<mark_weaver>sneek: later tell civodul: po/packages/POTFILES.in contains "gnu/packages/weechat.scm" which no longer exists, and this caused the following error for someone: *** No rule to create the target ”../../gnu/packages/weechat.scm” needed for ”guix-packages.pot-update”. Stops.
<Piece_Maker>guys, ive got a slightly nooby question... im trying to add a second user to my guix desktop.scm.. ive basically copied everything from the desktop.scm default from (cons (user-account) down to the (home-directory "dir")) - is this enough, or do i need to add more to that?
<Piece_Maker>i think for 'core' applications ill do it in the OS config as my dad is 'non-techical' but generally knows what he wants... so ill let him do his own thing but put some standard defaults in the config ;D
<mark_weaver>and btw, I should mention that "guix pull", which is roughly analogous to "apt-get update", only updates guix for the user who runs it, so each user who runs "guix" needs to run "guix pull" to get the updated package descriptions.
<Piece_Maker>so how does one update the software in your system profile? guix pull + reconfigure?
<Jookia>If you're going to instruct your dad to update it maybe it'd be best to make it so he can do it as a user
<mark_weaver>that means that you need to run it as your normal user so that packages installed in your user profile are up-to-date, and you also need to run it as root, because root is the user who must run "guix system reconfigure" to update the system-wide profile
<Piece_Maker>Jookia: ill be keeping on top of the system updates, im not that evil :P
<Piece_Maker>so, (as root) guix pull + guix system reconfigure = updates the stuff in OS config.scm... guix pull + guix -U = updates stuff in my own users /home
<mark_weaver>suitsmeveryfine: well, this is not an area of expertise for me, but at this point it would be easy for you to try various other things. running "guix system reconfigure" after merely changing gnu/services/xorg.scm will be very fast.
<Jookia>jmd: There's a hurd-wip branch, you could try that :)
<civodul>jmd: not soon, but that's a likely GSoC project
***civodul changes topic to 'GNU Guix | http://gnu.org/s/guix/ | videos at https://gnu.org/s/guix/help/#talks | 0.9.1 is in the works | https://libreplanet.org/wiki/Group:Guix/GSoC-2016 | avoid pastebin.com, it blocks Tor | channel logged: <https://gnunet.org/bot/log/guix>.'
<NiAsterisk>what's up. this powwow software is fun in licenses: public domain, gpl2, gpl2+ and gplv3 in headers. I searched for a newer version, found one, it now includes gplv3.. what do you do in this case? name every file :D?
<Jookia>NiAsterisk: public domain or not licensed?
<NiAsterisk>wait. it might be that cc0 is gone, i have to check this for every file. so if it's a combination of gpl2,gpl2+,gpl3, I do (license: gpl2 gpl2+ gpl3 and name all the indivual files? can't be right
<Jookia>Hmm, what other projects are multilicensed
<NiAsterisk>wild. can they do that? this is strange... the 1998 version of catrw.c was public domain, now it is gpl2
<NiAsterisk>hm.. if I have gpl2+ and then there's gpl3+, gpl2+ is somewhat inherited, covered by, included into, transfers to (sorry I lack the word to describe what I want to say) gpl3+ ... so I can just simply state the licenses and name no files
<civodul>Jookia: we do have a 'dependencies' field in 'file-system' for similar reasons, but i agree it's not super elegant
<roelj>I'm trying to build a package, which fails with: /gnu/store/5s9lvpwiszryhv3zvqm35fayyvmrr2px-python-3.4.3: bad interpreter: Permission denied. And I don't know what this means in Guix.. I thought it could either not access python at all, when it's not part of the inputs, or access it normally when it is part of the inputs.
<Jookia>civodul: If we go crazy could we start thinking about this as functions that take an input + expected output of a file or directory? (luks /dev/sda hdd -> /dev/mapper/hdd), (lvm /dev/mapper/hdd matrix -> /dev/mapper/matrix-root,/dev/mapper/matrix-swap,/dev/mapper/matrix-scratch), (btrfs /dev/mapper/matrix-root -> /)
<Jookia>civodul: This would also fix the issue of dependencies since you could just match outputs to inputs
<suitsmeveryfine>But I've rebooted three times and it worked every time. Now I have two-finger scrolling and three finger tapping for middle click
<Jookia>rekado: Basically mapped-devices can't depend on each other and it's problematic for things like LUKS and LVM. The problem is that mapped-devices don't really have a 'target' that works well, and in some cases may require multiple inputs or multiple targets
<Jookia>rekado: I wrote a quick idea on how to unify filesystems and mapped-devices in to a system that allows this and has dependency resolution for free
<wingo>suitsmeveryfine: "without any hacks" what does that mean? :)
<Jookia>rekado: My current hack to work around all this is to remove dependency resolution and just load all mapped devices then filesystems, with an LVM mapped filesystem thing added
<suitsmeveryfine>wingo: it means I didn't have to remove any drivers from xorg.scm which I first attempted earlier yesterday evening.
<Jookia>Is there a way to copy a store to another partition, or use it to download things from
<suitsmeveryfine>I believe that libinput is now being used instead of synaptics, because synclient commands do not work.
<paron_remote>> To cope with these kinds of requirements, configuration-management systems tend to invent a domain-specific configuration language that (eventually) becomes Turing complete, starting from the desire to perform computation on the data in the configuration (e.g., to adjust the amount of memory to give a server as a function of the number of shards in the service). The result is the kind of inscrutable "configuration is code" that peopl
<paron_remote> were trying to avoid by eliminating hard-coded parameters in the application's source code. It doesn't reduce operational complexity or make the configurations easier to debug or change; it just moves the computations from a real programming language to a domain-specific one, which typically has weaker development tools (e.g., debuggers, unit test frameworks, etc).
<paron_remote>> We believe the most effective approach is to accept this need, embrace the inevitability of programmatic configuration, and maintain a clean separation between computation and data. The language to represent the data should be a simple, data-only format such as JSON or YAML, and programmatic modification of this data should be done in a real programming language, where there are well-understood semantics, as well as good
<NiAsterisk>"rain1 what I've often wondered is, where is any interesting/good discussion online?" mostly in the places where few people go to.
<NiAsterisk>"Jookia Why do people not like GNU again?" the most comments I get are "stallmans / gnu project visions are detached from reality of market etc etc.." up to the point where I had to criticize an author on the lack of research on (among other things) free software in general "people are afraid of open sourcecode" "all technology in the past 25 years was build for surveilance / etc etc".
<NiAsterisk>at some point I will translate my initial review of the book to english. it's so sad you won't even laugh about the lack of competence to search "what is free software -> gnu.org -> philosophy" etc
<amz3>I have a possible GSoC idea, I think I already talked about last year but I don't remember what was the result of the talk
<amz3>I throw the idea (again) here, to have feedback and maybe help me write a summary of the idea if it good enough
<amz3>the idea is to create a computing grid of guix boxes
<amz3>I am not very familiar or not familiar at all with existing solutions
<amz3>so I'm not sure how this could be an improvement over existing solution
<amz3>also I don't know what guarantees we can provide over the correctness of the result (has results been tempered somehow?)
<efraim>build in parallel across the boxes and compare the results?
<efraim>we already have --rounds and offloading, why not offload the rounds
<amz3>why I think about this and guix is because with guix, the job provider doesn't have to create a specific program to be run on some cluster with a custom programming semantic instead he just have to use his everyday programs and programming paradigm, split the dataset if any, and provide the job to a guix cluster
<amz3>I'm not talking about building programs, but running programs on guix boxes with guix packages
<amz3>the idea behind that is that guix org could then ask guixsd user to subscribe to the program to help research, we could maybe even give money to people helping the project but i don't beleive in this
<amz3>anyway, another advantage I can think is that since the grid use conventinonal programs the speed of programs would be better...
<amz3>you may say, why other distros didn't do it... you would pin a good point, but with guix reproducibility in mind... maybe it makes more sens for guix to do that
<efraim>with BOINC programs they're already shipping you a pre-compiled binary to run on your machine
<lfam>By "non-texinfo" I also meant "non-patch-formatted"
<lfam>Anyways, we've had this discussion at least once before. I think if you search the guix-devel archives you can find people's arguments one way or the other. I would read that and then bring it back up if you think there is a new angle on the subject, or if you just really want a wiki.
<lfam>My preference is to put it in the manual. I think that wikis are a second-best alternative to a comprehensive and up-to-date manual
<NiAsterisk>wow.. watching a video on the company RealDoll and with their research into AI implementation, it is moving to a scary, interesting direction almost Android like you see in the game "RememberMe"... weird future.
<lfam>NiAsterisk: There is a documentary about those dolls and their owners. Fascinating! *VERY* bad date movie
<NiAsterisk>even if you talk to people and try to point out security risks they introduce, they are in denial. I tried it again last weekend... the app which controls the intrusion detection and smoke detectors in a house of someone I know who works for a larger company.
<rain1>i got the same warning but i think the cert is good
<lfam>"And you will be the first in line when there is additional work to distribute!" Haha
<fhmgufs>It's a pity that guix.org isn't available ...
<NiAsterisk>a_e: the associazione guix europe, could this be used for for example getting all current copylefts signed over / dualcopyrighted to it? I have no idea why I ask this, there was a question which faded.. something with copyrights and like it is handled in gentoo with the gentoo foundation being the only name mentioned in the headers.
<a_e>NiAsterisk: No, this is not a goal. If there were any copyright assignments to make, it should go to the FSF.
<NiAsterisk>fhmgufs: guix.is would be... with my registrar it would be twice the price including optional privacy whois etc, but the normal price is just ~30 euro.
<NiAsterisk>w/should I have the enormous amount of money 10 years would cost with orangewebsite.com for an .is (compared to nic.is), I would opt to donate 10 years of guix.is to transfered ownership to guix. but that's like 689 € vs 300 € normal price.
<NiAsterisk>i think you could just say shoes as i like to interchange some words which have the same functionality. there are books/wikis out there who deal with the origin and explanation of these sentence constructions
<paron_remote>suspicious ownership or permission on `/gnu/store/xmpk0cj27i8bb54b4b3biniqgmwpyck9-fdm-1.9'; rejecting this build output
<NiAsterisk>i use kamm bürste (comb, brush) interchangeable. doesn't make much sense, same with a selection of other things I don't care much about in language i use daily.. it's just not that important for my personal dictionary, where as somebody in the business which deals with hair would think it's very important.
<lfam>paron_remote: IDK never seen that before. Maybe the build process trieds to chmod the output dir?
<NiAsterisk>for people not following the list that much: an oportunity for me to maybe add another hydra slave, or whatever guix needs, maybe even a third mirror, came up. i send an email to the devel list about this.