IRC channel logs


back to list of logs

<rlb>mala: haven't looked at it carefully, and don't know offhand exactly what's preventing your code from working, but in case you hadn't found it -
<mala>rlb, gosh that is both useful and kind of terrifyingly complex
<rlb>I suppose subprocesses are generally somewhat tricky, but not sure if there's any simpler approach for this case (in part, since I still haven't investigated why your attempt didn't work)...
<jpoiret>mala: your with-error-to-port redirects current-error-port to a port that is not backed by a file descriptor, and thus will get dropped and replaced by /dev/null when spawning a sub-process
<jpoiret>you need to open a pipe instead
<chrislck>sneek: botsnack
<chrislck>\o/ \o/
<unmatched-paren>does guile provide constants for pi and e anywhere in the standard library? searching the manual for / pi / and /euler/ turns up nothing of use
<unmatched-paren>or do i just define them myself as approximations?
<unmatched-paren>one of the manual's examples contains a (define pi 3.14...) line, so i'm guessing so
<dthompson>unmatched-paren: nothing defined in guile. I define them in my own projects when needed.
<RhodiumToad>how precise do you want the value? :-)
<RhodiumToad>(* 4 (atan 1)) is likely good enough for most purposes
<dthompson>guile's partial evaluator compiles (* 4.0 (atan 1.0)) to 3.141592653589793 :)
<RhodiumToad>which is correct to that number of decimals...
<dthompson>just a fun fact that peval handles trig functions :)
<unmatched-paren>i don't really need it incredibly precise, but that atan approximation is good to know
<unmatched-paren>(i'm just messing about in the repl, really)
<RhodiumToad>the atan result should be as precise as the representation allows, in this case it's representing an inexact number using a 64-bit float
<RhodiumToad>(acos -1) is also available
<RhodiumToad>hm. there's no builtin way to do a square root with better precision than floats?
<mala>jpoiret, thank you!
<sevan>hi, I've almost got Guile 3.0.9 built on ancient OS X 10.4 running on powerpc. I have 2 questions, does Guile intentionally require a compiler with C11 support? (it has a redefinitions support requirement which C99 compiler will lack). I'm currently stuck on how to provide an implementation of dprintf(3) or if it's necessary, build gets to linking and breaks since it's not there on ancient OS
<sevan>X. I see an m4/stdio_h.m4 which is meant to check for its declaration but I don't see anything in config.log
<sevan>I'm building with GCC 5.5 which has C11 support, since GCC 4.0.1 supplied by Apple only support C99 and so chokes a couple of places where libguile/scm.h is imported and there's a redefinition in the file importing.