<lfam>How can I make Guile not abbreviate backtraces? <apteryx>lfam: there's supposed to be a COLUMN environment variable taken into account, but I haven't been able to make it work. <apteryx>Perhaps because on Guix most Guile code is run by the daemon so you have to inject this variable into the daemon's environment (service). <spk121>apteryx: it is "COLUMNS", not COLUMN <lfam>apteryx: I have set COLUMNS=9999 in the systemd service file but it doesn't have any effect :/ <lfam>I figured out the problem I was having. I don't need to read the backtrace anymore. But I wish I knew how to view it when I need to <lfam>It's possible that it gets reset by Guix somehow <lfam>I see that it gets set to 200 in (guix self) ***rekado_ is now known as rekado
<rekado>alextee[m]: instead of (when (not …) …) you can use (unless …) <rekado>alextee[m]: you also don’t need the begin on line 119 because that’s implicitly provided by “when” <rekado>same two points on lines 124 and 125 <rekado>if you want to improve option handling somewhat (it may not be worth it for a short script like this) you can take a look at srfi-37 args-fold <rekado>but I think your script looks fine <alextee[m]>thanks! i'll use those anyway just so i have them there for future reference <rekado>I haven’t used guile-config myself yet, but I always wanted to give it a try. <alextee[m]>it looks really useful for an actual program with --help and --version and all that <rekado>yes, and I think the declaration looks nicer than the somewhat cryptic way srfi-37 options are declared. <chrislck>alextee[m]: now is a good time to learn emacs which will multiply your .scm productivity :) <hugo>Did my message about him and emacs get delivered? I seem to have some problems with my irc app ***mood_ is now known as mood
***sneek_ is now known as sneek
<civodul>there's no standard bytevector-append, alas <brendyyn>does that alas mean you think there should be? <brendyyn>ok well i thought there might be a Good Reason for it. How can I define that procedure then? <brendyyn>do i have to just make-vector a new one, and measure all their lengths etc...? *brendyyn notes Racket has vector-append <mwette>create bv3 with size(bv1)+size(bv2), then use bytevector-copy! <brendyyn>I'm translating some python code into guile. I've never written python though. it's an interesting experience. <civodul>you could define a GOOPS method if you want to use + for everything :-) <spk121>I wonder how hard it would be to extract g_module from GLib to replace libltdl? I find libltdl's error message uninformative, but, unfortunately libtool is effectively unmaintained <wingo>spk121: i would dearly love to kill our ltdl dep <wingo>i gave it a look at one point but it was gnarlier than expected <mwette>civodul: at least dynamic-link works on my guile on macOS; let me check lower level <spk121>civodul: google says there is a macos man page for dlopen() ***sneek_ is now known as sneek
<spk121>brendyyn: the README for guile-gi should now be closer to correct for Guix <mwette>verified on 10.14.6, program using only /usr/bin/gcc, and dlopen("/path/to/lib", RTLD_NOW) <spk121>gnulib's dlopen would get your everything common except mingw & native windows <brendyyn>spk121: do you think my implementation is good? <brendyyn>for bytevector-copy!, eldoc is only showing the first 3 arguments in the modeline ***oni-on-ion is now known as ear-the-art
<bricewge>civodul: Are there any non blocking procedure to get bytevector? I can't manage to find one. <civodul>bricewge: not as such, but see "suspendable ports" <mwette>bricewge: maybe (ice-9 futures) can help <mwette>oops, missed the context -- nevermind