freenode/#lisp - IRC Chatlog
Search
1:47:47
KahMue
Hi there! Does anybody know how, in qtools, to connect a slot to a signal that provides Qt-type like QPoint. I need to connect a slot to a QWidget::customContextMenuRequested - signal, that provides a 'const QPoint & pos'.
8:15:47
splittist
no-defun-allowed: isn't it usually Kaplan? (How about ML Scheme - or is that too Maoist?)
8:24:35
no-defun-allowed
splittist: I would have picked someone like Makhno, who was on the receiving end of all that, as a namesake, but the almost-assassination felt more relevant.
8:25:48
phoe
minion: memo for KahMue: (declare (connected my-widget (custom-context-menu-requested)))
9:27:41
no-defun-allowed
I think the semantics of Mathematica are quite different to Lisp, and even to many languages. Something to do with pattern rewriting maybe.
9:28:47
no-defun-allowed
(And, maybe you can use it, but paying $344 for a programming language makes it hard for me to test.)
9:30:46
sarna
hi, can I somehow get a warning with this code? `(make-array 3 :element-type 'invalid)`
9:32:40
scymtym
sarna: in SBCL 2.0.1, i get a warning "undefined type: INVALID" when compiling that code
9:32:54
no-defun-allowed
It might be possible to write (some of) the evaluation rules in Common Lisp. https://reference.wolfram.com/language/tutorial/Evaluation.html describes it quite accurately.
9:49:45
no-defun-allowed
Yes, fexprs. They're not really macros as they don't generate new code, but they get all their arguments unevaluated and choose which ones they evaluate. That seems similar.
9:59:11
beach
Speaking of Mathematica, my thesis advisor offered our high-performance rewrite engine to Stephen Wolfram, for use in Mathematica, but Wolfram turned down the offer.
10:01:11
beach
Some day, I should try to implement our engine as a Common Lisp program. The compiler for it was written in Franz Lisp at the time, and it generated VAX assembly code. I am not sure whether it would be fast enough if Common Lisp code were to be generated.
10:18:11
beach
The rewrite engine? No, not particularly powerful. Just very simple and with very simple semantics based on rewriting, so it uses outermost evaluation. With this evaluation strategy, it is possible to define things like IF as functions, which is not possible with call-by-value languages.
10:38:33
no-defun-allowed
In a parallel universe where I asked follow up questions (such as "What constitutes a database in the programs you want us to write?"), I would have written an algebra system for my computing project last year instead of an accounting system. That could have been useful.
10:41:10
no-defun-allowed
Sure, but it would require less fluffing up than the need for compressing an "inefficient" text database and writing a compression algorithm.
10:41:23
no-defun-allowed
I haven't heard very many nice things about Matlab; why would you want to run it?
10:42:35
no-defun-allowed
beach: For Clomptroller? I only got a result for the whole unit (which also had my case study of a computer network) which was a B.
10:43:11
beach
no-defun-allowed: That sucks, i.e. that you don't know the results for individual parts.
10:43:39
no-defun-allowed
Maybe I can email my teacher and see if he filled out the rubric. I think he was supposed to.
10:46:41
no-defun-allowed
However, if that is representative of what I achieved with Clomptroller, I would say that I blame myself for a. picking a domain I had little prior knowledge in, b. not taking enough time to learn said domain prior to planning a program that uses it. I think that forcing the students to use the waterfall model was also detrimental to the programs produced, and I found some of the steps redundant (like writing
10:46:41
no-defun-allowed
pseudocode for the compressor when I had already written Common Lisp code for it).
10:47:56
no-defun-allowed
no-defun-allowed: Maybe you should wait until you have more than two things to say before numbering them. Nice try though.
10:51:23
no-defun-allowed
Then there were some other "impedance mismatches" where we had to write out test tables and some other laborious methods of testing, whereas I was more comfortable writing test suites; and we were supposed to interpret the "design" stage as interface design instead of protocol design, drawing mockups instead of writing mock interfaces, and so on. In conclusion, I was quite disappointed with my performance in that
11:40:20
no-defun-allowed
I should not be lamenting over below average results while half asleep (nor is it exactly on topic), so I'll head off for the night. See you later.
13:05:29
jmercouris
strangely though, *standard-output* -> #<SB-SYS:FD-STREAM for "file /Users/jmercouris/.local/share/next/standard-out.txt" {100936F393}>
13:06:14
jmercouris
it seems that I need to add an appender to log4cl to also append to *standard-output* or so