freenode/#lisp - IRC Chatlog
Search
1:55:06
theemacsshibe[m]
This is my server. It costs $400 to run for 12 days and runs at 1.2 gigahertz with 8 SPARC III cores.
1:59:42
theemacsshibe[m]
Alternate impression: I would tell you about my servers, but I can't get Ethernet working.
2:39:23
luggager
lispers may (or not) be interested in these fascinating reactions and discussion of the lisp machine {"and" :[{"rel" :[{"name" : "print"}, [{"name" : "x"}], {"host" :[{"name" : "print"}, [{"name" : "x"}]]}]},{"pred" :[{"name" : "print"}, [2]]}]}
3:35:21
drmeister
The CLHS says "If displaced-to is non-nil, make-array will create a displaced array and displaced-to is the target of that displaced array. In that case, the consequences are undefined if the actual array element type of displaced-to is not type equivalent to the actual array element type of the array being created. "
3:36:02
drmeister
I'd like to implement something where I can create displaced arrays of element type byte8 into other arrays.
3:36:57
drmeister
This is to write out the memory contents of a particular array - like write out the bytes of a specialized array of floats.
3:39:06
drmeister
Bike: I need to have an array of byte8 to pass to the pzmq library - I can't control how they are written out.
3:44:26
drmeister
Currently I get an: (ERROR "Cannot displace the array, because the element types don't match") if I try it.
4:42:54
iqubic
Is there a emacs lispy tutorial I can go through anywhere? Like an interactive tutorial I can play around in?
6:04:38
phoe
drmeister: since the consequences are undefined, you can define them yourself. In particular, you can do something that in C would be more or less char** array_of_bytes = ...; float** array_of_floats = (float**)array_of_bytes; which is just casting.
6:05:52
phoe
I also imagine that it would be sane to keep an error message in that case, because Lisp is strongly typed compared to C, and I think that most cases where one would want to displace array of elt-type A to array of elt-type B are simple mistakes and not voluntary actions.
6:06:47
phoe
Now that I think of it, I'd add a new keyword arg to MAKE-ARRAY, something like :CASTING-DISPLACE-TO that works like :DISPLACE-TO except it does just what you describe on the low-level.
7:01:55
Chream_
TMA: was a bit harsh or? i guess someone here in #lisp would know where an emacs tutorial is?
8:17:49
beach
Chream_: Sure, but there are a lot of very knowledgeable people in #lisp, so if we were allowed to discuss any topic that someone might know, then this channel would drown in off-topic discussions.
8:23:36
SAL9000
phoe: Yeah, I'm aware. IRC is sometimes more convenient for real time conversations though, thus my question
8:24:28
Shinmera
And you can reach didier on twitter, which may or may not be faster in case of problems.
8:29:13
TMA
Chream_: sure. those probably frequent #emacs too. I am still of the opinion, that iqubic would be better off asking there, probabilistically speaking -- there is a greater proportion of those willing to discuss things emacs there than it is here. I might have worded it too terse. It looks like iqubic was helped there faster anyway.
11:25:47
beach
It is a somewhat labor-intensive task, so if anyone feels like helping out, that would be great.
11:27:20
beach
What I have decided to do is to modify the text in place, and then refer to some appendix for the original text whenever I decide to update some text in the spirit of WSCL.
11:31:15
phoe
beach: Yep, I see, except you're creating a modification of the standard where I'm creating a clarification of it.
11:32:51
beach
Some TeX macro calls, like beginSection etc. I just replace by standard LaTeX macro calls.
11:33:24
beach
Some others, like semantic markup things like newterm, I define a macro for and don't modify the original.
11:40:05
phoe
beach: yep, fire up the issues on GitHub that can be picked up by other people if they want to.
11:40:44
beach
blurgh: Please, we don't need to know what time it is in every time zone in the world.
11:56:09
phoe
Will be in Malaga at 18:00 my time, it'll take me a little bit longer to go to Marbella.
12:02:47
beach
Sure. We'll be there around 14:00 or so. Hotel Villa Marbella. Show up for a beer if you like.
12:03:35
phoe
So we'll end up in Marbella later. I still don't know what the bus travel is from Malaga to Marbella.
12:33:11
phoe
PCL is a good place if you already know some programming. Get yourself a Portacle for an all-in-one Common Lisp development environment.
12:36:00
phoe
the most essential one is SLIME. the best way of getting it is installing Quicklisp first, and then (ql:quickload :quicklisp-slime-helper)
12:36:03
Shinmera
Regarding ELS I'll be arriving in Malaga at ~9:00 and with the Bus I should be in Marbella at around 10:30
12:37:28
phoe
He's mostly talking good stuff, though several things he said in that book are simply wrong.
12:37:59
random-nick
why is installing SLIME through quicklisp preferable to installing it through ELPA?
12:38:02
phoe
Scheme is oriented towards functional programming where CL is as multiparadigm as a language can get.
12:38:35
phoe
random-nick: it'll get upgraded with each Quicklisp upgrade. For some reason, I prefer that to the mess that I often ran into with ELPA/MELPA.
12:39:01
phoe
Shinmera: the last time I attempted to troll on #lisp you were the first one to complain about it
12:39:34
White_Flame
"Lisp" is the name of a family of languages which includes Scheme, but "LISP" is also the name of the first implementation, and languages whcih derived from it and recombined into "Common Lisp" are still basically known as "Lisp" when no other modifier is involved
12:41:56
siraben
I've already ordered Practical Common Lisp, should arrive in a couple days. I like how it very, well, practical and that Lisp isn't just for cool meta-circular evaluators :)
12:44:55
makomo
\h is supposed to include both the "classical" horizontal whitespace and some unicode stuff like no-break space, etc.
12:45:01
siraben
What can I expect to be able to understand well after reading through Practical Common Lisp?
12:45:17
phoe
in 100% conformant ANSI CL, we have a perfect bijective mapping between lowercase characters and uppercase characters; Unicode does not implement anything like that.
12:45:20
makomo
i mean, i doubt i'll have documents with such horizontal space, but having it for completeness would be nice
12:45:35
phoe
siraben: reading, no idea; actually following through with the exercises, you'll grasp the basics of CL
12:50:03
smokeink
I'm currently reading the last 2 chapters, about making interpretors and compilers in common lisp
12:50:39
smokeink
so far I've only read the first few chapters, now I'm reading the last few chapters
12:51:39
smokeink
cuz I don't need to understand the Eliza talking bot right now , but I want to understand how this thing works: http://lisperator.net/slip/ , it seems to be based on PAIP
12:53:45
Bike
paip is mainly about historical AI. it has some basic stuff about interpreting and compiling lisp, but you'd need to know more t han that for an actual implementation
12:54:55
smokeink
what other books do you guys recommend , for understanding how to make a lisp ? Lisp in Small Pieces ?
12:55:18
drunk_foxx[m]
I've started PAIP and On Lisp in parallel after Land of Lisp, seems like a great flavor to me. There are also quite great excercises in the book so that you don't just read and repeat
12:57:15
drunk_foxx[m]
It's especially funny to compare it to some of the modern bots, and understand that actually not much progress has been made since those days
12:58:07
smokeink
can SBCL easily be ported to WebAssembly (which funny enough, is based on S-Expressions) ? https://www.youtube.com/watch?v=yEYtwmI7bDg ( WebAssembly hello world )
13:00:17
p_l
WebAssembly could be generated from SBCL's compiler, but AFAIK there were some issues in making it a lisp friendly place
13:02:15
p_l
I don't remember everything, but I seem to recall some of type checking and function call semantics were iffy?
13:02:49
p_l
That said, you could probably port a CL implementation to wasm, maybe not necessarily SBCL
13:05:19
siraben
Siri is really bad, any deviation from a specific phrasing makes it say "Sorry, I don't understand" or other canned responses
13:06:15
drunk_foxx[m]
> drunk\_foxx[m]: Why has work on Eliza-like assistants gone dark after the 1990s?
13:06:54
nirved
back at the end 90s i came to the conclusion that human mind state can be represented in 50-70 dimensions (each dimension is a duality), haven't checked if this is correct though