IRC channel logs


back to list of logs

<nalaginrut>morning guilers~
***michel_mno_afk is now known as michel_mno
<jockej>hello again
<jockej>i'm still trying to build guile-emacs
<jockej>but i'm having some problems...
<jockej>it fails trying to compile newdisp.c i think, and spews gigabyts of text to stderr
<jockej>I thought I read somewhere that guile-emacs only works in the terminal
<jockej>so should I try some other options to configure?
<taylanub>my experience was the opposite; that it only works with GTK
<taylanub>what configure flags are you using?
<jockej>none :) only --prefix
<jockej>or, well, defaults, I should say
<nalaginrut>I've ever tried guile-emacs for a while, it works fine for me (fine means it works and occasionally halt). Dunno know how about it now
<jockej>hm, so what ae the options one should set?
<jockej>or is it better to try --with-x-tookit=no?
<taylanub>jockej: with gtk installed on your system and without any configure flags other than prefix, I would say
<taylanub>also what OS and compiler version are you using?
<jockej>linux 3.18.6 and gcc 4.9.2
<jockej>arch linux, if it matters
<jockej>seems I have gtk2 and gtk3 installed
<taylanub>I somehow remember the dispnew issue but not the cause or solution :\\
<taylanub>jockej: you're building both bipt's guile and emacs repos, right? not upstream guile.
<jockej>yes, I built bipt's guile repo as well
<jockej>then I copied stoff from there to /usr/include and /usr/lib
<jockej>until I got configure to run
<jockej>it should be possible to build guile-emacs without x support though, rigth?
<jockej>shouldn't that be a lot faster?
<jockej>I just wanna try and see if I can get it to run
***michel_mno is now known as michel_mno_afk
<jockej>dispnew.c: At top level:
<jockej>dispnew.x:2:46690: warning: inline function ‘BUFFER_OBJFWDP’ declared but never defined
<jockej>Makefile:342: recipe for target 'dispnew.o' failed
<jockej>it failes here...
***michel_mno_afk is now known as michel_mno
<Onslauth>Hi guys, sorry to bother you, I was wondering if there might be a work around to a large let block causing a stack overflow?
<Onslauth>I have a let block with about 30K entries in it, and I get the following error: 2015-03-09T15:01:21 W cell load failed: vm-error, (vm-run VM: Stack overflow ())
<Onslauth>Is there a tunable I might be able to increase to solve this?
<lloda>see here
<wingo>30k entries, wow :)
<wingo>would be interested to see what master would do with that
<wingo>Onslauth: there is an env variable
<Onslauth>I already set it to 500,000 and it works now
<wingo>fwiw in master if a variable is not used any more, its slot is reallocated to another var, which will mean that your stack frame in master would be much smaller
<wingo>also, master has no stack limits
<Onslauth>Is there significant differences between 2.0.11 and master ?
<wingo>new vm, new compiler, yeah a bunch of stuff
<Onslauth>It good enough to use on production?
<wingo>probably not :)
<wingo>i use it, but it's much less tested
<Onslauth>I might give it a go later on
<wingo>good luck, let us know how it goes
<Onslauth>I do have a quick question
<Onslauth>At the moment we have to read values from a csv file, and create objects out of them
<Onslauth>What we currently do is read the file, write another file with something like (make <object> #:name x)
<Onslauth>Then load that file in with (load filename)
<Onslauth>Is that the fastest way to do it?
<wingo>there are always faster ways to do things, just a question of how much effort should be spent :)
<wingo>how does it work for you? is it frustratingly slow, or do you not notice?
<Onslauth>Doing it this way is super quick
<Onslauth>the files are about 40k lines, which generate a new file to load with about 35K entries in the let block
<Onslauth>I originally tried creating and object each time I read a line from the csv, and that took about 15 minutes give or take
<wingo>i guess you load the file more often than the .csv changes?
<Onslauth>They are in lock step, so csv is sftp'd in and then the new file is generated and loaded
<Onslauth>Do you have an estimated release date for the master branch?
<wingo>hard to say; we are ready to start prereleases, but probably it will be 9 months before a final release
<wingo>greets, civodul
<civodul>howdy wingo!
<Onslauth>Ok, thanks :)
<dsmith-work>Morning Greetings, Guilers
<nalaginrut>wingo: oops, is down?
*nalaginrut guess tekuti crashed
***dje is now known as xdje
<taylanub>yeah, 502 Bad Gateway from nginx
***michel_mno is now known as michel_mno_afk
***cluck` is now known as cluck
<wingo>civodul: soooo
<wingo>do we enable (texinfo reflection) now
<wingo>makes sense, no?
<wingo>i mean, we can enable it in master and maybe stable-2.0
<wingo>and make improvements incrementally
<civodul>what do you mean by "enable"?
<wingo>civodul: i mean, at the repl, (texinfo reflection) should have been loaded
<wingo>for its side effects
<wingo>so that texinfo docstrings are parsed
<wingo>and so that (help (foo bar)) prints docs for all exports of (foo bar)
<wingo>greets xdje
<civodul>wingo: at the REPL, yes
<civodul>makes sense to have it enabled by default
<civodul>in master at least
<wingo>or stable-2.0 too, dunno
<wingo>it's been there unchanged for a long time
<wingo>but whatevs
<civodul>yeah, sounds rather harmless
<wingo>i have some documentation to update but we should get on the 2.1.x release train
<civodul>that would be neat!
<wingo>it's ready except for those last doc things
<civodul>we also need that 2.0.12 out, but that's more for mark_weaver & i i guess ;-)
<wingo>heh that will be great too :)
<wingo>we are slack
<wingo>slack high-fives
<wingo>and on that note, off to zzz i think
<dsmith-work>civodul: Might I suggest a few 2.0.12 RCs before hand?
<civodul>dsmith-work: probably a good idea, yes