| |
| |
| |
|
Page: 1 2 3 4 5 6 7
Comments:
<0> aja: ok, you get copyright
<1> I'd look it up on m-w, but it's not responding.
<2> sidewinder: I also invented "Viagra is a great recursive acronym".
<1> And my paper dictionaries are at home.
<3> aja: oh darn, I should have found a way to use that to poke fun at pg instead of "Span".
<2> cliini: I think the determination that Jingoism is extreme nationalism, possibly active and directed outward, is fairly good.
<4> My electronic dictionary says "extreme patriotism, esp. in the form of aggressive or warlike policy"
<2> Jabberwockey: And apoligies for mocking Germans - 50+ years on, it's probably turning into a stale joke.
<4> But it doesn't have any etymology which is why I checked Wikipedia
<4> Also I can cut'n'paste from WP, but not from the dictionary widget
<5> aja: No worries, I enjoy making fun of us as well, as long as the jokes are intelligent and not the usual stereotypical, lame-arsed attempts.
<1> The correct definition of the technical term jingoism is the full text of Jingo, by Terry Pratchett.
<1> But it's too long to paste on this channel.
<6> cliini: That's why we have lisppaste ;)
<1> I think lisppaste has some sort of limit, less than needed in this case. :-)
<5> aja: Btw: Here's one for you: Georg Friedrich Hndel was both British and German. What does that mean?
<1> That nationalism is silly?
<7> cliini: but then, you have to calculate a standard error on the 5-dictionary consensus (;
<1> How many nationalities did Einstein manage to gather? Four*
<2> Jabberwockey: He couldn't cook and he made very long operas?
<7> Jabberwockey: that he was the german hipster prototype (:
<5> No. That means that he had this overwhelming urge to invade Poland and conquer the world, but he was too polite for doing it.
<8> I'd say that 'Jingoism' means 'whatever those scary people are doing that is leading us to war'.
<5> antifuchs: Hi, you :)
<7> hi jabberwockey!
<2> Jabberwockey: Full disclosure: I'm Canadian, and EVERYONE makes fun of us. Even us.
<5> aja: I like Canada.
<2> Jabberwockey: :-)
<7> no wonder you do. it's so far away... (;
<1> I almost claimed I've never been to Canada, but that's not strictly true.
<5> aja: Never been there. But of what I heard, it's an amazing place. Amazingly cold and lonely at some places, a little weird with its version of French, but nice.
<1> I've seen the Niagara Falls from the Canadian side, from a boat. .)
<2> Jabberwockey: Well, its hard for me to be objective. It is, however, huge and very sparsely populated for the most part, so "lonely" is accurate, as is "cold" - but global warming is working out OK for us.
<1> There are bits of Canada which are hardly cold.
<6> Like Vancouver.
<1> And surprisingly, most Canadians live in those bits.
<6> People also live in Winnipeg, but they're just nutty.
<5> I liked the part from "Bowling for Columbine" where Michael Moore walked into houses in Canada to find that none was locked :)
<1> The bits which are truly cold have so low population counts that it is almost silly.
<2> Jabberwockey: And I'll give you the "weird French" thing. Rural Quebecois have difficulty making themselves understood to European French speakers.
<6> Jabberwockey: Yeah, but the people were _home_. I grew up in a bad neighborhood in Philly, and we didn't lock our doors when we were _there_.
<2> Jabberwockey: In the larger cities, they're locked. Vancouver and Toronto aren't all that different from major cities in the US.
<5> sellout: Sorry, but just as you mentioned Philly there started this playback in the back of my head. Please wait while I try to get Bruce Springsteen out of my ears, ok
<5> ?
<1> sellout, that's city for you. You know you are in a rural environment when there is a broom keeping the door shut when people are not home. .)
<6> Switch to Will Smith.
<5> sellout: Ew.
<6> The Roots?
<6> G. Love?
<5> sellout: Never heard of them.
<5> Sorry, doesn't ring a bell.
<6> Man ...
<9> Sorry, I couldn't find anything for ....
<6> Dieselboy?
<5> Heard the name, never heard the music.
<10> moin
<5> Moin moin!
<10> can anyone of you point me to a site that tells me how the s expression string has to be parsed?
<6> Anyway, that's why I grabbed the Fresh Prince. At least you know who it is.
<2> bourbaki: What do you mean by "The s expression string"
<5> sellout: Well, I have to admit, some of his stuff isn't too bad.
<11> bourbaki: maybe http://www.lispworks.com/documentation/HyperSpec/Body/23_.htm - but that's a bit sparse
<5> (read-from-string "...")
<6> I'm a fan ... it's silly and poppy, but that's better than acting like some kind of tough.
<5> sellout: THAT's true!
<10> aja i want to write a small program that can parse expressions like the ones in lisp but i dont have a good idea how to parse it when there are nested expressions like (foo (bar ..))
<6> Anyway, this started out offtopic, and I'm wandering.
<6> It's time for sleepl.
<6> also, sleep.
<5> sellout: And I needed weeks to get this "Welcome to Miami" out of my head.
<5> sellout: Sleep tight.
<8> bourbaki - a small program in what language?
<2> bourbaki: Hrmm. That's not all that hard, given a fairly limited available syntax (Norvig goes over it in PAIP). If you want a general solution, you're sort of getting into writing a reader.
<10> ayrnieu c#
<12> bourbaki ... nickieben
<13> (newbie question) is there a way to get an object's type in CL? (i can use typep to only do type comparisons)
<1> type-of
<2> sreeram: type-of
<13> thanks
<14> sreeram: what do you want to use the type for?
<1> But of course an object can have multiple types that satisfy typep, but only one of them will be returned by type-of.
<2> sreeram: describe is also useful for clos objects
<13> NIL satisfies (typep nil 'symbol) (typep nil 'list).. and i was just trying to find out what its type really is...
<1> It's perhaps not the way of thought that is most useful with types.
<1> Of course, in this particular case there is an answer: the type is NULL.
<1> But for lots of built-in types the spec doesn't specify exactly what type-of will return.
<15> The CMUCL/SBCL way for integers is sort of amusing, if nothing else.
<13> cliini, i was checking if sbcl implements NIL as a null pointer, or as just another constant symbol, or as a special object...
<1> sreeram, you can't know how it is implemented from what type-of returns.
<15> Oh, SBCL is a bit less amusing, I notice, now.
<13> cliini, thats true...
<12> i think an object has a [virtually] infinite number of types
<1> NIL can be implemented in lots of ways. In SBCL, it has both a valid symbol layout and a valid cons layout, I think.
<13> but i can find out if nil has a its own special type or not...
<1> Well, it does, in CL, by definition. That special type is NULL.
<12> ***uming that type=set and 'is of type'='belongs to set'
<13> (typep nil 'cons) -> nil
<1> It's not a cons, but it's a list.
<13> i'm building a toy lisp interpreter in python.. and am currently using None for nil.. and i realized that causes too many special cases...
<1> Nothing that can't be solved by another layer of indirection.
<13> right! :)
<15> sreeram: A cl***ical way is #1=(#1# . #1#), and a special-case in cons, I believe.
<16> there's an explanation on how nil is implemented in sbcl is src/compiler/generic/objdef.lisp
<15> Ignore the "special-case" part. I don't know what I meant.
<2> sreeram: try (eq nil '()) for funzies. Nil is a list - which means that the way it is implemented has some vagueness.
<14> sreeram: the cl*** of an object is what determines its implementation, not the type
<14> sreeram: perhaps cl***-of will give you what you want.
<8> (cl***-of '()) #<The BUILT-IN-CL*** NULL>
<13> hmm.. (cl***-of t) -> symbol... (type-of t) -> boolean
<8> sure, t is much simpler.
<2> ayrnieu: How does that work with the old "all cl***es are descended from T" thing?
<17> the cl*** T has nothing to do with the object T
<18> sreeram: there's a language to describe types as arbitrarily detailed sets, while cl***es are more like types in C etc.
<18> there's e.g. the type "(or (integer 4 42) symbol)", but no cl*** that is similar at all.
<2> frodef: sreeram: I was going to express it as "type has more to do with how something is represented in memory, and cl*** has more to do with the chain of inheritance." Not sure of the accuracy of that, though.
<18> how about: cl***es are the terminal nodes in type expressions?
<18> (is that even true?)
<14> no
<14> cl***es and types are completely separate
<14> but a cl*** has a corresponding type
<15> aja: That's probably not very accurate. (deftype prime () '(satisfies primep)) doesn't say much about how it's represented in memory, for instance.
<14> representation and implementation are a cl***-issue, not a type-issue
<14> types are about sets of objects
<16> Xophe: I did some benchmarking for the clos type-checking stuff. it slows down most kinds of slot writes by an order of magnitude, so it's probably best done only for (safety 3)
<18> zhivago: right, but substitute "cl***es" with "the types corresponding to cl***es"..?
<17> jsnell: really? I thought I'd arranged it so that if there was no :type declaration there was no speed penalty
<17> darn
<2> rydis: That's a good point. I suppose I went with the thought that you often select type on how it is implemented, but that might be incidental to the problem.
<2> s/implemented/represented in memory
<16> Xophe: sorry, I meant for slots with a :type
<17> ah. I can take another look at it all at some point -- I understand more about clos now than I did then -- but not for the next 10 days or so
<14> frodef: well, I'm not sure what terminal means here
<14> frodef: you can talk about types which are subsets of the set of instances of a cl***
<16> Xophe: I'm doing some work on it right now, so no hurry
<16> though I wouldn't mind your opinion on a couple of corner cases
<18> zhivago: terminals of the type expressions such as "(or integer symbol)" .. except things like satisfies.
<14> frodef: I guess so. :)
<19> jsnell pasted "the cases" at http://paste.lisp.org/display/25320
<20> good morning
<21> good morning
<22> jsnell: they're interesting cases, particularly the second one
<22> there's some low-hanging fruit in the permutation-vector optimization stuff that would optimize inheritance-test2 to look exactly like inheritance-test
<7> argh
<7> usocket has no support for :external-format and uses :element-type 'character on socket streams?
<7> wtf!
<22> scratch that recommendation, then
<7> indeed /-:
<23> antifuchs: have you been the one recommending usocket a few days ago...? :)
<23> antifuchs: it just needs to propagate the external-format param. but there's not even sign of trying to do that... socket-connect doesn't even get the params
<7> attila_lendvai: yeah, I just discovered that it doesn't do that
Return to
#lisp
or
Go to some related
logs:
#bind
does shockwave work on ubuntu
hooray beer wav
#css
'concat' is not a recognized function name.
#python
logitech mx510 mouse xorg.conf
gentoo ntp no reply
force clvm removal ubuntu
deadroot + iis
|
|