IRC channel logs

2016-06-06.log

back to list of logs

<galex-713>davexunit: did you found the problem with sly/uniform stuff/shaders.scm?
<davexunit>galex-713: it's not a problem with sly.
<davexunit>it's a problem with your graphics drivers not supporting the version of GLSL that sly requires
<ArneBab>mario-goulart: It took me a while to understand macros, too.
<ArneBab>mario-goulart: sorry, that was intended for marusich
<dsmith>Monday Greetings, Guilers
<fantazo_>hen someone would want to use the compiler tower to compile to something specific how difficult is that to do? or asked differently has anybody done something like that who isn't in the core guile dev "team"?
<fantazo_>when
<fantazo_>sorry.
<davexunit>fantazo_: it's not that hard. I tried it recently and was very happy with how easy it was.
<fantazo_>davexunit, what did you translate?
<davexunit>my favorite example of an outside project using this is guile-xcb https://github.com/mwitmer/guile-xcb
<davexunit>it actually reads the XML specification for XCB and compiles it to Scheme code.
<fantazo_>davexunit, that's really cool. and I also didn't think of that the xcb protocol has an xml spec.
<dsmith>sneek: macros?
<sneek>I've heard macros is http://hipster.home.xs4all.nl/lib/scheme/gauche/define-syntax-primer.txt
<csed>Huh. I wonder if you could uh, drop a Guile shell into dwm.
<csed>Like, the suckless WM.
<suppi>hello
<suppi>I heard there's an effort to incorporate guile into emacs. does anyone know if that work is open source?
<OrangeShark>suppi: https://www.emacswiki.org/emacs/GuileEmacs
<OrangeShark>there is a link to the repo there
<OrangeShark>it is also available if you use the guix package manager, the package is called guile-emacs
<wingo>meep
<ijp>ribbit
<wingo>ijp!
<wingo>you are a person with opinions
<ijp>you bet
<wingo>did my recent consternation about textual output primitives provoke you into any particular opinion
<ijp>was this on the mailing list?
<wingo>yes
<wingo>also on irc but the writeup is http://thread.gmane.org/gmane.lisp.guile.devel/18337
<civodul>ACTION looks
<wingo>yay, i nerdsniped a ludo
<wingo>a confusing issue!
<wingo>i wish in retrospect that we had named (ice-9 binary-ports) to be (ice-9 binary io) or (ice-9 io binary)
<ijp>my gut says (ice-9 textual-ports), but it is currently hungry and prone to making bad decisions
<suppi>OrangeShark, thanks. is this the one? it seems it hasn't been updated in a while http://git.hcoop.net/?p=bpt/emacs.git;a=summary
<ijp>if you did want a grab bag, you could call it (ice-9 io) or (ice-9 ports)
<wingo>(ice-9 ports) is a thing in master
<ijp>oh right, so it is
<wingo>it has all the porty bindings
<wingo>but it is all re-exported to (guile-user) because compatibility
<wingo>so adding something to (ice-9 ports)'s interface adds it to the default environment
<wingo>which is nae so good
<ijp>right
<wingo>we could rename (ice-9 ports) but probably it already has the right name, dunno
<wingo>ice-9 textual-ports would be fine i guess, dunno what else would go in there
<wingo>maybe r6rs-like functions, dunno
<wingo>get-line, get-string-all, get-string-n!, get-string-n, and maybe the datum functions... dunno
<wingo>interesting that r6rs does not provide `display' except in its legacy module
<wingo>put-datum is `write'
<wingo>ok as per civodul on ML, R6 it is via (ice-9 textual-ports) :)
<civodul>R6 is the thing!
<wingo>:)
<paroneayea><wingo> ijp!
<paroneayea><wingo> you are a person with opinions
<paroneayea><ijp> you bet
<paroneayea>haha
<wingo>:)
<wingo>good evening paroneayea :)
<wingo>did you get a chance to look at the memcached stuff? it works!
<wingo>also the web server works
<wingo>with ethreads and such
<paroneayea>wingo: I haven't, I'm at the w3c socialwg face to face meeting
<wingo>rrrrright
<paroneayea>but I'll be looking soon
<wingo>godspeed you noble catherd
<paroneayea>thanks :)
<wingo>:)
<ijp>that reminds me, didn't I write a guile-memcached?
<wingo>ijp: maybe you did!
<wingo>i made a wee implementation just as an example, it's not a real thing
<ijp>ah no, "scheme-memcached"
<wingo>i just wanted to be resistant to slow loris attacks
<wingo>i dunno why that is my obsession
<daviid>wingo: I did prefer (ice-9 textual-ports), while reading, fwiw
<wingo>daviid: excellent!
<daviid>because such a module may later be filled with other stuff pertaining to it 'per defs'
<wingo>the manual on ports is cringeworthy bad, sheesh
<wingo>"Ports are created by opening," wat
<civodul>heh
<daviid>folks! I did submit my GI project to GNU [although a mockup at this time and prob for the next 6m to a 1y] and it has been pre accepted: it will be part of GNU anyway, but rms would prefer another name, so help! here is the actual name and why
<daviid>GNU Gbank
<daviid>a Guile binding for GObject Introspection
<daviid>Unimaginatively named after Sbank [5], unimaginatively named after
<daviid>pybank, a Python binding for GO bject Introspection.
<daviid>:)
<daviid>if anyone can come up with a better [short, it is used everywhere in the code] it would be nice
<paroneayea>daviid: oh cool
<dsmith>daviid: Something like gobi ?
<dsmith>daviid: Or maybe something compeltey different? Like gorfle-snork ?
<daviid>dsmith: gobi guile object binding ?
<dsmith>Sure
<daviid>gorfle-snork what beast is that ? :)
<dsmith>Something made up on the spot.
<dsmith>IT's good to have a name that is uncommon. Make for better global internet searches
<dsmith>Makes
<daviid>dsmith: let me paste the comment rms made
<dsmith>Oh no! Another RMS approval contest!
<civodul>:-)
<daviid>yep, i had to change KsĂȘ and found Foliot, which I now like very much too :) but one, Gbank sounded really great to me and I am idea less
<daviid>here the comment reported by the gnuevaluator, John Darrington: 1. The name. In contrast to how things were 30 years ago, today rms likes to have names which are non-cryptic, and convey meaning to the layman. Can you think of an alternative to "GBank" ? I know it is hard to find a short, name which is not already taken ...
<civodul>ACTION thinks forwarding private messages is unkind
<daviid>oh, this is not private, by all means
<daviid>by no means
<civodul>you sure?
<dsmith>ACTION stuffs fingers in ears and starts chanting La La La La
<civodul>:-)
<daviid>civodul: yes I am sure
<dsmith>Hm.. A short meaningful name that means "Guile Bindings for GObject Introspction"? Oi
<dsmith>daviid: Ask him for a suggestion
<jordyd>I'm trying to figure out the best way to modularize my code. I tried using define-module but use-modules doesn't look in the current directory for modules. I also tried a simple load but then I get warnings about possibly undefined variables.
<jordyd>Is there a standard way to do this?
<roelj>jordyd: Have you tried adding the current directory to the GUILE_LOAD_PATH?
<jordyd>roelj, I haven't, though I considered it. It seems a rather hacky way to go about it.
<dsmith>Which is easisly done when testing with the -L option.
<jordyd>Is that pretty much the standard way to do this sort of thing?
<jordyd>I was hoping for a way that would kind of "just work". I guess I could modify %load-path in the main file and then I wouldn't have to rely on setting an environment variable every time.
<dsmith>Well, you generally do not want '.' in your load path for the same reasons you don't want it in your $PATH.
<OrangeShark>jordyd: ya, try adding it to your GUILE_LOAD_PATH, several projects use a script called pre-inst-env which sets it up
<jordyd>dsmith, Is there a better way then?
<dsmith>daviid: Introspection: http://www.thesaurus.com/browse/introspection I kind of like "scrutiny"
<dsmith>jordyd: Won't this module be eventually installed somewhere on the existing load path?
<jordyd>dsmith: I mean, it could, but I would just be using it internally for my app
<dsmith>jordyd: I don't know.
<jordyd>Hmm... maybe I can write it in such a way that it would be useful to third parties
<jordyd>But still there should be some standard way to do this kind of thing imo. Something for the guile developers to think about maybe?
<daviid>civodul: sorry, my attention has been grabbed by a collegue on sonmething else, but the message was copied to gnueval, not just for me, and then, we'are talking choosing a name for a GNU Project, and the idea(s) behind what is today considered a better name ...
<daviid>dsmith: yes I will ask for everybody's idea, including rms himself of course
<daviid>dsmith: so far I like gobi, but maybe the acronym letters are the wrong order? not sure what you ment exactly
<daviid>dsmith: tx by the way
<dsmith>daviid: Oh, gobi could mean different things I guess. Guile OBject Introspecion, GObject BIndings, probably more.
<daviid>dsmith: cool! I will propose it, let's see, and still look for some alternatives, anyone fancy a name let me know...
<dsmith>daviid: The thing that comes immediately to mind is "navel-gazing" ;^}
<daviid>:) explosive