***Guest67293 is now known as lispmacs
<lispmacs>how do you read/set the current working directory in Guile? <lispmacs>i embedded guile in a C application, and registered some primitive functions with scm_c_define_gsubr <lispmacs>can I get those primitive functions in a module that can be loaded by other modules I create? <lispmacs>I don't see like a scm_define_module function <lispmacs>problem is I create a module of helper functions, but thhe helper module does not see the primitive functions when I use it <lispmacs>i suppose I could load a file that defines a module that exports those primitive functions? <lispmacs>or, i guess, since it is scheme, I could embed that file in the C file and eval it <enderby>hi, wondering what the correct way to document functions in scheme/guile is ***ng0_ is now known as ng0
<dsmith-work>lispmacs: The typical way to expose C funcs as a module, is to also have a companion .scm file. <dsmith-work>lispmacs: Though writing stuff in C less desired these days. Scheme is better, and the comiler is gettting more awesome all the time. <d4ryus>is there any guile dbus library or bindings for libdbus? <str1ngs>d4ryus: if you use Gio then possibly guile-gi or g-golf would work for you <d4ryus>ok, guile-gi and g-golf looks interesting, thanks! <str1ngs>if you need to use signals guile-gi is the most mature of the two. <str1ngs>also you can still use guile-gi and g-golf and make your own typelib. that would be a last resort though <dsmith-work>d4ryus: My mistake. It's guile-avahi I'm remembering, which *does* use dbus, but it's the avahi code that uses it. <malaclyps>d4ryus, if you get something working, I'd love to see it. I played around getting dbus to work with guile, but eventually just ended up parsing qdbus's command line output <d4ryus>malaclyps: funny you say that, i just started doing exactly that :D <ArneBab>wingo: re encoding 1.8/2.x: ah, yes, that’s going to make a difference <ArneBab>wingo: I should re-do the comparison with LANG=C LC_ALL=C — don’t know when I’ll get to that <wingo>ArneBab: LC_ALL=C isn't great either fwiw <wingo>the optimized cases are iso-8859-1 and utf8 <wingo>what if... what if we fix exception types in 3.0 <wingo>like, native exceptions are condition objects instead of key plus args <wingo>already we had to implement a wrapper for r6rs. <wingo>i would prefer to layer r7 on top of guile rather than on top of r6 fwiw <wingo>so that would suggest pushing conditions more towards guile core. <ArneBab>wingo: current exceptions are hard to use; an improvement sounds good, as long as compatibility with r6rs and r7rs is possible.