IRC channel logs


back to list of logs

<gnusosa>Quick question: is it normal that Glib build is stuck in this test 'PASS: network-address 52 /gresolver/resolve-address/12'
<paroneayea>attempting to package guile-emacs
<paroneayea>I'm stuck though on building the guile branch
<paroneayea> is what I have
<paroneayea> the error I'm getting
<paroneayea>there must be some kind of script for affecting the tools that configure calls...
<paroneayea>the #!/bin/sh and etc
<paroneayea>hm :(
<paroneayea>I think I need help to continue. But I've made progress!
<paroneayea>I don't really understand why my usage of substitute* isn't working, basically
<paroneayea>solved it.
<gnusosa>what was it?
<paroneayea>gnusosa: I needed to patch the #!/bin/sh before I called
<paroneayea>well, needed to patch the reference to be the input
<paroneayea>the guilemacs guile branch is now compiling
<mark_weaver>paroneayea: instead of patching, just do (zero? (system* "sh" "")) in the autogen phase (which should be put after unpack)
<mark_weaver>if you search for that expression, you'll find several occurrences in gnu/packages/*.scm
***rohan_ is now known as ajnirp
<paroneayea>mark_weaver: yeah I'm doing that, but it's one of the scripts that's called in that's the problem..
<phant0mas>hey mark_weaver the problem isn't with flex, it's with m4 when cross building for hurd
<phant0mas>all the tests fail with seg fault, maybe the best course of action is just to disable them for i686-gnu on m4
<phant0mas>for your yesterday question, in order to build the cross toolchain for hurd I only needed flex which targeted linux
<phant0mas>and all the packages built with that toolchain , when target=i686-pc-gnu, didn't need flex
<phant0mas>only mig needs flex
<df_>what's the standard way to get wifi with wpa working?
<df_>also what's the situation with term(cap|info)? TERM=rxvt-unicode is apparently not recognised
<paroneayea>hello #guix
<paroneayea>civodul: I have some WIP packages for guilemacs-guile and guilemacs-emacs
<paroneayea>about to send to the list
<paroneayea>"make check" fails in guilemacs-guile and I don't know enough to have it figured out
<paroneayea>I think maybe having guix packages of guile-emacs might increase chances that people might try it out though...
<civodul>paroneayea: yes, it's a good idea
<civodul>that's nice
<paroneayea>just shot to the list
<mark_weaver>the texlive-texmf-2014.x86_64-linux build on hydra has been going for almost 40 hours so far
<mark_weaver>I guess most of that time is spent trying to transfer it over the net
<mark_weaver>we should probably arrange to have those builds done only on slaves that are closer to hydra
<civodul>it's really a tricky problem, but a major annoyance :-/
<civodul>eventually we should do something smarter, like using content-based address stores
<mark_weaver>to avoid transferring files when the destination already has a file of the same hash, I guess?
<mark_weaver>it might be good to do that only for files above a certain size. for small files, the overhead of the mechanism would probably overwhelm the savings
<DusXMT>I wonder, where are service definitions for dmd stored?
<mark_weaver>DusXMT: gnu/services/*.scm
<DusXMT>In a running system, that is
<mark_weaver>DusXMT: looks like the services you selected are combined together into /gnu/store/*-dmd.conf which is referenced in /run/booted-system/boot
<civodul>dirty hack: try "cat /proc/1/cmdline", which contains the dmd.conf file name
<civodul>or what mark_weaver writes, indeed :-)
<DusXMT>Thanks :)
<DusXMT>I'm thinking of trying dmd outside of guix, so I'm trying to collect information about it
<mark_weaver>civodul: btw, I've been unable to get resolution of *.local names working on my system. I recently switched my config to use %desktop-services and (name-service-switch %mdns-host-lookup-nss), but it still fails
<mark_weaver>it's not important to me, so no rush, but just letting you know that there's a problem
<mark_weaver>I guess I should file a ticket
<civodul>mark_weaver: does avahi-resolve-host-name work?
<civodul>does nsswitch.conf mention mdns?
<civodul>and does /proc/$(pidof nscd)/env has an LD_LIBRARY_PATH that contains nss-mdns?
<civodul>that's the checklist :-)
*mark_weaver looks
<mark_weaver>civodul: "yes" to questions 2 and 3.
*mark_weaver looks for avahi-resolve-host-name
<mark_weaver>avahi-resolve-host-name works
<civodul>so it works! :-)
<civodul>hmm, ok
<mark_weaver>oh wait, it's all working now
<mark_weaver>well, nevermind :)
<civodul>aah, i like that ;-)
<civodul>BTW i'm glad i fixed wicd for my use case
<civodul>this code is pretty horrid
<mark_weaver>okay, so printing.
<mark_weaver>I have a networked printer that's very compatible with free software (an HP)
<zacts>I always get HP printers
<mark_weaver>when I try to print from evince, it doesn't show the printer
<zacts>I find they have the best support for free software
<zacts>(in my experience at least)
<zacts>but I have yet to get a lazer printer
<mark_weaver>I remember with Debian it was very easy to get it set up with cups. not sure how to do it here
<zacts>all ink-jets for me thus far
<zacts>mark_weaver: I think hp has an 'open source' site on how to set it up via cli
<mark_weaver>specifically this is an HP C309a
<zacts>let me find it
<civodul>mark_weaver: it's taken care of by GTK+ with CUPS
<zacts>oh, do you not even have a web browser?
<zacts>a graphical web browser?
<civodul>somehow i see my partner's printer, which is connected on their own pc
<zacts>because you can always set up CUPS via a port on icecat or whatever
<mark_weaver>zacts: yes, I have a web browser. GNU IceCat
<civodul>i wonder why something this sophisticated works :-)
<zacts>yeah, CUPS let's you config via icecat via a certain port
<mark_weaver>ah, right
<zacts>but you have to modify the CUPS .conf file, wherever it is, to allow permissions to do this
<mark_weaver>except that I have no cups daemon running.
<zacts>and that can get kind of buggy / messy if you don't know what you are doing
<mark_weaver>the printer is advertising itself over mDNS
<zacts>hrm... yeah you'll need to start a cups daemon, do you know how to do this?
<zacts>ah, hrm...
<mark_weaver>I can see it in "avahi-browse -avr" and name resolution now works to its .local address
<zacts>well, I would look into it more to help, but I currently don't have the time, perhaps later today.
<zacts>I'm not an expert at setting up printers without CUPS though
<zacts>mark_weaver: although, I think the FreeBSD guide has doc on how to do this kind of thing though
<zacts>the handbook
<zacts>for lpr and things without CUPS, and direct printer connections
<zacts>(I tried this kind of thing back when I used FreeBSD)
<mark_weaver>zacts: thanks, but the problem I'm having is likely specific to GuixSD.
<zacts>ah ok
<mark_weaver>well, it's not urgent
<zacts>well, sorry I couldn't be of more help, but I tried @ least. :-D
<mark_weaver>civodul: in the example OS desktop configurations, which include wicd, I think "netdev" should be added to the user's supplementary groups. last I checked, that was needed to allow the user to configure wicd
<mark_weaver>should I go ahead and add it?
<zacts>hey mark_weaver was that emacs parentheses mode for editing scheme called paredit?
<zacts>I can't remember which one you mentioned
<mark_weaver>zacts: yes
<zacts>ok cool
<mark_weaver>highly recommended
<zacts>thanks, I'm going to try to teach myself this one
<mark_weaver>zacts: here's a video demo'ing it:
<zacts>oh, let me check it out
<zacts>I've fully converted from vim --> emacs now, with emacs bindings too
<mark_weaver>I should warn you, though, that it's a steep learning curve. it will be painful to use until you have become somewhat proficient.
<mark_weaver>ooh, nice!
<zacts>ok, well vim was this way
<zacts>and emacs
<zacts>so no biggie
<mark_weaver>it's well worth the effort to learn. I don't know how I lived without it
*DusXMT thinks he'll go ahead and install it now, emacs
<zacts>mark_weaver: the mention of paredit being about manipulating AST rather than text
<zacts>that kind of makes me think of the difference in philosophy of emacs and vim
<zacts>vim is for editing *text*, emacs is for editing *ideas* (in a sense)
<zacts>I kind of think
<civodul>mark_weaver: re netdev, right, go ahead
<civodul>i had forgotten that one
<mark_weaver>civodul: okay, thanks!
<mark_weaver>zacts: I don't think that's really true.
<zacts>why no?
<zacts>*why not?
<mark_weaver>I don't know what it means to edit an idea, honestly, and I've never read anyone write anything like that about Emacs.
<mark_weaver>RMS also prefers to simply edit text, and doesn't use anything like paredit.
<zacts>mark_weaver: well, operating at the level of an AST, versus the level of characters
<zacts>is an example
<mark_weaver>sure, but that's not part of the philosophy of emacs.
<zacts>you are operating on abstractions of text, chunks with meaning, versus characters / letters
<zacts>hm.. ok
<zacts>well, /me is just an emacs idiot! I do declare this! :-D
<mark_weaver>AST-style editors like paredit are quite new, and historically it has not been something that emacs developers are in favor of, afaict.
<mark_weaver>heh, I wouldn't use the word "idiot", but we all start out in a state of ignorance until experience reduces it :)
<mark_weaver>it's great that you're willing to step out of your comfort zone and learn a new tool :)
<zacts>mark_weaver: have you ever met any of the authors of SICP?
<zacts>or gotten your SICP autographed?
<zacts>(if I may ask)
<zacts>s/your SICP/your copy of SICP/
<mark_weaver>for a couple of years I spent most of my time in the GNU project office at the MIT AI lab, adjacent to Gerry Sussman's office and two doors down from RMS and Hal, so yes, I saw them every day.
<mark_weaver>but I never asked them to autograph my copies of SICP
<mark_weaver>(I have both the first and second editions)
<zacts>ah cool!
<mark_weaver>Gerry kindly allowed me to sit in on his classical physics class, which uses SICM (Structure and Interpretation of Classical Mechanics)
<zacts>yeah, well if you know someone that well, it could be odd to ask for an autograph
<zacts>oh really, neat!
<zacts>how was the class?
<mark_weaver>yeah, it's just not something that I tend to do, in general. not that there's anything wrong with it :)
<zacts>and how far through the SICM curriculum did you get?
<mark_weaver>I confess that I was too busy with other things at the time to do the exercises, which meant that my understanding of the class became increasingly superficial as the class proceeded, and eventually I just stopped going. too bad, because it was a very interesting class. maybe some day I'll have enough free time to finish it.
<zacts>heh, yeah
<zacts>I'm actually hoping to get through SICM either near the end of this year, or middle of next year
<zacts>I did recieve the 2nd edition of the book
<zacts>mark_weaver: is it true that Hal helped co-found the FSF?
<mark_weaver>yes, both Hal and Gerry took RMS in, let him keep his office in the AI Lab even after he quit his job there, and have been helping to support him ever since.
<mark_weaver>the FSF came a bit later, and they were both founding members I believe
<zacts>oh neat
<zacts>mark_weaver: were you an actual student at MIT, a teacher, or just a worker / researcher?
<mark_weaver>none of the above
<zacts>a slacker?
<mark_weaver>I just showed up one day and started helping RMS
<zacts>that's cool
<mark_weaver>for a while I worked as the FSF staff sysadmin
<mark_weaver>I've also been a member of MIT's all-volunteer radio station for over 15 years
<mark_weaver>but this is all off topic here :)
<zacts>ah, yeah sorry
<zacts>but, interesting nonetheless!
<mark_weaver>no worries!
<zacts>well, heading home from a cafe. bbl
<kete>interesting story, mark_weaver