freenode/#lisp - IRC Chatlog
Search
20:16:42
Bike
you could think of those java and C++ functions as being multiple functions with different types that happen to have the same name. the compiler can unambiguously figure out which definition a particular call will use. lisp has apply so in general it can't. of course you could still make it mostly free with some pretty basic compiler macroing
21:56:13
Shinmera
And the award for Longest Symbol Name Ever Devised goes tooooo.... ASDF (of course), with ASDF/BACKWARD-INTERFACE::*DEPRECATED-FUNCTION-STYLE-WARNING-ENABLE-ASDF-BINARY-LOCATIONS-COMPATIBILITY-NOTIFIED-P*
22:15:09
jcowan
beach: That was a very old version of the model. The latest version allows arbitrary objects (in the Lisp sense) as primitives.
22:17:09
gabc
Hi! I'm trying to understand conditions, and I keep getting an "Unbound condition slot" error. I reproduced it here, can anybody can see the issue? Thanks! https://plaster.tymoon.eu/view/2369#2369
0:20:02
MrtnDk[m]
I got my common lisp system up and running with emacs, slime et all on my raspberry pi, thanks to some of your guys. Now I am trying to get it running on my laptop. I installed sbcl, slime (from both ubuntu and melpa), but I can't seem to get the repl running. I'm sure I'm forgetting something basic.
0:24:14
MrtnDk[m]
Searching for program: Ingen sådan fil eller filkatalog, lisp (No such file or directory, lisp)
0:34:12
johnjay
iirc don't you have to issue two commands, one to start hte repl and one to start slime-mode?
0:35:04
MrtnDk[m]
johnjay: I evaluated the spell I got from Xach, and then did "M-x slime" in the lisp buffer, i guess.
0:39:52
MrtnDk[m]
Xach: Yeah, but my system doesn't ... which is why I translated the error in parens. 😀
0:40:43
MrtnDk[m]
johnjay: Strange. Slime is enough for me, but I have to C-x C-o back to the lispy buffer.
0:41:26
MrtnDk[m]
johnjay: Yeah, "filkatalog" is "file catalog" in danish, which now a days is known as "directory" in English.
0:42:13
MrtnDk[m]
johnjay: Oh ... my program is very small .. maybe it is compatible with Emacs-lisp and didn't realise .. .let me test.
0:44:47
johnjay
if you open a file test.lisp and just put (+ 2 2) in it and do M-x slime what happens if you do C-x C-e? the same?
0:45:25
MrtnDk[m]
I wonder if I did something else. I put the (setq ...) thingy in my init file, so I will try and restart emacs, and do that. ... I will save test.lisp first though.
0:46:03
johnjay
i'm not as concerned because it seems the lisp process in the repl is the same one it's using for C-x C-e and such
0:46:58
MrtnDk[m]
Yes .. it works .. but maybe because I have swank (or what it is called) running already?
0:50:05
MrtnDk[m]
It still works. I do have some slime in my initfile, although I don't remember putting it there.
0:51:48
MrtnDk[m]
Hmmm .. it just seems to be amount a list of packagenames sent to "package-selected-packages".
1:09:38
MrtnDk[m]
johnjay: I installed both as Ubuntu package and using "M-x package-list-package", locate slime and install it. That requires you to have configured Melpa in your init file though.
2:08:51
fiddlerwoaroof
jcowan: yeah, at this point I've learned half a dozen different ways of writing programs, and I've generally noticed that each style has people who say it's "only good in theory", people who wonder why anyone would use anything else and people running large systems on the style.
2:10:17
fiddlerwoaroof
SQL, in particular, isn't my favorite language, but once I understood the relational model somewhat, it can be really productive to learn how to model and solve certain sorts of problems in that paradigm
3:19:42
beach
For the record, I am totally convinced both about the beauty and the usefulness of the relational model. I don't think I have said anything else.
6:09:44
beach
I just watched a talk on YouTube entitled "Soldiers and Scouts: Why our minds weren't built for truth" by Julia Galef. I think her research might provide insight into the reason for the dismal state of the software industry, and software development in general.
6:09:46
beach
In particular, she reiterates a result that I have seen before, namely that a high degree of education in a domain makes people MORE LIKELY to misinterpret information so as to support their existing point of view.
6:12:06
beach
I am sure this research has some interesting implications with respect to the use of Common Lisp, both in industry and in academia, but I am not yet sure what those implications are.
6:42:52
flip214
beach: because then people have "more to lose" (more data, more education) if their viewpoint was wrong
6:47:22
beach
flip214: That might be related, sure. I am not sure the full explanation is as simple as that, though.
6:49:35
beach
I am more interested in the consequences. It means that people with more knowledge about programming-language semantics, automatic memory management, compiler design, language design, software engineering, etc., are less likely to use that knowledge accurately to make an informed choice of what programming language to use.
6:50:15
beach
... at least if they have already formed an opinion in the domain, like people with many years experience in some particular language or programming style.
6:50:17
Shinmera
beach: There's also been studies showing that with political issues highly educated people are /less/ likely to change their minds when presented with contradictory evidence.
6:51:22
White_Flame
also, computing is just a job now, not an enthusiast pursuit of exploration and doing better things with these universe-creating boxes
6:52:23
White_Flame
those with the older viewpoint of exploration and advancement are a stark minority
6:53:09
White_Flame
yep, but it has more active effects on the average person, who now has computing devices in their primary activities
6:54:50
White_Flame
in terms of people not having reasonable evaluations of the state of the art & overall direction
6:55:23
beach
I am thinking this: If training programs like university teaching programs teach languages such as C++ or Java, and the trainees get used to those languages, then giving them even more training is going to make them less likely to want to switch.
6:57:08
White_Flame
2) I want to do X, what tool do I need to accomplish it better than my existing ones?
6:59:03
flip214
White_Flame: well, you can solve _anything_ with a turing-complete language... doesn't mean it's a good idea
6:59:47
White_Flame
I think the main question here is when is it best to try to introduce new languages?
7:00:16
White_Flame
I think people who grew up on assembly, BASIC, COBOL, etc, were left hoping for even greater things
7:00:53
White_Flame
while with modern languages, a beginner could go a long way before finding lack in the language itself, for the purposes they're using