IRC channel logs

2014-05-06.log

back to list of logs

***sethalve_ is now known as sethalves
***fangism is now known as fangism-ctrl-Z
<nalaginrut>morning guilers~
<civodul>Hello Guilers!
<zacts>hey can guile do pcre?
<ArneBab_>zacts: google says yes: https://github.com/hdasch/guile-pcre
<zacts>but it's not part of core guile
<zacts>hm.. that's cool
<ArneBab_>sneek: later tell mark_weaver: I created a first rough draft of an SRFI for wisp: http://draketo.de/proj/wisp/srfi.html
<sneek>Okay.
<ArneBab_>sneek: botsnack
<sneek>:)
<wingo>moin
<ArneBab>moin wingo
<nalaginrut>I wonder why there's no SRFI for crypto
<ArneBab>that’s true, yes. Especially given the prevalence of crypto in academia
<nalaginrut>yes
<nalaginrut>but seems there's no any progress after 2004
<nalaginrut> http://srfi.schemers.org/srfi-discuss/mail-archive/msg00080.html
<nalaginrut>and srfi for SQL
<nalaginrut>I found my ssql is similar to SchemeQL, I'll take closer look
<ArneBab>I just stumbled over deferrred evaluation: I there’s a difference between (vector-set! world target ( + target-value change))(vector-set! world source (- source-value change)) and ordering it the other way
<ArneBab>diff is calculated via (let* ((change (* A source-value))))
<ArneBab>maybe that wasn’t actually the error, rather a bug in my code…
<wingo>what good would a srfi for crypto do? better to have a good implementation than a standard.
<ArneBab>portable crypto would be nice - but yes: good implementation first
<nalaginrut>agreed ^
<ArneBab>sorry for the earlier noise: I had a bug in my code.
<nalaginrut>I have to copy/paste crypto modules for my projects every time, it would be better if it's srfied
<wingo>being srfi is no guarantee of quality, and can work against quality by imposing unnecessary restrictions
<ArneBab>by the way: I finally managed to create a draft of a srfi for wisp: http://draketo.de/proj/wisp/srfi.html
<nalaginrut>well, I think it's inevitable to compromise for standardization
<ArneBab>nalaginrut: it would also be possible to standardize the best implementation
<ArneBab>you just need to agree on what is bestb
<ArneBab>and that’s not easy…
<nalaginrut>yes, I agree
<wingo>compromise is a bad thing in a crypto library...
<nalaginrut>hmm...yes, there's more to be considered except efficiency
<nalaginrut>maybe I should constrain my words to digest hashing functions...
<nalaginrut>this sounds easier
***fangism-ctrl-Z is now known as fangism
<stis>Hmm compiling master to play a little with wingotoys
<dsmith-work>stis: Be prepared to wait longer.