| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10
Comments:
<0> wtf!? http://www.biblio.com/books/44457124.html I know we've talked about expensive books here before, but this takes the cake
<1> I suppose suspend debugging would inevitably lead me to reading mjg59 sooner or later
<1> http://lkml.org/lkml/2005/6/6/85 "The behaviour is very non-deterministic"
<0> actually, this one is more expensive: http://www.ilabdatabase.com/php/highlights/detail.php3?lang=&book=3 but that seems a bit more understandable
<2> we need temporary brain fusion as a more direct way of communication. This email business is rather tedious.
<3> ick, I wouldn't want to fuse with somebody else's brain when I don't know where it's been...
<4> lichtblau: around?
<5> When I try to install asdf I get package "SB-EXECUTABLE" not found. What do I do?
<6> Mitja: what implementation are you using?
<5> SBCL
<6> asdf comes with sbcl
<6> to load it into sbcl, type (require 'asdf) or similar
<5> Should (require '<package>) normally return NIL?
<7> yeah
<5> Xach, thanks
<5> and hnaz_ :)
<8> does anyone else feel a little weird about the most recent franz news post on planet lisp?
<9> hehe
<10> ...indeed.
<10> The first post yesterday?
<0> lemonodor: we should offer cheap licenses for SBCL to LW and ACL users
<3> yes, seems to me that to be in bad taste
<7> lemonodor: the 'jump to 8.0'?
<3> heh, 8.0 must be much better than 0.9.15
<3> almost NINE TIMES BETTER!
<1> "For ACPI to work correctly, all the parts have to work correctly." # my only observation is that Mr Holmes seems to be constipated again
<1> we could do an emacs and drop the leading major version
<8> i don't think there's anything wrong with the announcement as such. it's fine, it's just that it seems to have crossed the line into advertising. and maybe (free) corporate advertising isn't appropriate on planet lisp
<1> Franz News has always been really rather close to advertsiing, in my view. for what little that' sworth
<8> franz may feel it's appropriate in the context of their feed.
<8> well, typically that feed seems to be about new releases and new features. sometimes some promotional material.
<8> none of that has made me feel weird. maybe because this one specifically targets a competitor.
<8> it just feels more like something you'd see come up as a google ad more than the others have.
<1> yaeh, it was an odd one to see, I'll give you that
<1> but Franz has always been the "odd one out" in planet lisp terms anyway
<1> at least the others all seem to have been written by identifiable individuals instead of marketing departments
<8> i actually like having franz news show up. i kind of hoped that it would motivate some of the other vendors to publish tasteful, appropriate feeds hoping to get on planet lisp as well.
<11> For this kind of marketing, I think LW ought to be given equal time (***uming they would use a feed for that sort of thing) or else neither should. But I realize I don't get to vote on planet lisp policy.
<10> Wouldn't it be easist and most politic to send a polite message to Franz about this?
<11> My guess is that they would respond with "we have no control over who subscribes to our feed."
<12> Why not slip a quiet word to the LW marketing people?
<8> that's what i was thinking. i was just curious if i was the only one that felt it might not have been appropriate.
<8> why, because we want a lispworks ad up next?
<13> should probably run that by Xach first
<11> lemonodor: no, I certainly don't, but just looking at it from the standpoint of fairness.
<1> Riastradh gets the award for most sensible and least viscerally satisfying course of action
<1> can't we just go and start a fight somewhere?
<11> heh
<10> junrue67, sure, but they might also be willing to separate the content of their feed. It's still worth a shot.
<11> Riastradh: ok sure
<14> I don't even mind seeing "we're running a special on Allegro; see this link for pricing details"
<15> pricing details from Franz? hah
<14> It has been known to happen occasionally. They recently did publish a price for Allegro 8.0 on OS X.
<5> What is the equivalent to *standard-output* function in sbcl?
<14> Your question doesn't make much sense. *standard-output* is a special variable, not a function.
<15> actually the sleazy part of the ad (IMO) was implying that you'd really be getting something for paying the same price as lispworks costs
<15> but it seems that you still need to pay the runtime licensing fees on top of that
<5> chandler, CL-WHO uses a function named *standard-output*.
<5> An example using CL-WHO, rather.
<15> it very probably doesn't
<0> where's that commercial solicitation from sb-studio?
<14> While it is entirely possible to lexically bind *standard-output* as a function, it would be incorrect for any implementation to supply a function definition.
<15> any more than (with-open-file (*standard-output* ...) ...) is using *s-o* as an function
<14> It's more likely to me that a package import was missed, a macro was treated as an ordinary function call, and special syntax supplied by the macro was instead interpreted as a function call.
<0> why do I still get an odd number of keyword args when attempting to call (defun foo (&rest rest &key moose &allow-other-keys) like (foo :lame :moose :bar)
<16> Becaus eyou provided only two keywords: :lame and :bar and no value for :ba.
<16> Try (foo :lame 'lame-value :moose 'moose-value :bar 'bar-value)
<0> pjb: but that's what I want, in this case. I was hoping that rest would get all the args and moose would be nil
<0> hence the &allow-other-keys (see the note in CLHS 3.5.1.6)
<16> slyrus: &key overrides it: it imposes the odd number of rest arguments.
<16> s/odd/even/
<17> "If &key is specified, there must remain an even number of arguments" 3.4.1.4
<0> 3.5.1.6 led me to believe otherwise, but ok.
<17> 3.5.1.6 starts out by saying "An odd number of arguments must not be supplied for the keyword parameters."
<0> yes, but the next paragraph, which describes the consequences of breaking this rule, would seem to indicate that if &allow-other-keys is specified, no error need be signalled.
<17> it then goes on to discuss when an error _must_ be signalled, but it never says it's ok to have an odd number of keyword arguments
<0> ok
<17> I admit the wording is a bit confusing
<14> slyrus: I find that when I start playing games with &rest &key &allow-other-keys it's usually best just to go back to a &rest argument and pick values out of it as necessary
<18> hello everbody
<12> "thinking pants"?
<0> thanks chandelr. now I have a nice remove-keywordish-args
<0> chandler`, even
<12> ... GRR.
<12> STUPID @%$%$ INTERNAL-ERROR mechansim. :-/
<19> Be calm. It'll be ok :)
<12> What a pain in the ***, though. This should be a -trivial- proof-of-concept, and there's no bloody hook to hang it on.
<12> (Or, possibly slightly more accurate, no -obvious- hook.)
<20> what is it?
<12> ... I take that back. The signal-handling mechanism is the hook I need.
<12> spiaggia: I'm trying to find a scenario in stock SBCL wherein I can convince it to restore an existing signal context.
<12> And, of course, the obvious case is SIGINT.
<12> Now, what was the magic I had come up with...?
<12> Hrm...
<12> spiaggia: Want a proof-of-concept instruction single-stepper in SBCL on x86/Linux?
<12> (Oh, and while I think of it, some of the research I've done since the original SBCL/Win32 work indicate that I didn't have to be anywhere near as paranoid with the context information for exception handlers.)
<20> nyef: very interesting. I'll be off for a few days, but I'd like to hear about your progress when I come back.
<12> What? Aarrgghh!!
<21> nyef: I think I can do a lisp-level stepper :)
<12> pkhuong--: Oh?
<21> yeah, with my continuations code
<21> just have to change the bind macro
<12> Heh. Ah, right.
<21> prompt for a value/continue, etc.
<12> Honestly, I don't think I'm going to develop this much beyond a proof-of-concept.
<12> At least, not now.
<21> it'd get pretty tedious, though, since every subcomputation would stepped through.
<21> +be
<10> pkhuong--, on dynamic binding and continuations: While the semantics of the interaction between non-local control transfer and dynamic binding is specified in Scheme, there are no partial continuations, so it gets tricky.
<12> ... You know you have to worry when your signal context tells you that your interrupt occurred at #xffffe402 on a 32-bit system.
<22> Riastradh: I think I'll provide a windup-protect (? any standard name for protecting forms w/ a prelude) in addition to unwind-protect. Build dynamic binding on top of that with side effects and if something must be single shot, etc., that can be specified on top of these two primitives.
<10> It's called DYNAMIC-WIND in Scheme.
<22> k.
<12> SAPs are immutable, aren't they?
<23> hmm I'm sick and tired of debugging using (format). I'm writing my own loggig library
<23> somewhat modeled aftel log4cplus, basically log4j architecture but without fancy stuff, ie no category, instead logger names are strings as in package.whatever.outer.inner format
<23> plus I won't have (defstruct loggingevent) to avoid consing, will just p*** strings around
<23> well anyday
<11> maxm: perhaps log4cl would be a good starting point. http://common-lisp.net/project/log4cl/
<23> junrue67: see my comments at the end of log4cl on cliki
<11> ok, I will :)
<23> basically its a nice name, but unfinished and ****s :-)
<11> ah, well then never mind :-)
<24> there you go then, an ideal starting point on which to build on
<25> anyone using s-http-server with sbcl
<12> xristos: I'm not, and wasn't planning on it, but, now that you've indicated the existance of such a thing, I'll have to look into it.
<12> Meanwhile, I need to get some sleep.
<25> one thing that needs to be done is to cure us from such ailments as sleep
<25> half of our lives wasted
<26> yeah I agree
<27> dulouz pasted "I don't Understand the :|| Z" at http://paste.lisp.org/display/23858
<28> hi, can anyone enlighten me about that?
<26> lol, I have no clue
<28> that's unfortunate!
<26> I mean I have no clue what :|| is.
<29> doesn't look right -- :initarg :,s
<29> specifically the :,s
<26> yeah
<28> well i'm expecting it to end up as
<28> (Z :INITARG :Z :INITFORM 1 :ACCESSOR Z)
<28> just messing around, learning, messing up mostly :)
<28> do i need to escape the ":" character or something?
<26> I always thought that you didn't
<26> not with macros
<26> but maybe I'm mixing something up
<26> **** it, guinness and lisp don't work together
<26> gnight y'all
<28> hah
<30> just to take part, I find that two pints actually improves my code
Return to
#lisp
or
Go to some related
logs:
#physics
+copy files +ubuntu +home +sudo
ubuntu netserver lh3
remove warning Parsing of undecoded UTF-8
ricbonco
Unknown symbol pci_pretty_name
#suse
#ubuntu
multimedia codecs opensuse
#freedesktop
|
|