freenode/lisp - IRC Chatlog
Search
17:23:48
phoe
God damn it, I'd have sent an important paper today instead of waiting for the damn notification that wasn't coming.
17:25:30
beach
You are not supposed to wait around for the notification. Normally, you should have plenty of time to accommodate the referee remarks (provided your paper was accepted).
17:26:39
flip214
phoe: If you'd like a review of the paper (what is it about?), I may be volunteering. beach knows my performance, if you need some second opinion about me.
17:27:46
phoe
So I don't know what I should write on the paper to my university - whether I'm going as a speaker or an attendee.
17:27:46
flip214
well, perhaps you want another independent look anyway, and be it only for next year's try?
17:29:06
phoe
And I have a paper written since I'd like my uni to help me with my plane tickets and so on, since they have a budget for that sort of thing - except I don't know who I am going to be at that conference.
17:29:33
phoe
flip214: yes please. I'll make a few final fixes today and send you a link to the paper.
17:31:35
beach
phoe: What do mean by "I don't know who I am going to be at that conference"? Do you have multiple personalities?
17:38:02
phoe
My university might/will have different reactions depending on whether I'm an attendee or I'm a speaker.
17:39:40
phoe
Except now I'm in a limbo - if I write that I'm an attendee while it turns out that I'm a speaker, I'll be missing out on higher rates, but if I write that I'm a speaker while it turns out that I'm an attendee, I'll be lying in formal papers and will need to straighten that out.
17:41:37
beach
phoe: Maybe you could ask the program chair for an advance notice, citing your conundrum.
17:43:11
phoe
Well, yes, that's a point. Especially now that they've silently extended the date of notification.
17:48:26
beach
"Dear Program Chair, I have submitted a paper to ELS (submission number #), and expected the notification of acceptance yesterday. It was not until today that I noticed that this date has been moved. I depend on a contribution from my university in order to have the means to travel to ELS, and they need more time to make a decision than there is between the new date and the conference itself. Would it be possible to obtain a
17:49:51
beach
phoe: The other thing you can do is to start the process, saying you won't know until March 6 whether your paper is accepted or not.
17:59:49
fiddlerwoaroof_
I see two or three on quicklisp, but I'm wondering if anyone has nay experience with one of them
18:17:41
pebblexe
hey, how do you convert a string to a simple-string? cl-ppcre says my string fell through, it wanted a simple-string
18:23:11
didi
I want to print a long list prefixed by #\; in every newline. How do I do it? When I eval (format t "~&; ~A" (make-list 42)), format breaks the list with newlines but each newline is not prefixed by a #\;.
18:28:00
sirkmatija_
reposting question, since my internet connection died after posting it last time :/
18:28:14
sirkmatija_
Cl-charm get-char function doesn't detect keypresses, unless I hold the key down for longer time or press it multiple times. I have evaled both enable-raw-input and enable-non-blocking-mode. Anybody knows where is the problem? Thanks in advance
18:28:16
sirkmatija_
Oh, I also use get-char with :ignore-error t and loop it until it returns non-nil output (key char)
18:30:06
pebblexe
actually, it turns out I was handing ("test" "test") to concatenate instead of "test" "test"
18:32:01
pjb
didi: or you may want to use pretty printing and ~<, but I don't know that well enough to tell you exactly how.
18:43:06
pjb
Petit_Dejeuner: #'foo and (function foo) read as equal forms, if you have the standard #' reader macro bound in the *readtable*, and *package* is bound to a package where cl:function is accessible.
18:46:03
jurov
i came up with `',variable to obtain quoted symbol from variable and it works, but it's such a wart, is there more beautiful way to do it?
18:48:48
mhd
where does emacs decide to have l, lsp, and lisp as its known Lisp extension for automatically putting buffer into lisp mode?
18:53:10
didi
I used to think `let' was ripe for parallel computation, until I wrote an evil `let' that would totally break if the value forms are not computed in sequence. Does `let' guarantee the value forms are computed in sequence--I'm not talking about binding the variables--or I just got lucky?
18:54:00
mhd
Handy log: (loop for (x . y) in auto-mode-alist when (eq y 'lisp-mode) collect `(,x . ,y))
18:54:00
mhd
(("\\.l\\'" . lisp-mode) ("\\.li?sp\\'" . lisp-mode) ("\\.ml\\'" . lisp-mode) ("\\.asd\\'" . lisp-mode))
18:54:38
Xach
varjag: ah yes, it's color-table-entry, which returns a value that can be broken out with color-rgb
18:56:27
pjb
mhd: or you can put everything in a single regexp (concat "\\." (regexp-opt '("l" "lsp" "lisp" "asd" "ml")) "$") --> "\\.\\(?:asd\\|l\\(?:i?sp\\)?\\|ml\\)$"
19:05:05
mhd
@pjb, I guess ACL users emacs init files all contain (add-to-list 'auto-mode-alist '("\\.cl\\'" . lisp-mode))
19:06:10
pjb
mhd: <- irc; Everybody, since there are free software acl libraries. You may find some in quicklisp.
19:28:31
drmeister
I've spent a couple of hours trying to track down what this means and what might be wrong: #<TEST-RESULT GRAY::STREAM-READ-LINE :FAIL Unexpected end of file on #<TRIVIAL-GRAY-STREAMS-TEST::TEST-STREAM>.>
19:33:35
drmeister
But the ECL source code doesn't bind the READ-LINE symbol to anything - I don't see how it could invoke GRAY::STREAM-READ-LINE
20:16:15
Xach
varjag: it really only needs to fetch the small metadata files. everything after that will be fetched on demand.
20:16:32
Xach
varjag: as a courtesy, it updates whatever you had installed already, but if it stops partway, it will just get them when needed.
20:35:06
fiddlerwoaroof_
Xach: is there a way to get a quicklisp system's source repository without quickloading it?
20:37:09
fiddlerwoaroof_
ym: that error messages sounds a bit like openbsd is treating your executable like a shell script
21:01:31
Xach
the client will likely never validate ssl certificates. it *will* verify cryptographic signatures very soon.
21:02:27
Xach
i have all the code to do the work done, but it must be integrated with the current dist setup.
21:26:05
Xach
it will be an openpgp public key. it will be embedded in the bootstrap file, which is available via https.
21:26:26
Xach
the goal is that everything the bootstrap and client does can be verified by other tools.
21:30:12
jackdaniel
drmeister: it is used only for thread safety (concurrent access), see src/c/gfun.d (esp comments)
21:40:37
TruePika
It should be possible to detect their disappearance, but my plans involved part of the bot having a helper around each PLANE, to simulate TCAS
21:42:53
TruePika
I'd probably do something akin to a reinitialization to a dummy class which issues a condition
21:57:47
TruePika
hash test would need to be equalp, but there will only ever be up to 26 planes in the world
21:58:43
TruePika
though, I'd theoretically not need to use an alist (just a regular list), on second thought
22:04:29
X-Scale
http://www.cl-http.org:8002/lispdocuments/draft-proposed-ANSI-Common-Lisp.pdf seems to be gone. Any idea where one can find it now ?
22:10:10
X-Scale
I guess I've found a similar document -> ftp://linux4u.jinr.ru/pub/misc/symbolic/lisp/dpANS3/book.pdf
22:10:14
Xach
X-Scale: ok. the source material from which it is derived is freely available and widely copied.
22:11:56
dim
in the case of pgloader the perf impact is quite small, when measurable, because it's all IO bounded (network and disks) anyway
22:39:55
fiddlerwoaroof_
phoe: yeah, lisp is surprisingly versatile for web programming, despite the somewhat low-level apis it provides for string manipulation
22:52:10
fiddlerwoaroof_
That's true, I generally find that the various string manipulation libraries don't fit together very well and don't have some options that are useful. I.e. split-sequence takes a count argument, but that determines the number of split fields returned, not the number of splits made.
22:52:41
fiddlerwoaroof_
In Python, it's really useful that 'a b c d'.split(' ',3) returns ['a','b','c d']
22:53:13
fiddlerwoaroof_
On the other hand, lisp makes it really easy to fill in the gaps, in a way that few other languages do
23:01:42
TruePika
unable to optimize because: could not optimize away %SAP-ALIEN: forced to do runtime allocation of alien-value structure
23:23:44
pjb
fiddlerwoaroof: (split-sequence:split-sequence #\space "a b c d" :count 2) #| --> ("a" "b") ; 4 |#
23:25:06
pjb
fiddlerwoaroof: (let ((string "a b c d") (n 3)) (multiple-value-bind (splitted next) (split-sequence:split-sequence #\space string :count (- n 1)) (append splitted (list (subseq string next))))) #| --> ("a" "b" "c d") |#
23:26:32
fiddlerwoaroof
pjb: I'm not denying that it's easy to write your own, I'm just saying that all these little things introduce friction into the sorts of string processing I frequently do
23:27:12
pjb
fiddlerwoaroof: I agree. That's why you build up your own library. Or you may design a consistent string processing library.
23:27:15
fiddlerwoaroof
Lately, I've found a #{} reader macro that mirrors Clojure's #() macro for defining functions is useful
23:28:14
TruePika
but I'm most worried because the form being complained about is almost part of SBCL
23:29:03
fiddlerwoaroof
Tcl would make sense because macros there are essentially string transformations
1:58:45
pillton
You'll probably have to be more specific. I've solved graph problems by solving a linear system.
2:00:55
didi
pillton: I currently implement graphs using adjacency-lists and a hashtable. The hashtable hold the vertices as keys and the values are the adjacency-lists. I am wondering about different implementations and if anyone investigated this problem space in lisp before.
2:12:53
didi
I incidentally came across the passage that I mentioned: https://jaxenter.com/disadvantages-of-purely-functional-programming-126776.html "With a garbage collected imperative language, the relationships between the vertices and edges of a graph can be expressed using weak hash tables. The garbage collector will then collect unreachable subgraphs for you."
2:23:04
Petit_Dejeuner
didi: "It took 50 years for normal people to dilute the smug weenies to the point where you can get a useful answer about functional programming on social media."
2:31:07
fiddlerwoaroof
I like it how he tells us that the Lisp community had wrong arguments about Lisp's goodness, but doesn't bother to share them
3:02:04
didi
pillton: I am thinking about it. I want to retain the original vertices names, so I am thinking of using a map--possibly a hashtable--for names and indices and an array for the graph proper.
4:53:58
jack_rabbit
Is there a way to inspect the SBCL "heap"? I'm curious about the kinds of objects that are alive.