Are we close to a Lisp boom ?
in [Lisp]
|
Prev: ANN: SymbolicWeb v0.1 (quite alpha) w/ source code this time
Next: Greenspunning ML (revisited)
From: Robert Uhl on 6 May 2008 11:39 Ken Tilton <kennytilton(a)optonline.net> writes: > >> IMO, not until CL goes through another standardization process, not >> for the language this time, but for a few libraries: comm, stream, >> unicode, thread. > > Yeah, the damn thing is unusable as it is. It's not that it's unusable (that's clearly false, as every project written in CL clearly demonstrates); it's that it's less usable than would otherwise be the case. It's not an issue of black-and-white: it's an issue of lighter vs. darker grey. There was a time when CLOS was not standardised; as The Art of the Metaobject Protocol demonstrates, it could always be rolled by hand. Surely you agree that a single standardised CLOS is better than a dozen similar-but-incompatible OOP libraries? In the same way, standardising libraries would be useful. The existing pretty-much-similar sockets libraries should be standardised. Gray Streams should be standard. Unicode should be the new standard, with clearly-defined migration paths for legacy encodings (I'm not up on my Unicode specs--perhaps there are some suggestions already there). Threading and multi-processing should be standardised. Something interesting would be standard message-passing based on the Erlang model. I don't know if there's any agreement that the Erlisp model is the right way to do this though. It would be nice for CL to be as far ahead of the curve again as it once was. Garbage collection and closures are pretty common now; optional and keyword arguments are not unknown (c.f. Python); macros are at a tipping point (people realise they need them, and are trying to figure out how to get them in irregular languages); CLOS-style generic functions are in a similar position; conditions still haven't caught on; I think that integrating some of Erlang's features might be a way to do for multiprocessing what CLOS did for OOP. >> Too bad there isn't a benevolent angel that could fund such >> an expenditure <hint>PG</hint>. > > Nah, he went broke trying to do a start-up with CL, a Web store I think. PG's too busy reinventing the wheel with Arc to do anything for CL. -- Robert Uhl <http://public.xdi.org/=ruhl> People who do technical support for a living are bitter, twisted and uncharitable. Eight hours a day of telling people what's already in the manual [...] results in a steady and inexorable progression towards a state of depressive sociopathy. --dansdata.com
From: Robert Uhl on 6 May 2008 11:43 Edi Weitz <spamtrap(a)agharta.de> writes: > >> I'm no longer writing CL, I'm writing a dialect of CL that is >> dependent on the success or otherwise of my vender. > > How is that different from C/C++? With standard C plus standard POSIX, you're pretty much certain that your app will run anywhere important. You're not certain that it'll run fast or particularly well, and of course there are those edge cases that need to be taken care of--but my perception is that C+POSIX is much more reliable a platform than Common Lisp. It's also much more low-level and much more prone to segfaults and security holes. At the moment I'd rather program in non-portable CL than in portable C, but that's me. -- Robert Uhl <http://public.xdi.org/=ruhl> Remember, democracy never lasts long. It soon wastes, exhausts, and murders itself. There never was a democracy yet that did not commit suicide. --John Adams, 1814
From: Duane Rettig on 6 May 2008 11:43 pjb(a)informatimago.com (Pascal J. Bourguignon) writes: > Duane Rettig <duane(a)franz.com> writes: > >> EL <eckhardnospam(a)gmx.de> writes: >> >>> Spiros Bousbouras schrieb: >>>> If it's true that as we progress in time, successive >>>> fashionable languages resemble Lisp more and >>>> more then Lisp's turn should come at some point. >>>> Do you agree with this argument ? If yes, would >>>> you say we're close to a Lisp boom ? >>> >>> At least you guys made it on rank 16 in this <ironic>very >>> meaningful</ironic> index here: >>> http://www.tiobe.com/index.php/content/paperinfo/tpci/index.html >>> >>> And in the "cleaned up" list here on rank 11: >>> http://cdsmith.wordpress.com/2007/06/18/is-tiobe-fatally-flawed/ >>> >>> Not bad, eh ;-)? >> >> Notice he had to re-split the Lisp and Scheme categories, in order to >> make his list believable - would anyone have accepted a list like that >> where Lisp/Scheme ranked in positions close to Fortran and C? >> >> :-) > > And I notice a CL in 27th position too. Ah, you're talking about the Tiobe site. I was talking about the "cleaned up" list, which currently places Scheme at #12 and Lisp at #13. I was the one that originally suggested to the tiobe people that they combine Lisp and Scheme. > If we added Lisp+Scheme+CL, > the sky's the limit! Actually the first position, but good enough :-) As far as I know, CL is actually included in that mix for the Tiobe data. It's hard, though, to infer contextually that a particular instance of "CL" stands for Common Lisp. I read at one point how they get their data, and that's what led me to make the suggestion to combine. I personally place very little faith in numeric measurements like tiobe's list and LOC measurements, mostly because they can be abused by both sides of an argument. But for what they're worth (i.e. an interesting number) they are numbers, and they make possible a narrow comparison that wouldn't be otherwise available. But take or leave the list; the guy who wrote the "cleaned up" list (which he apparently now retracts becase he's seen the light about taking any list too seriously, including his) was obviously incensed that his own pet language, Haskell, was rated less popular than APL (which nobody has a keyboard for anyway :-). Everyone is going to find reasons to hate the list, and it should be a lesson to those who ask for "proof" that Lisp is better than something else; it's not going to happen, because in deconstructing the meaning of the word proof one finds that there needs to be an observer to the proof/test, and such observation is always going to be subjective. -- Duane Rettig duane(a)franz.com Franz Inc. http://www.franz.com/ 555 12th St., Suite 1450 http://www.555citycenter.com/ Oakland, Ca. 94607 Phone: (510) 452-2000; Fax: (510) 452-0182
From: Robert Uhl on 6 May 2008 11:47 Pertti Kellomäki <pertti.kellomaki(a)tut.fi> writes: > > There is lots of stuff out there for which it does not really matter > whether one is using CL, Python, Ruby, or any other half-sane > language. In that situation, the language where you can just say > "import X" for stuff like regexps, HTTP, HTML etc. is going to > win. And for good reason, I might add. Yes. Fortunately, for the important Lisps one just goes to weitz.de and grabs CL-PPCRE (faster Perl-compatible regexps than Perl!) and Hunchentoot. HTML generation is another issue, although I'm partial to CL-WHO (also from Dr. Weitz). For a good portion of the '&c.' weitz.de is your one-stop shop... -- Robert Uhl <http://public.xdi.org/=ruhl> Customs officers enter into a Faustian bargain whereby they are given absolute power in exchange for their sense of humour. Hitler's dad, you will remember noddingly, was a Customs officer. And _Hitler_ thought he was a nasty piece of work. --Mil Millington
From: Edi Weitz on 6 May 2008 11:51
On Tue, 06 May 2008 09:43:16 -0600, Robert Uhl <eadmund42(a)NOSPAMgmail.com> wrote: > With standard C plus standard POSIX, you're pretty much certain that > your app will run anywhere important. How many useful and/or successful Windows apps have been written in pure standard C plus standard POSIX? -- Lisp is not dead, it just smells funny. Real email: (replace (subseq "spamtrap(a)agharta.de" 5) "edi") |