, but my config.scm might have examples, particularly lines 165 or 263:
<XWill1>ok thank you
<XWill1>It should be close to the line 165, but with a string-append, nice config.scm btw
<rk4>pkill9: out of curiousity, what system do you use that config.scm on? [thinkpad?]
<pkill9>yeah, thinkpad x240
<luciddreamzzz>what is a good browser if looking to use substitutes?
<luciddreamzzz>guessing icecat will take a while :3