freenode/lisp - IRC Chatlog
Search
15:21:58
pjb
shrdlu68: if your file contains (cl:defun cl-user::foo (cl-user::x) (cl:+ cl-user::x 42)) and assuming no foul play with the *readtable*, then you can set the *package* to any package, loading the file will always read the same sexps!
15:22:40
pjb
shrdlu68: Similarly, as soon as the file contains a toplevel in-package form, it controls the package in which the following unqualified symbols are interned.
15:23:26
pjb
dlowe: usually, people write (in-package "BAR"). But if FOO::IN-PACKAGE is not CL:IN-PACKAGE, (in-package "BAR") can have different effects.
15:49:16
slightlycyborg
Hi. I have x predicate calls. What is the best way to sum up the number of predicates evaluated as true?
15:53:44
dlowe
I was trying to think of a way that didn't involve a temp list, but I didn't think of anything without macrology
15:55:44
loke
slightlycyborg: If every comparison uses the _same_ pred, you want to do: (count-if #'pred LIST-OF-VALUES)
15:57:16
slightlycyborg
Ya, the #'identity raised an interupt in my mind and I saw that pattern. I am not using the same pred for this instance though.
15:58:31
dlowe
If every comparison uses the same value and different preds, (count-if (lambda (p) (funcall p value)) LIST-OF-PREDICATES)
15:59:05
dlowe
(defmacro prog-count-if (&body forms) (let ((counter-sym (gensym "COUNTER"))) `(let ((,counter-sym 0)) ,@(mapcar (lambda (f) `(when ,form (incf ,counter-sym))) forms) ,counter-sym)))
16:01:36
dlowe
(tagbody 5 (defvar count 0) 10 (if (not (pred x)) (go 30)) 20 (incf count) 30 (format t "I am so cool. count=~a~%" count))
16:02:18
Xach
slightlycyborg: this steady drip-drip-drip of new info is excellent for stretching out the discussion and exploration of solutions!
18:29:52
fiddlerwoaroof
(length (flatten (mapcar (lambda (f v) (if (funcall f v) (1) ())) functions values)))
18:37:41
_death
jackdaniel: heh.. at first I was puzzled.. TRON operator? then I recalled TRON/TROFF...
20:53:30
aeth
LdBeth: If what you said was valid then it would constantly be setting a to 1 and then a to 2 every iteration of the loop. It doesn't seem to be valid in SBCL, it says "duplicated variable". verisimilitude has the answer.
20:53:44
verisimilitude
In any case, you should be using the integer stepping clauses for this, instead.
20:54:51
aeth
(loop :repeat 5 :for a := 1 :then (1+ a) :do (format t "~A~%" a)) is equivalent to (loop :repeat 5 :for a :from 1 :do (format t "~A~%" a))
20:55:11
aeth
The advantage of the second form is that you can add a terminating condition right in it instead of writing your own elsewhere.
20:56:03
LdBeth
well, Emacs' cl-loop don't signal an error on duplicated variable and just complains "unused variable a" when byte-compiled though
20:56:10
aeth
You can also count 0-based and by steps greater than 2. (loop :for a :from 0 :below 10 :by 2 :do (format t "~A~%" a))
20:57:39
aeth
LdBeth: The way I would have interpreted it if it was valid would be to have a set to 1 and then incremented by 1, so both a's would be used, sort of like (let* ((a 1) (a (1+ a))) a)
21:23:04
fiddlerwoaroof
it has most of the features of CL's generic functions, but adds &context arguments and a couple additional kinds of specializers
21:23:34
fiddlerwoaroof
the &context arguments let you dispatch based on things like the currently active mode, etc.
0:43:31
slightlycyborg
I am trying to get asdf to load a system that I placed in ~/common-lisp. I load quicklisp on sbcl boot. I am getting "component <system_name> not found" after (asdf:load-system "<system_name>")
0:46:31
slightlycyborg
I think I found the error. I named my file <system>.asdf instead of <system>.asd
3:01:09
hectorhonn
verisimilitude: hello verisimilitude. cl just keeps surprising me with its features
3:01:29
verisimilitude
It's a rather comprehensive language, as it was designed to unify several dialects.
3:01:54
verisimilitude
Being practically oriented, it also includes many facilities that certain programs would require; the various time facilities are a good example.
3:02:56
fiddlerwoaroof
restarts, multimethods and method combinations were the features that most attracted me.
3:03:10
aeth
More importantly, CL was designed to be (just barely) efficiently implementable on 1990s hardware.
3:04:03
aeth
CL has two decades of Moore's law over newer languages of comparable features since people tend to target the hardware they have right now.
3:05:04
aeth
hectorhonn: generally the advice on IRC is to not ask to ask but to ask because IRC is fairly slow moving for a chat protocol.
3:05:08
verisimilitude
Common Lisp has certain restrictions in place that let a concerned implementation optimize well. Newer languages, such as Python, have been ``designed'' without this foresight and the results are clear.
3:06:06
aeth
verisimilitude: Interestingly, Python's only 6 years newer than CLtL. I think one of the key differences is that by 1990, C had won so languages were going for niches (like scripting) whereas CL was designed for the whole stack.
3:08:10
hectorhonn
if, in a macro, i do (let ((sym (gensym))) `(defun ,sym () (...))), will this defun get garbage collected at the end of the macro?
3:09:08
fiddlerwoaroof
But, if there are no references to the name, I think it will eventually get gced
3:10:01
verisimilitude
It would only be garbage collected once it becomes unreachable, but that's not enough.
3:10:21
verisimilitude
I'm just letting him know it's possible the symbol is saved, which would prevent this.