freenode/lisp - IRC Chatlog
Search
12:41:26
beach
jmercouris: I believe it is good in that it allows for several views into one buffer, and each view can have a different way of presenting the contents, including a different parser for the buffer contents.
12:41:57
jmercouris
I'm going to attempt to replace our own buffer implementation with one of the supplied ones
12:43:10
jmercouris
what's strange to me is that you can insert something like (cluffer:insert-item y "aj")
12:43:54
jmercouris
I don't fully understand the principles behind the implementation yet, but I at least can use simple-line so far :-D
12:46:50
jmercouris
maybe I will figure it out in time, anyways, with your blessing, i will try this
12:46:56
beach
Anyway, standard-line uses a gap buffer which is efficient. simple-line uses a list as I recall.
12:47:28
beach
Sure, I'll try to assist as much as I can. I may need some time to re-read the code to answer questions.
16:33:26
jackdaniel
hm, now that I think about these timeouts I think that we should not put a timeout in acquire-lock
16:39:01
jackdaniel
ralt: also, pointing out potential problems in the code is something to be thanked for, not to be sorry about, hence thank you! :)
16:39:46
ralt
It's more the "point out a problem, point out another problem in the next iteration" thing I'm apologizing for
16:41:35
ralt
jackdaniel: your argument probably works in the general case but starts failing in the case of under pressure environments
16:42:46
ralt
Well, if you do the timeout for lock, then every signature needs to have the timeout, including the count ones
16:43:58
ralt
Just spawn 1k threads reading the value, another 100 threads modifying it after taking a lock, and see if everything dies
16:44:59
jackdaniel
regarding timeout for count ones, why would that be necessary? if I call mailbox-empty-p I would assume, that it will wait until the lock can be taken
17:01:31
axion
I'm having a lot of fun working on a new library for a change for creating generative art :)
17:09:27
axion
Nah, though I started working on a port the same author's ConvChain a while ago, which can be guaranteed to terminate, unlike WFC :)
17:10:20
axion
Just doing basic noise algorithms and a composition pipeline https://gist.github.com/mfiano/0ae825095bd26c4b6eda3e203b38da7d
21:33:25
jmercouris
I'm looking at Qlot right now, at this function specifically https://github.com/fukamachi/qlot/blob/master/install/quicklisp.lisp#L16
21:34:46
jcowan
How can I find standard functions that return two values, either the result or #t if there is a result, or #f #f if there isn't one? hashref works like this, but what else
21:37:35
jmercouris
(ql::http-fetch "http://beta.quicklisp.org/quicklisp.lisp" #P"/Users/jmercouris/Downloads/quicklisp-JBCUCRBI.lisp")
21:38:05
jmercouris
can anyone confirm if ql:http-fetch is working on their machines as I expect it to be for myself?
21:38:45
jmercouris
I also tried the variation: (ql-http:http-fetch "http://beta.quicklisp.org/quicklisp.lisp" #P"/Users/jmercouris/Downloads/quicklisp-JBCUCRBI.lisp")
21:39:12
_death
function-lambda-expression returns 3 values, that you can think of as generalized booleans.. same goes for get-properties
21:40:23
Shinmera
Not sure about standard functions, but a bunch of libraries make use of that convention.
21:49:38
seoushi
http-fetch works for me. Had to be in the ql package tho since http-fetch is not external