IRC channel logs


back to list of logs

<lfam>I'm going to start converting all the uses of git-reference to match what's recommended in the manual. I guess it could go in core-updates.
<rain1>I have a question about writing a package defintion
<rain1>I think that I've managed to get it to do everything perfect except I still need to pass LD_LIBRARY_PATH:
<rain1>Is there something I can inside the package definition itself to inform guix about that lib64 directory?
<lfam>rain1: Did you grep in gnu/packages for LD_LIBRARY_PATH? You will find some inspiration there
<rain1>I did grep for that yesterday, but let me try again and study it more carefully!
<rain1>ah there's a post install wrap-program thing I could do
<rain1>that sounds quite fitting so i'll try it out
<rain1>although thats on gnu-build-system
<rain1>I added a wrap-programs phase and I can see that it really happened
<rain1>but i still get the error
<rain1>(wrap-program "midori" `("LD_LIBRARY_PATH" ":" prefix ("lib64"))
<rain1>does this look reasonable?
<lfam>rain1: It might be hard-coded somewhere. I'd use `strings` on the binary and grep the source code
<janneke>rain1: aren't you missing an `out' in there?
<janneke>something like (string-append out "lib64")
<rain1>oops, yes I am!
<janneke>rain1: have you checked the generated wrapper?
<rain1>I thought 'prefix' was taking the place of 'out'
<janneke>yeah, stubled over that last week
<rain1>linux mint was hacked
<rain1>(via wordpress)
<Jookia>ah, at least GPG signing picked up the hack, yeah?
<rain1>maybe a lesson to learn from it is to use static sites as much as possible
<rain1>they seem to be really really keen on MD5 hashes
<rain1>so I don't know if there is any GPG involved
<lfam>Somebody asked me for a link to a Guix presentation. Which video should I send them?
<Jookia>It's so great that I was working on GRUB patches then accidentally garbage collected my store after removing some temproots I thought were the problem'
<lfam>Ah, never garbage collect the store until you have no works in progress ;)
<Jookia>so i get to compile guix again
<Jookia>on my dual core t400
<lfam>Maybe it's time to take a break ;)
<Jookia>i've been at it for a few days, mainly testing the patches
<Jookia>testing, rebasing, cleaning :)
<lfam>A good use case for FS snapshots IMO
<rain1>this was my attempt to package midori
<rain1>I still haven't been able to get the LD_ stuff going
<lfam>rain1: Without trying it, I wonder how it fails? I usually have to build strings so that "lib64" would be "/lib64"
<lfam>Typically "out" does not include the trailing slash
<rain1>hm this is a bit frustrating
<rain1>I'm trying to test our your change
<rain1>but guix package -r midori ; guix gc ; then applying the edit and installing it
<rain1>it just succeeds instantly without rebuilding anything
<rain1>may I rm the 8 different midori things I have in /gnu/store/?
<lfam>rain1: That means that nothing has changed
<lfam>No, never do writing or deleting in the store.
<rain1>I wish they weren't there though
<davexunit>what does it matter?
<davexunit>you can gc them manually if needed and nothing else refers to them
<rain1>they're troubling me a bit since I'm struggling to make my package work and I don't want anything interfering
<lfam>If the installation is returning instantly, then the package definition you are building with is generating one of those 8 versions
<lfam>rain1: I'll try your package definition after my current compile.
<davexunit>rain1: it's almost certainly not interfering
<rain1>i suppose adding the / helped but wasn't enough then?
<rain1>ill look into the gnu packages more to try to understand better
<lfam>rain1: Can I suggest a change to your workflow? I think it will make things a little easier to test. Rather than using `guix package -i foo` and `guix package -r foo` to test your changes, use `guix environment --ad-hoc foo`. That will build foo and put it on your PATH, but it won't register foo against garbage collection.
<rain1>oh yeah I definitely should be doing that, i'll try it now
<lfam>You just need to ctrl-d out of the environment shell to get back to normal
<rain1>thanks very much for all the good advice!
<lfam>When you are done
<lfam>Alright, I'm downloading the midori dependencies. Might take a few minutes
<rain1>it's kind of you to test! cheers
<lfam>rain1: What is zeitgeist?
<rain1>I don't really know but it gives me the creeps
<lfam>Ha, okay. But how did you choose to disable it?
<rain1>I just when I try to install a program and it pulls in zeitgeist
<rain1>and a whole slew of KDE dependencies along with it
<rain1>I just hate*
<rain1>I think it's something 99% of people don't know is on their computer, or what it does
<rain1>(and would not miss it)
<lfam>I've never heard of it and don't have real experience with KDE.
<rain1>It would be really easy to enable it though, just removing the obvious line
<lfam>I'm going to reenable it while testing
<rain1>oh okay!
<lfam>rain1: I looked at the nixpkgs definition for midori. I noticed that one of our regulars contributed to that file in the past: izysong. Perhaps izysong can help you :)
<lfam>I will say I think their configure script is broken. Setting the prefix as instructed (with --prefix) has a bizarre result: CMake Error: The source directory "/tmp/guix-build-midori-0.5.11.drv-0/build/--prefix=/gnu/store/3g4n0n5d1vf2bkkc0sz5n2gm4vdw60d0-midori-0.5.11" does not exist.
<lfam>Howdy paroneayea
<paroneayea>hi lfam
<paroneayea>lfam: hope to get the rest of your last few comments on the mediagoblin deps addressed tonight
<paroneayea>then I think I can push 'em, yeah?
<lfam>Yeah, I think so. Time to get on to the Javascript deps?
<paroneayea>lfam: I'm going to procrastinate on that one :)
<paroneayea>I actually think I need to talk to davexunit about what to do to see if we can hit something more interim
<davexunit>I think we cheat a bit for now.
<davexunit>I don't know exactly how mediagoblin does things, but if we can substitute anything that uses bower/npm with the upstream unminified source files for now, that would be OK.
<paroneayea>davexunit: make a jquery-cheater package? :)
<paroneayea>davexunit: hm
<paroneayea>I'm not sure jquery will work without being "compiled"
<paroneayea>in the unminified steps
<davexunit>last time I looked at the jquery website they had both minified and unminified versions for download.
<paroneayea>if we could just run unminified js though yeah, that would be fine by me
<paroneayea>davexunit: ah ok
<paroneayea>yeah I'll look again
<paroneayea>davexunit: well
<paroneayea>davexunit: I think I saw the same, but they still somehow took a bunch of files
<davexunit>a minifier as a build step is an optional thing.
<paroneayea>and packed them into one?
<davexunit>but that's not a big deal.
<paroneayea>yeah I guess we could do that
<davexunit>it's still the source code.
<davexunit>unless they did other things to it.
<paroneayea>jquery and the css files are the only things we "need" to get a v0 version of mediagoblin packaging in guix
<lfam>I can't believe what I'm reading ;)
<paroneayea>I'd like to have the full "proper" deps, which also includes pdf.js and video.js, etc
<paroneayea>lfam: hm? :)
<Jookia>lfam: Believe it!
<davexunit>perhaps those are also available in unminified form?
<paroneayea>yeah maybe
<paroneayea>as an aside, it'll be interesting of guix is the first distro to have mediagoblin packaged for it.
<davexunit>this would be enough to start with. we can iterate from there.
<lfam>I'm being silly. I had misread that davexunit was suggesting we used minified js. In any case, I know that packaging mediagoblin will be a progressive endeavour
<paroneayea>brb! hitting the shower, then back to packaging
<rain1>I think there is a problem with the pidgin-otr package? does anyone use it?
<rain1>i installed pidgin and it, but the otr plugin isn't showing up
<rain1>cp /gnu/store/v6h...rid-pidgin-otr-4.0.1/lib/pidgin/ ~/purple/plugins/
<rain1>this makes it work
<rain1>sorry ~/.purple/plugins/
<Daneel_Olivaw>Hi. My installation of guixsd completed with no errors. I rebooted. Now I'm getting grub rescue saying that /grub/i386-pc/normal.mod is not found. Help?
<rain1>I don't know if this will help but I have that file in: /boot/grub/i386-pc/normal.mod
<Daneel_Olivaw>I'll move it.
<rain1>(maybe something was pointed at / instead of /boot?)
<Jookia>Do you ahve a separate /boot ?
<CompanionCube>one random question that may be useful in the future: is it possible to rollback a guixsd install/system without using grub?
<Daneel_Olivaw>I have three partitions. A swap. A /mnt/boot and a /mnt
<Jookia>Daneel_Olivaw: Hmm. /mnt is where / is mounted, yes?
<Jookia>Where was the normal.mod originally
<Daneel_Olivaw>I have a /boot/grub/i386-pc/normal.mod. Should I copy it to /grub/i386-pc/normal.mod
<Daneel_Olivaw>I never moved it.
<rain1>I don't think it should be moved
<Jookia>Could you paste your grub.cfg, the output of the 'mount' command and your system config?
<Daneel_Olivaw>I can't. This computer has no other OS on it. All I can really do is upload pictures.
<Daneel_Olivaw>Fixed it.
<Jookia>Ooh, how?
<Daneel_Olivaw>It's gets to the grub os selection now. I select the OS. And it can't find the command initrd linux and search
<Daneel_Olivaw>I had my boot partition mounted on /mnt/boot. Should have been /boot
<Daneel_Olivaw>But it still won't boot because of these missing commands.
<Daneel_Olivaw>It's weird. These are the boot commands and every one of the three is missing.
<rain1>maybe the boot partition was mounted wrong during install too?
<rain1>and it didn't manage to install those commands
<rain1>should I report the pidgin thing or something?
<rain1>my mistake
<rain1>I'm supposed to add export PURPLE_PLUGIN_PATH
<rain1>but I didn't notice since I installed from inside emacs
<paroneayea>all mediagoblin python deps now pushed to guix master \\o/
<davexunit>paroneayea: awesome!
<xd1le>so trying to create bootable usb, is it normal for filesystem to be ext4?
<Jookia>For Guix I think it is as it's not an ISO but a disk image
<xd1le>because the startup says "no bootable device found"
<xd1le>but it works with ISO's for example
<xd1le>(i don't know what i'm doing)
<Jookia>It uses GRUB as the bootloader rather than ISOLINUX
<xd1le>so i guess you're saying it's not going to work on my machine?
<Jookia>Hmm, is this the downloaded image>
<xd1le>Jookia: ?
<Jookia>Where did you get the bootable USB image from
<xd1le>followed the instructions, so xz -d that
<Jookia>Then you dd'd it to a USB?
<xd1le>running `file` on that decompressed file, it says it's a "DOS/MBR boot sector"
<xd1le>i did `dd ... && sync`
<Jookia>And have you fiddled with your BIOS to boot from USB
<xd1le>not sure how to do that?
<xd1le>i know i can boot from usb because i just did that with an ubuntu one
<xd1le>ah right, you mean the bios settings
<xd1le>*it can
<paroneayea> egads
<xd1le>ok let me see if it works on this machine..
<xd1le>shoot, forgot i can't try it right now
<xd1le>anyway, on that machine, it says "cannot find bootable device" even without the guix usb stick inserted
<xd1le>(so i probably stuffed it up somehow)
<xd1le>it's an old one i found
<Jookia>Some old machines (such as my eeepc) treat USB drives as hard drives, so you need to add them to the boot orders
<xd1le>but, putting an ubuntu usb stick in it, and it boots up into that
<xd1le>oh it's probably not old as in old, just old as in unused for a few years
<Jookia>Maybe the BIOS does some magic?
<xd1le>but not like 10 years old or anything i don't think
<xd1le>maybe, i have no idea... :(
<Jookia>Playing with the settings should be able to fix it, no sad frowning
<xd1le>sadly, i can't seem to find the keybinding to do that
<Jookia>paroneayea: No HTTPS? Fantastic
<xd1le>usually it's esc or f12 or something..
<Jookia>xd1le: del, esc, f1, f2, mash them all multiple times when booting
<Jookia>i think some machines have used insert to access bios which made me feel pretty bad
<xd1le>also they were using sha1 right?
<xd1le>okay lemme try
<xd1le>Jookia: is it possible to accidently delete this bios settings stuff?
<xd1le>or should it definitely be there?
<Jookia>xd1le: if by accidentally you mean with an SPI flasher yes
<Jookia>xd1le: otherwise no
<xd1le>nah, then i guess it should be there
<xd1le>omg i found it
<xd1le>and i changed to legacy bios instead of uefi whatever that means
<xd1le>and i have this distro "solydxk" on it, wtf
<Jookia>paroneayea: To add insult to injury they used a six character password: "upMint" for their DB
<xd1le>ugh now i have to find the key for the boot menu again
<xd1le>esc not working
<Jookia>xd1le: UEFI is not great
<xd1le>Jookia: wait what, was this some sort of injection hack?
<Jookia>xd1le: wordpress exploit -> backdoor implant -> root access
<xd1le>wow wordpress
<xd1le>this brings up an obvious question
<xd1le>does guix have gpg sigs?
<xd1le>(i should have checked for this today)
<Jookia>They're on the download page
<xd1le>ah yep, i see em
<xd1le>so like, mint does not?!
<xd1le>because then we could use them to verify
<paroneayea>xd1le: but whether guix should check sigs of packages too...
<paroneayea>xd1le: I just replied to an email thread about that :)
<xd1le>paroneayea: oh i know that, i mean about the images
<Jookia>mint has md5 hashes over a HTTPS connection, but in this case hashes wouldn't help as they could be modified
<xd1le>yeah, which is why gpg could help i think
<xd1le>to provide the authenticity
<Jookia>only if you verify the signature
<lfam>xd11e: you didn't check the sig when you installed? ;)
<xd1le>lfam: i haven't installed
<xd1le>Jookia: yeah true, as always
<lfam>Earlier today I printed the paper mentioned here:
<lfam>"The Update Framework"
<lfam>It discusses these problems and apparently offers a solution / mitigation
<lfam>But I haven't read it yet...
<xd1le>HAHAHA IT WORKS! :)
<xd1le>Jookia: ^
<Jookia>xd1le: Narly dude
<xd1le>now, how to check if wireless works..
<xd1le>docs say to use ifconfig
<xd1le>how can i do a wireless network scan?
<Jookia>do you have iwconfig
<xd1le>yes, on this boot image
<xd1le>as in, i have whatever is on the image, obviously
<Jookia>oh, i mean iwlist
<Jookia>iwlist wlan0 scan
<Jookia>replace wlan0 with your wifi card listed in iwconfig/ifconfig
<xd1le>yeah i just figured that out
<Jookia>do you have wifi?
<xd1le>and did it, "interface doesn't support scanning : Network is done"
<Jookia>lucky! also try 'iwlist wlan0 up' or iwconfig
<Jookia>i acn't remember
<xd1le>ah shoot
<xd1le>yup just did ifconfig wlp3s0b1 up
<xd1le>"missing free firmware (non-free firmware loading is disabled)"
<xd1le>ah well
<Jookia>wifi :(
<xd1le>do wired connections usually need nonfree firmware?
<Jookia>usually not
<xd1le>would the interface more likely be "enp4s0f0" or "lo" ?
<Jookia>lo = loopback w* = wifi, e* = eth
<xd1le>okay, it'll still be helpful for packaging stuff for guix if it works, but was hoping to use it on the go too..
<Jookia>USB dongles or replace the wifi chip :P
<xd1le>what are usb dongles?
<xd1le>and how to replace the chip?
<xd1le>is it easy?
<Jookia>depends if you can take apart a laptop or the machine- USB dongles are portable wifi things that you plug in and give you wifi
<xd1le>as in, do the usb dongles come with their own wireless cards or whatever?
<xd1le>i might research on replacing chips
<Jookia>No, dongles are wifi cards
<xd1le>okay i'll research that too
<Jookia>you plug them in to your USB and they show up as a wireless device you can use to connect
<Jookia> you can see a few here
<xd1le>the laptop has a lot of screws on the bottom, probably easy to take that apart at least
<xd1le>but it's a small laptop, so not sure if the sizes will be standard and everything
<Jookia>that also includes cards
<Jookia>note: i haven't bought from thinkpenguin
<Jookia>from what i can tell most laptops use the same wifi card slot
<Jookia>lspci could tell you what card you're on and you can look it up
<xd1le>Jookia: none of those say "dongle"?
<Jookia>oh, dongle's just the term for a wireless USB device since they stick out the side... it's actually kind of a terrible term, the proper term is 'adapter'
<xd1le>ah i see
<xd1le>so all those ones that say "usb adapter" is what you mean
<xd1le>so as in, i can just plug that in
<lfam>I want to make a function git-version that will generate a version string out using the variables 'commit' and 'revision', as shown in the julia package.
<Jookia>xd1le: and you get wireless support
<lfam>So that the package version field can just be (version git-version)
<xd1le>and it will act like anoter wireless card, or..?
<Jookia>it'll act like another wireless card- you can connect to networks, it does what wifi does
<xd1le>as in, you know how my current wireless card on it doesn't work?
<xd1le>this could replace that, or is it something else?
<lfam>xd11e: It's a replacement card
<Jookia>'replace' as in, put it inside your machine or using it
<xd1le>replace as in using obviously
<Jookia>then yeah
<xd1le>because it's a usb connection
<lfam>You can do networking over USB
<xd1le>wow that's so cool
<xd1le>so does think penguin work with linux-libre?
<Jookia>lfam: My computer supports ethernet/debug/networking over USB, it's cool
<Jookia>yes, from what i know all their products work with linux-libre
<lfam>I think any USB should support it with the right software support
<Jookia>Not so fast, there are terrible USB ones with firmware
<Jookia>I've had bad experiences with USB wifi adapters that need kernel patches to stay alive and use proprietary firmware
<lfam>My question about this git-version idea is how does one refer to variables in that situation?
<lfam>Maybe I should ask on #guile
<lfam>Oh, my example is libuv-julia, not julia
<xd1le>Jookia: okay i'm probably going to get one of those, cool! :)
<xd1le>hey doesn't that mean it could work on any laptop
<Jookia>xd1le: Well, weigh it against replacing the internal card- you'll lose a USB port (though it's probably more secure)
<Jookia>What do you mean?
<xd1le>as in, on any laptop that requires nonfree firmware for the wireless card
<xd1le>as long as i don't use it at the same time ofc
<Jookia>Yep, just insert it and it'll work
<Jookia>I think you can use two cards at once (if you had two wireless networks) if you're up for advanced routing
<xd1le>it has 3 usb ports so it should be good
<Jookia>cool then'
<xd1le>but a lot of oem laptop models these days only have like 2, with which, replacing the internal one would be the better option, yeah
<Jookia>the average amount of (free) USB ports I have on my machines are two
<Jookia>Though I say this as the machine I'm on has two free ports in a hub since I have three things plugged in and it has two ports
<xd1le>hey do you know of a command to check battery life?
<Jookia>Hmm, nope
<xd1le>ah nws
<lfam>You can dig out of the sysfs
<lfam>I mean, dig itout
<lfam>dig it out
<Jookia>you did it
<lfam>Look in /sys/class/power_supply
<Jookia>you won't beat my bad spelling though, no matter how hard you try
<lfam>Figure out which power supply is your battery, and then do `echo $(((energy_now * 100) / energy_full_design))`
<lfam>Or something like that.
<xd1le>instead of energy it says "charge", but yes, thanks!
<lfam>Don't use that command line
<lfam>But that's the idea
<xd1le>i can just do `mkfs.ext4 /dev/sda3` to format a partition right?
<xd1le>i.e. to delete everything there
<xd1le>sorry for these noob ot questions
<xd1le>also, how do i "mount the target root partition under /mnt" ?
<xd1le>hmm... how do i shutdown
<xd1le>(i know with systemd the command is `shutdown`...)
<xd1le>sneek: later tell civodul i think the guixsd image should have vi in it, what do you think?
<sneek>Will do.
<Jookia>xd1le: 'halt'
<Jookia>xd1le: and yeah
<xd1le>Jookia: thanks
<xd1le>yeah to mkfs.ext4?
<Jookia>xd1le: thats how you format a partiton, and you do mount /dev/sda3 /mnt to put it under mount
<xd1le>Jookia: thank you
<Jookia>not sure if its included but you can use cfdisk or parted to do partitioning too
<Jookia>(parted if definitely included)
<xd1le>fdisk is included
<Jookia>oh, i've never used fdisk
<xd1le>which i have some experienc with
<Jookia>cool then
<xd1le>i've also used gparted, which i assume is the graphical version of parted..
<Jookia>i don'
<Jookia>i don't know if it is
<Jookia>they're very different interfaces (command line vs GUI)
<xd1le>ah, well i'll hopefully figure it out
<roelj>I'm running into an error with loading Guile modules when compiling Guix:
<roelj>Am I missing a Guile package?
<rekado>xd1le: the image already includes a couple of editors IIRC; among them are zile and nano(?)
<rekado>roelj: is this Guix modified or plain from git?
<roelj>rekado: Plain from git
<roelj>git clone ...; ./bootstrap; ./configure CFLAGS="-Wall -Wextra -O2"; make -j4
<efraim>I add guile, guile-json, gnutls and libgcrypt when I run make
<rekado>this looks like a syntax error
<efraim>don't think you need make -j4, it's currently multi-threaded within the one guile thread
<efraim>so that may introduce new race conditions
<rekado>updating my clone now to see if I can reproduce this
<roelj>My master is at 1034331a112097e5fdbc2bb73b288443ad85148a. That should be the latest..?
<rekado>I just updated and ran "make" in "guix environment guix".
<rekado>I do not get an error.
<roelj>Could it be a gettext problem?
<roelj>Because on the line it errors, it uses "_", which I guess is something for translations..
<roelj>rekado: Thanks for checking :)
<roelj>I think I've found the issue. I was using Guile 2.2 from the guile repository.
<xd1le>rekado: yeah i know, but they don't work with brains accustomed to vi-style keybindings:)
<roelj>Can git tags be used as versions in Guix?
<roelj>Ideally, (commit <commit-id>) => (tag <tag-name>)
<amz3>héllo, I read civodul is not against using another P2P stack to distribute guix
<amz3>I think IPFS is a good candidate for this task, is there anyone that knows about it?
<efraim>never heard of it, do they have a website with more information?
<efraim>another option I can think of would be tahoe-lafs
<xd1le>i've heard about it at least, so i'm assuming many more here have also heard of it
<swedebugia>xd1le try 'upower --dump' to see the percentage left.
<efraim>looks familiar now that I see the website
<efraim>built from go, so me/codemac/others should get more on top of finishing packaging it if we want to try it ou
<xd1le>swedebugia: thanks
<xd1le>anyone know how to make 'ifconfig' available for root?
<xd1le>i.e. under which package ifconfig is in?
<NiAsterisk>iproute2 iirc
<NiAsterisk>wait. no
<amz3>never heard of tahoe-lafs
<NiAsterisk>but iproute2 is a replacement, xd1le
<xd1le>hmm, i just realised
<xd1le>i need ifconfig or something to get internet
<xd1le>but i need internet to get ifconfig or something
<rekado>xd1le: you can always do "guix package -i vi" to get the editor you want.
<xd1le>rekado: yeah i know but still
<rekado>ifconfig is part of the image afaik
<xd1le>but ifconfig is there in my user account
<xd1le>but it says "command not found" when i run it under root
<xd1le>(by doing `su`)
<rekado>as root you could run /home/xd1le/.guix-profile/sbin/ifconfig or similar
<xd1le>and apparently i need to run it under root
<xd1le>because otherwise it says operation not permitted
<xd1le>rekado: thanks, i'll try that
<xd1le>there's no .guix-profile
<xd1le>this is under guixsd
<xd1le>so.. where are the paths to the profiles?
<xd1le>what's that?
<rekado>/var/guix/profiles/per-user etc
<xd1le>okay thanks
<efraim>its a distributed file store
<rekado>xd1le: root could also just install it, no?
<xd1le>rekado: no internet
<xd1le>wait, can it?
<rekado>also: isn't it part of the system profile as per /etc/config.scm?
<xd1le>i did the barebones one
<xd1le>so it has like tcpdump iirc and one other package
<xd1le>and like the base system packages
<rekado>xd1le: how did your user profile end up with ifconfig?
<xd1le>i don't know?
<xd1le>it was on the boot image
<xd1le>usb boot image
<xd1le>as root
<rekado>xd1le: you wrote: "<xd1le> but ifconfig is there in my user account"
<rekado>can't you just locate it and then run it as root?
<xd1le>ah okay figured it out
<xd1le>sorry it's in another location bc it has to have a wired connection, so takes time..
<xd1le>so root *does* have ifconfig and everything
<xd1le>and i just got internet
<xd1le>but for e.g.
<xd1le>if i log in as user
<xd1le>and then run `su`
<xd1le>then there's no ifconfig
<xd1le>but there's also no ls
<xd1le>soo... possibly a much bigger issue...
<xd1le>i should try sudo ls...
<xd1le>i tried locating the ifconfig
<xd1le>but there's nothing under /var/guix/profiles/per-user/...
<xd1le>like i did ls -a
<xd1le>do you think i did installation wrong?
<xd1le>(installation said no errors reported or something similar like that though)
<xd1le>okay sudo ls works
<NiAsterisk>did you just setup guixsd?
<xd1le>so i can still use sudo until/if i can fix su
<xd1le>NiAsterisk: how do you mean?
<xd1le>yeah, this was my first guixsd install
<NiAsterisk>applications are profile based.. basetools like ls should be available to every profile if you used desktop-profile (or what it was called)
<NiAsterisk>so if you install iproute2 as $user, then it's on your system, but it's not in every profile, only of those who do guix package -i iproute2
<xd1le>NiAsterisk: i used the barebones you see here:
<xd1le>with a few modifications
<xd1le>(it comes with the usb boot image)
<xd1le>NiAsterisk: yeah i know that
<alezost>xd1le: system packages are in /run/current-system/profile, so ifconfig is here: /run/current-system/profile/bin/ifconfig
<xd1le>btw, should i install applications to root or user, or my choice?
<xd1le>also, where should my user specific .scm go?
<xd1le>er, i should just reread the documentation again..
<NiAsterisk>location of the config.scm for the system does not matter
<xd1le>ah right, because it's specified on the cli
<NiAsterisk>I prefer to have as little as possible under root user. it's your choice, your system.
<xd1le>cool, thanks
<xd1le>so to be clear, root does have ls, ifconfig etc. as expected
<xd1le>but when i run `su` from my user account to run commands as root
<xd1le>ls isn't even there
<NiAsterisk>do you run su or su -
<xd1le>what's su -
<NiAsterisk>eh... hm.. i use it for so long i don't even know what's the exact difference.
<NiAsterisk>right now at least
<xd1le>ah, it's just makes it a login shell
<xd1le>maybe that makes a difference though..
<NiAsterisk>that. right
<xd1le>yup! it works!
<xd1le>NiAsterisk: thanks!
<xd1le>i guess that just means that just `su` doesn't source the bash login file, which has the appropriate $PATH's or whatever
<xd1le>but obviously an interactive login shell does
<NiAsterisk>hrm. as long as we don#t have a browser equivalent to torbrowser: why does setting socks5 to 9050 not work for proxy in icecat?
<swedebugia>amz3: thanks for the link to IPFS!
<NiAsterisk>it's just not easy searching for solutions when the browser is named almost like an animal
<NiAsterisk>getting more results for cats and snow than what I really want
<efraim>in iceweasel I set remoteDNS also
<efraim>I also send my traffic through privoxy
<amz3>I think does a good job at buzz word bingo, but doesn't describe yet what it provides. AFAIK it's possible to have a client library push data in the IPFS an retrieve it
<NiAsterisk>okay, I never used privoxy.. maybe I should use that. but I am used to just using torrc
<amz3>they take a great time dealing with the underlying infrastructure/code to explain that it's a really versatile platform. A problem remains is that it linked to some bitcoin-based tech which might be a problem
<NiAsterisk>does it use the blockchain? because blockchain ages and scales terribly
<NiAsterisk>ipfs is on the list of things I need to read into
<NiAsterisk>for some time. other papers take time
<amz3>NiAsterisk: the main features don't require a block-chain
<NiAsterisk>is there no privoxy service for system/services/networking.scm (needed at all)?
<xd1le>yeah privoxy is nice, definitely should be packaged by guix if it isn't already
<xd1le>it's to search for icecat things in general, in my experience
<NiAsterisk>it is packaged.
<xd1le>tor browser should also be packaged ofc
<NiAsterisk>it can't because firefox has parts which are not compatible, but we are soon trying to work around that.
<xd1le>oh so there's just no service
<xd1le>ah nice
<xd1le>(that you are working on it)
<NiAsterisk>well, Jookia.. and I did package torbrowser before (inofficially) for gentoo. the best would be to try and get in contact with torproject to find out why they keep the non-free parts.
<NiAsterisk>so it will not happen now. but sometime this year
<xd1le>ah i see, that's unfortunate
<xd1le>perhaps it's a byproduct of torbrowser deriving from firefox
<xd1le>which has nonfree parts, i think?
<NiAsterisk>which is what I said some moment ago :)
<xd1le>ah right
<xd1le>i'm so tired rn -_-
<roelj>Is there a mirror:// for github too?
<alezost>roelj: no
<NiAsterisk>efraim: do you use privoxy on guixsd? how do you configure it without a shepperd service? write one yourself?
<efraim>NiAsterisk: I'm still on debian
<NiAsterisk>ah, right.
<NiAsterisk>i used to use tordns, but that's (currently) not an option when you try out and look into gnunet and related things. guess i'll have to figure it out myself to make it work.
<xd1le>i wonder if it's possible to have inter galactic filesystem
<vimuser>the universe is a file
<vimuser>generated by math
<NiAsterisk>yesterday I found a filesystem which claims to be scaleable AND use the blockchain. I have doubts, but i think i'll discuss it with another project.
<NiAsterisk>well not scaleable the way like for example bitcoin "scales"
<xd1le>inb4 the universe is a computer simulation and the theory of everything is basically, "everything is a file" (not a string)
<NiAsterisk>everything is in lisp.
<xd1le>ah so this is relevant:
<xd1le>(ignore the perl/python part at the end)
<xd1le>it's there i presume because it's from, i don't know where the original is from
<xd1le>ah ofc
<xd1le>i actually kid you not though
<xd1le>a few nights ago i dreamt where everyone was speaking in this weird simple language
<xd1le>but somehow you visualise it in literally s-expressions
<xd1le>and it's very easy to understand
<xd1le>that's what i remember
<xd1le>i just.. don't know
<NiAsterisk>cloning tor-browser.git takes its time. stil lfaster than xonotic
<efraim>I used to build firefox os for my nexus4
<efraim>that was a 6 hour build process, but downloading ~20GiB over git was also hours
<NiAsterisk>i had gentoo running on an 10 years old netbook with incredible small amount of ram. I was used to compiling, even with archlinux on that thing, I knew slow speeds from back when I started with ZenWalk and slackware, but this netbook... nope.
<xd1le>ugh try --depth 1 if you don't need all of it
<NiAsterisk>it was interesting to learn the difference i5 cpu and 16 gb ram vs centrino2 vs arm.. 45 min to 2 hours vs 16 hours..
<NiAsterisk>i know but i want it all
<xd1le>what type of crazy shit have they put in that repo to make it 20GiB...
<NiAsterisk>well it's an OS..
<xd1le>i mean, this is why i try to stick with only text files
<xd1le>i know it's not alway spossible
<xd1le>but the biggest repos i've ever gotten were like 2GiB
<xd1le>also it's supposed to be a smaller/simpler OS
<xd1le>i hope they're not bundling os images in their repo
<NiAsterisk>i don't know what size all relevant gentoo projects have.. i guess less than 20GiB. portage is small.
<efraim>I think there may have been a few blobs for drivers for different phones, but it ended up being like 135 submodules or something
<efraim>and I'm sure there was a fair amount of artwork
<xd1le>methinks it's just some build of up of non binary files
<xd1le>like yeah artwork
<xd1le>s/non binary/binary
<xd1le>although i have no idea how git handles binary files
<efraim>from watching some of joeyh's early git-annex videos, each blob is basically unique, and you can't really generate a diff
<xd1le>ah well there you go, i was speaking of optimization/compression or whatever to save disk space
<NiAsterisk>I will look more closely at torbrowser,firefox,iceweasel buildsets and sources in the next time to see if the non-free parts can be dealt with easy or can be patched and (best case scenario) be send to torproject, so that at build time there could be the option to remove those parts.
<NiAsterisk>on the topic of distributed storages, I thought it was only accessible in the hidden trac, but hit's also on the front page:
<NiAsterisk>in the trac (and soon on some wiki) are evaluations of those.
<calher>How do I change *just the hour* with date(1)?
<calher>Scared to do `date 022206'.
<calher>Dammit, no JWM.
<civodul>Hello Guix!
<sneek>Welcome back civodul, you have 1 message.
<sneek>civodul, xd1le says: i think the guixsd image should have vi in it, what do you think?
<calher>Actually, doesn't Zile have a mode that acts like vi?
<calher>Zile Implements Lua Editors.
<civodul>i don't know, i don't think so
<civodul>i think Zile died in an attempt to rewrite it in Lua
<civodul>probably time to push
<calher>That's probably broken, too.
<NiAsterisk>the image has zile and nano. isn't that enough?
<calher>civodul: I'm not going to try Guile on Zile because it's too hard to build stuff on Guix.
<civodul>calher: it should be easier to build stuff than elsewhere, should might want to share to misfortune on help-guix or some such!
<davexunit>glibc 2.23 is out
<wingo>yay rebuild :)
<davexunit>core-updates already has a few rebuild-the-world things in it
<davexunit>including at least one glibc rebuild triggering change.
<wingo>oh then we can merge the xorg update into it
<wingo>it works for me, fwiw
<wingo>It Works For Me™
<davexunit>core-updates is the place for this stuff
<calher>In urxvt --
<calher>£ screen -R
<calher>Cannot find terminfo entry for 'rxvt-unicode'.
<calher>Meh, using xterm.
<mark_weaver>it might be worthwhile to build the xorg updates in another branch, separate from core-updates, dunno.
<mark_weaver>maybe combined with the pulseaudio and gstreamer updates
<calher>I can't delete ~/.Xresourcees.
<mark_weaver>and libvpx. so many branches
<mark_weaver>civodul: do you think we should build the xorg/pulseaudio/gstreamer/libvpx updates in their own jobset, or should we do those in core-updates?
<civodul>mark_weaver: rather in their own jobset
<mark_weaver>sounds good, thanks!
<calher>Does the console use Terminus?
<efraim>OOM on the llvm-3.7.1 tests, failed 4 hours in
<efraim>I'll try again tonight
<mark_weaver>efraim: thanks for working on it
<efraim>pianobar magically built correctly
<keverets>there's not that much to pianobar, really. bit of web scraping and passing to libao
<keverets>still cool that it built correctly
<NiAsterisk>What's more likely: outdated website or outdated github readme? I don't get an answer from the developing team at the moment, but it's a difference between CMake and Cargo/Rust for building.
<wingo>mark_weaver: so are you handling the merge of wip-xorg-server-1.18 ?
<wingo>ACTION wants to get it on a train headed towards master
<mark_weaver>wingo: yes, I'll combine it with some other updates (pulseaudio, gstreamer, libvpx) on a branch and ask hydra to build it out.
<wingo>cool, tx
<wingo>choo choo :)
<efraim>for the audio updates, do gstreamer and friends need to be on the same version?
<mark_weaver>efraim: I'm not sure I understand your question.
<mark_weaver>it seems that all of our gstreamer packages will be updated to version 1.6.3
<efraim>gstreamer and the gst-plugins-good/ugly/libav
<efraim>yeah but it might be that we can cherry-pick some to update now and leave the others at their current version
<mark_weaver>hmm. which one(s) would you prefer that we avoid updating, and why?
<efraim>gst-plugins-good, they were the ones giving me the most problems
<mark_weaver>problems with version 1.6.3?
<mark_weaver>what kind of problems?
<efraim>i didn't try it extensively, but it had 1 or 2 tests it alternated in failing
<mark_weaver>efraim: I don't think that's a new problem. gst-plugins-good has been failing its tests non-deterministically since at least version 1.4.5
<mark_weaver>I have many failed build logs in my archives
<efraim>i see
<mark_weaver>wingo: btw, as a general rule, autoconf/autogen/autoreconf phases should go after unpack instead of before configure. the reason is that some of the phases before configure are needed especially on non-intel platforms.
<mark_weaver>some of those phases need to fix up the generated scripts, e.g. configure
<mark_weaver>wingo: also, sometimes you removed patches but forgot to remove them from dist_patch_DATA in
<mark_weaver>civodul: the package list doesn't seem to be updating correctly
<mark_weaver>the updater script ran last night, but the list on the website still says "Updated February 18, 2016"
<calher>query YstDawson
<civodul>mark_weaver: yes i noticed, but the weird thing is that the script itself works and "cvs ci" is successful
<civodul>so i wonder if it's not a more general problem with web pages
<civodul>maybe nully knows? :-)
<colonolGron>rekado, ?
<paroneayea>I try not to sign into anything google, but for legacy reasons I have an account
<paroneayea>I got a message that "firefox 38.0 on windows" signed onto my account and became alarmed
<paroneayea>turns out that's what Google thinks Icecat 38.6 is
<paroneayea>on guix
<mark_weaver>paroneayea: that's because IceCat spoofs the useragent to be Firefox on Windows:
<mark_weaver>by default, anyway. you can change it
<paroneayea>well no wonder thne!
<paroneayea>I was like, how did anyone brute force my randomly generated password in assword? It's super long? they must have compromised my machine...
<rekado>colonolGron: hi!
<rekado>I have to go now but might be online later tonight.
<lfam>How about a function 'git-version' that could generate version strings for git sources from 'commit' and 'revision' variables defined in the package definition?
<rekado>lfam: +1
<lfam>It's a long line of fiddly boilerplate to get that string right otherwise
<lfam>This would be my first Scheme function from scratch
<lfam>Would it be possible to define the function in, say, package.scm, and then have it use the variables defined in the package definition, as in pbtranscript-tofu?
<lfam>Or would those variables need to be defined in package.scm, or elsewhere?
<lfam>My terminology may be off...
<mark_weaver>lfam: scheme procedures can only access their own lexical environments, and what's passed in to them via arguments.
<lfam>Okay, I'm going to try banging my head against it for a while. I'm sure I'll be back with questions!
<Jookia>civodul is back \\o/
***zimmermann_ is now known as zimmermann
<roelj>What can I best do when the build files call "cc" directly instead of "gcc" or a CC variable? Should I patch the Makefile(s), or could I somehow set "cc" as a link to "gcc" (or alias, or ...)?
<Jookia>Doesn't gcc provide a 'cc' binary?
<wingo>meep meep
<roelj>Jookia: Apparently not.
<Jookia>wingo: o/
<wingo>greets :)
<mark_weaver>roelj: often it is sufficient to add "CC=gcc" to #:make-flags
<mark_weaver>hi wingo!
<calher>I'm going back to Trisquel.
<Jookia>roelj: I'd probably patch the build files then since they don't use CC
<Jookia>calher: -> #trisquel
<roelj>mark_weaver: Thanks, I will try that.
<mark_weaver>calher: okay, it's probably a better fit for you
<roelj>Jookia: Ok. I'm afraid that such patches will break because upstream is not going to care whether it breaks or not.
<Jookia>roelj: well, ask upstream to fix it as well. Submit the CC fix as a patch
<calher>Jookia: I think so.
<mark_weaver>roelj: I think it's preferable to just add "CC=gcc" to #:make-flags
<roelj>mark_weaver: I'm building again now, see if it works :)\\
<roelj>I certainly hope so
<roelj>/gnu/store/l7px210li6zviymgvp3cps6n48x7fgpl-bash-4.3.42/bin/bash: cc: command not found
<roelj>Too bad.
<mark_weaver>roelj: oh well
<roelj>So, can't I "alias cc=gcc" in the bash shell of the build process? :(
<NiAsterisk>why alias?
<roelj>So that invoking 'cc' will invoke 'gcc'?
<roelj>without changing the filesystem of the build..
<mark_weaver>roelj: there's no 'bash' running the build process
<NiAsterisk>" I think it's preferable to just add "CC=gcc" to #:make-flags" is meant to be added to the package definition.
<mark_weaver>there's guile running the build process, running 'make' which might occasionally run bash for individual commands, but maybe not even that.
<roelj>NiAsterisk: Yes, I've added it tot he package definition.
<mark_weaver>NiAsterisk: it seems that the Makefile runs 'cc' directly without using a CC variable
<mark_weaver>roelj: how many occurrences can you find of 'cc' being used directly?
<roelj>mark_weaver: I'm finding out now :)
<roelj>I must be looking in the wrong place, because I can find none directly
<a_e>mark_weaver: Hello, thanks for taking care of the multimedia updates on hydra!
<mark_weaver>a_e: no problem! It just started building
<a_e>Nice! I am a bit worried by the hundreds of not building packages. It is difficult to compare. Well, we have to make do! Hopefully this will go through well.
<roelj>Oh, I found something. the variable CC is not defined, but used in one of the Makefiles.
<a_e>calher: I just moved to GuixSD on my work machine over the weekend. I am rather thrilled by how well it is working. (Except that all change takes time, I am used to KDE but now need to learn Xfce). The up- and downgrade facilities are really pleasant. Also the on-the-fly tests of the configuration when doing a "guix system reconfigure", before a reboot would break everything.
<lfam>roelj: There is a basic example of how to work around that in the package definition of dvtm (gnu/packages/dvtm.scm)
<roelj>lfam: CC=gcc should fix that, right?
<lfam>I believe so
<davexunit>hmmm, Guix often lies to me about what files will be downloaded
<paroneayea>new glibc out patching multiple vulnerabilities
<paroneayea>not sure if any of them are critical or not
<davexunit>2.23 was released, yeah.
<a_e>davexunit: To me, it lies by omission. It works more than what it pretends to do!
<lfam>ACTION groan
<davexunit>a_e: yes, exactly.
<a_e>paroneayea: Maybe it does not matter so much for once.
<lfam>re: glibc
<paroneayea>yeah it might not matter, I'm not sure
<davexunit>a 2.23 update should go to core-updates, if it hasn't already
<a_e>There is already a core-updates branch that is being filled with patches, so we could build all this at the same time.
<davexunit>we are already rebuilding glibc there for at least one reason
<paroneayea>well at least one of them listed here is the one we arleady have patched
<paroneayea>* An integer overflow in hcreate and hcreate_r could lead to an
<paroneayea> out-of-bounds memory access. Reported by Szabolcs Nagy. (CVE-2015-8778)
<paroneayea>that's the only one that looks maybe worrying to me, but I haven't looked at how hard it is to trigger.
<mark_weaver>davexunit: yeah, glibc should definitely be updated in core-updates
<lfam>I don't want to try to cherry-pick upstream security updates of core components
<mark_weaver>civodul: btw, in case you don't know: we need to update our 'pkg-config' source URL to use https, because upstream now redirects to https. however, gnutls depends on pkg-config, so it's not clear how to fix this.
<mark_weaver>and more generally, there seems to be a widespread movement to deprecate http, so we'll probably need another plan for how use https sources for all packages, even those that gnutls depends on.
<paroneayea>gnutls bootstrapping? :)
<paroneayea>I guess it's not as "easy" as gcc
<paroneayea>this must be closer to coreutils as a requirement?
<davexunit>the answer may indeed be that gnutls is a bootstrap binary
<Jookia>maybe you might have to put gnutls in the bootstrap-
<davexunit>but I'd like to evaluate all other possibilities first
<mark_weaver>another option is to handle downloads differently, so that they can make use of resources outside of the build environment. since they are fixed-output derivations, the loss of purity here shouldn't matter, as long as we end up with the file we intended to download.
<calher>a_e: The upgrade facilities are *not* pleasant. I had to do stuff from a git repo.
<mark_weaver>https is not the only issue. a few months ago I wanted to use 'git-fetch' to download something that was needed for 'git'.
<a_e>calher: Yes, there are some rough edges there. I also switched to the git versions as soon as I could.
<davexunit>it's pretty pleasant to me.
<calher>system reconfigure didn't work, jsut jf;lksafjd.
<Jookia>mark_weaver: Perhaps impure downloads are the most useful idea here
<civodul>mark_weaver: yes i saw it and proposed a plan, but it needs more work
<civodul>we should open a bug report to keep track of it
<davexunit>I'm not big on the idea of making the downloads impure.
<mark_weaver>davexunit: well, consider this: suppose all downloads were done via a local proxy that runs outside of the build environment.
<davexunit>because the derivation no longer describes the process of downloading a file
<davexunit>it's missing things
<mark_weaver>you can consider that proxy to be part of the network, not part of the build.
<mark_weaver>and the rest of the network is *already* outside of the build environment.
<civodul>davexunit: derivations have the notion of "sources", which are files taken for granted
<davexunit>assuming that things are available on the host system is the reason I hate language package managers.
<davexunit>mark_weaver: a local proxy sounds interesting
<davexunit>but it's another daemon to run, etc.
<civodul>in the meantime, i wonder if we could come up with a minimal GnuTLS to address the issue
<wingo>probably easier to look for an http-hosted location for the pkg-config tarball :)
<civodul>right, we can also mirror it :-)
<davexunit>the downside there is that we have to mirror every tarball that is in the pkg-config dependency graph whose upstream URL uses https
<davexunit>there's only 1 now, but I expect that number to grow as time goes on.
<davexunit>I'm surprised we got on as long as we did without this problem, honestly.
<calher>"<mark_weaver> calher: okay, it's probably a better fit for you" --> "Thank God you're leaving. I'm tired of your negativity."
<wingo>calher: no need to be rude. guix is beta software and isn't ready for everyone.
<civodul>davexunit: i think we need to think about the short-term solution, and about the long-term solution, which are likely going to be different
<calher>wingo: I know mark_weaver is glad to see me leave.
<paroneayea>calher: we do have a code of conduct and ask people in the community to treat each other nicely
<paroneayea>indeed, it's a requirement for participating.
<wingo>was about to say the same thing
<calher>paroneayea: Does what you said translate to "We have a code of conduct and I'm about to threaten you with it"?
<civodul>calher: paroneayea says that the CoC is a requirement for participating
<paroneayea>calher: I don't believe that asking people to paritcipate nicely is a threat.
<paroneayea>nor requiring it.
<wingo>ACTION lost in a twisty maze of polkit
<civodul>wingo: glad you're looking into it, BTW :-)
<wingo>i don't know if i'll find the thing, but we'll see :)
<paroneayea>calher: anyway, yes it's a requirement. If you aren't willing to abide by treating people nicely in this space, then as goes the phrasing, "you don't have to go home, but you can't stay here".
<paroneayea>but if you're willing to be nice, then great
<calher>"be nice" turns into censorship really fast.
<Jookia>censorship of meanness is censorship i can get behind
<davexunit>it's not an issue of censorship
<calher>Jookia: People have to agree what 'mean' is, and that's tricky.
<Jookia>calher: Most codes of conducts include a list of things that are unacceptable, not sure if Guix's CoC includes it
<Jookia>Oh, it does
<paroneayea>calher: anyway it's pretty simple: the community has asked you to be nice and respectful of the people here
<paroneayea>it turns out that being rude and mean is another form of censorship: it ends up making it so many other people are afraid to participate
<davexunit>we don't need a CoC to say "practice common courtesy or you will be kicked"
<davexunit>the ops of this channel decide what is OK
<paroneayea>anyway, yep, "practice common courtesy or you will be kicked." It's up to you, calher, but I suspect you've been more than warned at this point.
<davexunit>we only recently got a CoC, which is a good formalization, but we didn't tolerate jerks in the channel prior.
<Jookia>calher: Speaking as a human being, I know you're upset because Guix isn't going that way and I assume you have some form of depression, but being angry and seeing everyone as an enemy or hurting you isn't going to make things better. It might make you feel justified or let you vent, but you need to find a better place than #guix for that.
<lfam>I'm interested in trying GNOME on GuixSD. Do I just need to add gnome to global packages and enable the desktop-services?
<wingo>lfam: so, yes to a first approximation
<davexunit>lfam: yup, that works.
<wingo>but there are some things that don't work afaiu
<lfam>Okay, I expect an approximation, not perfection :)
<davexunit>many services are still missing, but the shell works.
<davexunit>if you use a laptop, don't close the lid!!!
<lfam>Hey, it's QEMU
<wingo>davexunit: what happens when you close the lid?
<wingo>nothing i guess?
<davexunit>infinite suspend loop when you try to wake it up.
<wingo>why, i wonder
<davexunit>at least on my laptop
<davexunit>I'm not sure
<calher>Since I'm not good enough for you and you all hate me, I'm going to use human proxies to refine my messages because I don't know how to have a heart anymore.</ragequit>
<davexunit>this happens a few times a week
<paroneayea>well, we shouldn't have repeat behavior
<paroneayea>I think if they rejoin and do it again
<paroneayea>kick them
<paroneayea>and if they rejoin and do it again after that
<davexunit>enters channel, asks questions, gets angry that guix has rough edges, says something mean, we get mad, leaves.
<davexunit>I'm all for the ban hammer at this point
<davexunit>I'm not an op
<paroneayea>who is an op? civodul ?
<wingo>ACTION nominates davexunit as adjunct op
<mark_weaver>civodul and I have ops here, at least
<paroneayea>mark_weaver: civodul: I'd say, feel free to apply banhammer now, or the very next time this happens.
<paroneayea>there's no reason to waste everyone's time on this.
<paroneayea>anyway, on a happier note, I'm gonna get lunch!
<paroneayea>whooo lunch!
<mark_weaver>I'm not sure it's quite reached the point of deserving a ban, but if he gives up on Guix and uses something more finished and polished, that's fine with me, and I *do* think a polished system is a better fit for him.
<NiAsterisk>treating people nice, and therefore no freespeech (which includes hatespeech) is totaly okay. every discussion on "why" is wasted time for pleasant teamwork
<civodul>i would ban next time that happens
<mark_weaver>I won't object :)
<NiAsterisk>well, every discussion after the initial one.
<civodul>unless everyone think i'm naive :-)
<Jookia>SInce a whole bunch of devs are here, is it okay to add 'guix utils' to the VM building modules? It looks like nothing breaks when I did it and it gives a lot of useful functions
<davexunit>NiAsterisk: I'm very much a defender of free speech, including speech that some may find offensive, but people have to understand context. #guix isn't the place to let your pejoratives fly.
<NiAsterisk>okay. i think we just have to agree to disagree, as discussion on that is too offtopic for this space.
<davexunit>I think we mostly agree...
<lfam>It's just not productive to get mad here
<davexunit>yeah, exactly.
<Jookia>Signal to noise ratio is important
<lfam>Another reason not to get into a discussion about CoCs in general ;)
<davexunit>break over! back to work!
<davexunit>in this fantasy, you are assembling packages on an assembly line
<civodul>Jookia: (guix utils) is not supposed to be used on the "build side"
<lfam>davexunit: If only all packages were that easy
<davexunit>a sign hangs that say "0 days since the last accident"
<Jookia>civodul: Hmm, it has an interesting function I find useful 'make-symlinks'
<civodul>Jookia: ah, maybe we can do something then :-)
<lfam>How about a sign that says "x days since we last had to rebuild every package"? ;)
<Jookia>days since glibc rebuild
<davexunit>lfam: that's "accident" here ;)
<civodul>Jookia: you mean 'switch-symlinks'?
<civodul>isn't it enough to use 'symlink'?
<civodul>switch-symlinks is really when you need atomicity guarantees
<Jookia>Yes and no- in my patchset I've moved the generation of the GC root of grub.cfg to install-grub. It will error when there's already a symlink, and I also need it for more GRUB stuff in the patch after that one (yet to post yet)
<Jookia>No idea if the grub.cfg GC root should be installed regardless of GRUB's installation success, I'm doubting that for now
<Jookia>_hubertus_: Hey there
<_hubertus_>I try to use guix sd but bootable iso is not goog way for me
<_hubertus_>is there any chance to create bootable ISO?
<Jookia>_hubertus_: Are you on a system with EFI?
<_hubertus_>normaly I use Fedora
<Jookia>I mean, machine. BIOS
<Jookia>If you're on a machine with EFI you need to switch it to use a compatibility mode
<_hubertus_>but i try use guix on virtualbox, qemu
<Jookia>_hubertus_: If you use it on Guix or QEMU you have to use the image as a hard drive, not a USB key
<_hubertus_>but ther is no image file to download
<Jookia>From the Guix homepage there's a GuixSD image you can download
<_hubertus_>xz file?
<Jookia>Yeah, extract that
<_hubertus_>ok, thet what should I do?
<Jookia>You should have an image file (the same file name without the xz)
<Jookia>Do you want to try it in a VM or on a machine
<Jookia>Well load up QEMU with that as the disk image
<_hubertus_>OK, trying...
<lfam>_hubertus_: There are some instructions on running GuixSD in QEMU here:
<lfam>They aren't in the online HTML version of the manual yet, but you can build an HTML version of the manual if you want
<Jookia>civodul: In this case it'd be useful to have switch-symlinks, every alternative seems to lead to use switch-symlinks in this case though I might be in tunnel vision :P
<_hubertus_>Istruction fo VM is very nice. But if I'm rigt, first I should use guix app to build guixsd image?
<lfam>_hubertus_: Yes, those instructions are meant to help you boot a VM image that you would create with `guix system vm-image ...` command
<lfam>_hubertus_: You can also try `guix system vm ...`, which gives you a read-only operating system
<lfam>That's not as fun though
<_hubertus_>better to read guix manual it seams to be powerfull app
<civodul>Jookia: maybe you can post your patch when it's ready and we'll see from there
<civodul>Jookia: the switch-symlinks thing is a minor issue anyway
<Jookia>civodul: Okay- Patch should be posted within a few hours to be teared apart :)
<civodul>ACTION added a non-HTTPS URL for pkg-config
<_hubertus_>I have one more question. When you predict to release stable version?
<civodul>_hubertus_: the next beta release is due within a couple of weeks
<dannym>Does anyone have experience with collecting Linux oops crashdumps when they happen while X is running? Couldn't see any backtrace, but Num Lock was blinking...
<dannym>oops, Caps Lock I think
<lfam>Hm, somethin has changed between Guix and QEMU. I'm having all sorts of trouble
<_hubertus_>thanks for info and tips. See you to the next time
<lfam>When I try to reconfigure, I get an error from guix/scripts/system.scm, when doing [string-append "--root=" ...]. The error is something like "string-append: Wrong type (expecting string): #vu8(6 145 236 74 2 206 [...]". There are some more numbers
<lfam>This error repeats itself even when I try to reconfigure using the same configuration I initialized the system with. This is in QEMU
<lfam>Does anyone have any idea what could be wrong?
<lfam>Those numbers seem to always be the same
<wingo>i think i found a bug in polkit introduced in the last release, fml
<davexunit>lfam: not sure, do you see a full backtrace?
<davexunit>the #vu8 thing is the printed representation of a bytevector
<lfam>davexunit: I do see the whole backtrace
<davexunit>that would be good to include in a bug report or whatever.
<davexunit>there's a procedure that is expecting a string but getting a bytevector
<davexunit>may or may not be hard to fix if you can take a look at the problem code shown in the backtrace
<lfam>Here's the backtrace and the configuration file:
<davexunit>lfam: so, that corresponds to the boot-parameters-root-device selector foor the <boot-parameters> type
<davexunit>in gnu/system.scm
<davexunit>seems there is two ways for such objects to be created
<davexunit>one way is for it to be converted from an sexp read from a port
<davexunit>the other is a gexp
<NiAsterisk>so I get not much response from the hackerspace here who develop panopticon.
<NiAsterisk>and it seems like i need to package rust
<NiAsterisk>to package cargo
<NiAsterisk>to write something to load cargo packages
<NiAsterisk>to build panopticon
<davexunit>usually one avoids the language package manager
<NiAsterisk>and I found this: .. did this attempt go somewhere?
<davexunit>when making guix packages
<NiAsterisk>but their difference between gh and website is strange
<NiAsterisk>I could link to the thread in german.. but basically some parts in the source are missing.
<NiAsterisk>output is in english though, so I can forward it to guix-dev
<lfam>davexunit: Do you think I should file a bug report?
<davexunit>lfam: are you running guix from git?
<davexunit>if so, first try a full re-compile with 'make clean-go && make'
<davexunit>this will ensure that it's not an ABI issue
<lfam>No, it's with a "fresh" guix pull
<NiAsterisk>sending a translated email to the dev list, as the archive on das-labor is members only
<Jookia>Time to do a world rebuild woo
<NiAsterisk>oh, somebody just replied.
<lfam>davexunit: I want to see if the bytevector is something I recognize, and I found this info on how to convert between bytevectors and strings:
<lfam>But I don't know how to import those modules
<lfam>Or if I would need to install them
<davexunit>lfam: (use-modules (rnrs bytevectors))
<NiAsterisk>i should figure out how to make GNUS keep copies in folders when two lists are in To: or CC: / BC: / BeenThere.
<lfam>ERROR: In procedure utf8->string: ERROR: Throw to key `decoding-error' with args `("scm_from_stringn" "input locale conversion error" 84 #vu8(145))'
<davexunit>lfam: that means its not utf-8 encoded text
<lfam>Okay then
<lfam>I'm going to recreate the QEMU image from a git checkout, and if it still doesn't work, at least the case will be reproducible
<arianvp>Heya =D
<arianvp>Is this the same system as the patching system in nixos?
<arianvp>I'm on nixos with the nice glibc vuln im trying to fix
<Jookia>Security patching is broken on NixOS but AFAIK it's at the top of the list of things to fix
<arianvp>So it's currently also broken in Guix then right?
<Jookia>I mean Guix, not NixOS
<Jookia>It's fine in NixOS I think
<davexunit>arianvp: yeah our "grafts" system isn't working right now. we'll be working on fixing that.
<arianvp>I'm still not fully convinced on grafts though. It still needs to rebuild some dependencies right?
<arianvp>it's like tracing a path in a tree
<davexunit>it's not a full rebuild
<Jookia>Grafts basically build a new tree of your system but refers to old binaries with libraries patched
<davexunit>it's essentially just recursively replacing references
<davexunit>so new store entries for the whole tree, but much faster.
<arianvp>I also learned (the hard way) that the user profile is not patched by this system though (at least in nixos)
<arianvp>Have you guys explored any options how to also do this for the user profile?
<davexunit>arianvp: guix pull && guix package -u
<davexunit>updating the system is different than updating user profiles
<arianvp>but I mean. say you have gnu emacs installed in a user profile
<arianvp>(which depends on glibc)
<davexunit>the user needs to update
<arianvp>if I add a graft to patch the system glibc. There will still be an old version of glibc in the user profile right?
<arianvp>how would you patch the user-local glibc? Before an update has made it to the package repository?
<Jookia>arianvp: You'd edit your Guix and patch it yourself then update
<davexunit>which is the same way you'd do it for the whole system
<arianvp>makes sense
<davexunit>what would be nice is a way to detect vulnerable software in profiles
<davexunit>so that the user still has control over their profiles, but is more aware of security issues.