freenode/#lisp - IRC Chatlog
Search
4:36:59
stylewarning
Are you folks familiar with the POWER architecture? Is it similar enough to PowerPC for the existing SBCL or CCL to work on it?
4:38:07
jasom
stylewarning: POWER and PowerPC are largely synonymous; that being said if you're running on a ISA revision/OS combination not encountered before, a small amount of porting work could be needed
4:44:54
jasom
I think Power 8 and 9 are designed to be run LE for driver compatibility though; not sure if sbcl or ccl support LE power
6:12:14
beach
minion: memo for phoe: I started such a thing: https://github.com/robert-strandh/Spell
6:13:59
beach
parjanya: phoe was looking for Common Lisp code for an English dictionary for things like spell checking.
6:16:25
beach
I forget. It was some time ago. It is supposed to return some standard instances that give all possible interpretations of some word in English.
6:18:38
beach
I do need a spell checker though. I am getting increasingly dyslexic. So before I can abandon Emacs in favor or Climacs, or beirc, or ..., I need one.
6:19:14
parjanya
beach: btw, I’ve seen a "Lispy lists" mentioned on YT, is it a book, a file, what is it?
6:19:52
parjanya
beach: so far I could only imagine writing it in python : / my Lisp idioms are coming very hard to me
6:24:59
parjanya
beach: anyhow, this is what I saw mentioning "lispy lists": https://www.youtube.com/watch?v=iferkSFzkF8
7:32:57
whoman
of how complicated it would be to "translate" elisp into CL, i was thinking about this myself as i have some elisp code around
7:35:18
beach
I don't think it would be very hard. The tough part is to implement all the infrastructure that it takes to execute the resulting code.
7:35:55
beach
You could very well end up having to write a complete implementation of Emacs in Common Lisp in order for taht to work.
7:51:04
flip214
minion: memo for whoman: see https://github.com/adolenc/cl-neovim/issues/1 for some talking about "neomacs", which aims to get such a compat layer.
9:37:37
dxtr
oleo: That's literally exactly what I wanted but I couldn't quite find the correct stuff to poke at
9:38:51
oleo
i either use rlwrap sbcl for getting completion and history (you have to create the completion and history files too ofc)
9:41:28
beach
dxtr: If you don't get the CL-USER> prompt, there is something wrong with your SLIME installation.
9:42:48
beach
dxtr: Do you see any error messages, or any other messages that indicate that something is wrong?
9:43:55
beach
By default, you should get a buffer named *slime-repl ...* and it should have that prompt.
9:47:04
oleo
dxtr: i'm not using emacs at all by the way, i start right into mcclim and use it's editor climacs too
12:34:04
phoe
mingus: memo for Xach: https://github.com/phoe/fox5/blob/6d8f73882be6ddaf1f94d9a89153a38f1b0160fc/skippy.lisp
12:34:04
minion
phoe, memo from beach: I started such a thing: https://github.com/robert-strandh/Spell
12:54:18
phoe
Good. I will try to write an interface for GCIDE and once I have it, we can think on how we can link the two together.
12:57:30
phoe
minion: memo for Xach: https://github.com/phoe/fox5/blob/6d8f73882be6ddaf1f94d9a89153a38f1b0160fc/skippy.lisp
13:13:50
_death
phoe: the one with copy-seq.. why not allocate all the frames in the beginning and just write to each one
13:14:57
phoe
so it's logical to me that for drawing frame N, I will want the result of drawing frame N-1, which depends on result of drawing frame N-2, ..., all up to frame 1
13:15:19
phoe
so it's logical to me to draw everything on one frame and then just collect their copies as results.
13:16:12
phoe
could do, sure thing - except I don't see any obvious optimizations here, sans for saving one frame's worth of memory
13:17:47
_death
it should be much faster.. I've been doing image-related stuff w/ Lisp for some years :)
13:19:38
phoe
should be? why would allocating N images from the start and writing to them be much faster than copying an array N times after writing to it?
13:21:34
phoe
it works well enough for me so far, maybe I can make it faster - but I consider it fast enough so far.