some stuff about the 2002 International Lisp Conference in SF

243 views
Skip to first unread message

Dave Bakhash

unread,
Oct 31, 2002, 12:05:46 PM10/31/02
to
Hi,

I just got back home, missing the last day of this year's International
Lisp Conference. But for the people who were unable to make it there
this year, I'd like to say a few things about it, in the hopes that
people will find one another, and good things will result. I don't
intend for people to follow up, or for most people to care to read what
I have to say...but figured that I'd have wanted to get an idea of what
someone else had come back from the conference with, had I not been able
to make it there. This was my first such conference in CL.

There's a lot to say, so first off, I'll start by saying that it was
very well organized. The organizers did an incredible job, keeping it
timely and moving, yet relaxed and comfortable. Franz had several
people working around-the-clock keeping everything running smoothly, as
well as Raymond and the other folks who provided sponsorship and
support.

For me, it was amazing to finally meet people I've been talking to
directly or indirectly for years. Brigitte Bovy of Xanalys was
especially nice to meet in person, after years of knowing each other
over the phone.

There were some incredible speakers. I particularly enjoyed hearing
what RMS had to say, and despite not preparing anything for the talk, he
told an amazing story, shared a lot of insight, and did so with a warm
sincerity.

I spoke to RMS briefly about using CL instead of Guile (Scheme) as the
next extension language for GNU Emacs, and all he really had to say
about it was that it was "too big". On that note, I'll mention ECL...

ECL was presented by a really cool Spanish guy, a physicist now living
in Germany named Juan Jose Garcia-Ripoll (excuse me if I got the
spelling wrong). He presented ECL, a continuation of KCL, which is an
incredible project which has nearly a full ANSI CL working as a library
(about a meg), that you can call from C, and which compiles CL->C, and
readable C at that. It also provides a nice C calling interface to CL
functions, and gives you a fully-functional prompt with a REPL. This
work has already been ported across many Unixes (including Mac OS X) and
Win32 via Cygwin. I find this to be an excellent direction for a Lisp
implementation, especially as most of CL can be done with portable ANSI
C, and then leverage modern C compilers. Benchmarks supposedly put
compiled ECL code ahead of CLISP, making it a better bet. For those
interested, it's LGPL'd, and you should check it out. I'm looking into
putting some money together to fund a effort for ECL in GNU Emacs in
place of Guile as a better (and much more efficient) alternative.

Anther big speaker (for me) was Peter Norvig. I've read most of his
books, like many people here, and have a deep respect for him. He's at
google now, and discussed some of the real-world problems that Google
faces in their line of work. I shared his feelings that Common Lisp,
though having many advantages over other programming languages, was no
longer alone, and that many of the key features that CL had that were
not found in other languages have been adopted, gradually, making other
languages good (and often better) alternatives for certain applications.

I also got to talk to the guy who mostly wrote and maintains Open MCL,
Gary Byers, who was one of the nicest, most patient, and knowledgeable
Lisp guys I've ever met. Open MCL is another great Lisp, the engine
behind Digitool's MCL. He was extremely low-key, despite the swarm of
people with Macs that use and rave about MCL. I have never seen so many
Macs before (except product placement of Macs on TV and in movies). I
felt like one of a minority of people with a non-Mac laptop. In any
case, Open MCL has been ported to Linux (PPC), and is free, while
Digitool's MCL is coming soon to OS X, and if you buy it now, you get
the beta now for about $500 and then the stable addition when it comes
out (saving about $200 from buying it when it comes out, which IIRC will
happen around January...correct me if I'm wrong). Franz also passed out
their ACL 6.2 trial edition, which includes ACL for OS X. So now Mac OS
X is in the race with CL, with ECL, (Open) MCL, ACL, and others (CLISP,
etc.) I was also told that there's a working version of CLX on (Open)
MCL on OS X, which (along with XDarwin) allows a lot of graphical
CL-based programs work on the Mac (though they probably already did with
CLISP).

Roger Corman, author and maintainer of Corman Lisp for Win32, was also
there, and gave a great talk on multiprocessing in CL, and how it should
be implemented. I enjoyed the talk a lot, not to mention meeting in
person.

Even though we lived close by for many years, I never met Kent Pitman
until this past week. He's someone who shares many of the same concerns
as I do, but who's thought a lot more about them. Needless to say, you
learn a lot when you talk to him.

I am a big fan of the sdf public access UNIX system, and had the
pleasure of meeting Stephen Jones, who made this project happen. He is
simply the best deal in town with respect to hosting, accounts, speed,
bandwidth, and just about anything and everything you'd ever want from
someone hosting. He's also a Lisper, though not using CL (AFAIK) in his
day job at Marconi. Stephen was really cool to meet -- someone who has
really made things happen, single-handedly, and who is now profitable
with sdf. I would *strongly* suggest using and donating to

http://sdf.lonestar.org

Stephen is looking into providing Open Genera to sdf users, since he has
a license, and since he hosts with Alphas. I would suggest that for
people who want an email account for life, to have it at sdf. I don't
just mean the address...but the account. I'll be moving mine over soon
(thanks, Stephen).

On that note I learned some history about Symbolics, and what is now
Symbolics. From several people who have used Open Genera, it seems like
the ideal Lisp environment. I don't know if it's for me, but if it goes
up on sdf, I'm there. For those who don't know, Open General supports
X, and so it pops up an X window on the Alpha into a emulated Lisp
machine (at least, this is my understanding, as someone who's never used
it). As far as you know, you're really on a Lisp machine...and don't
worry that it's an emulated environment -- it's supposedly faster than
ACL on the Alpha for several benchmarks.

The guy who told me about Open Genera, who also told me about tons of
other stuff, is Carl Shapiro, who's very young, but don't let this guy's
age fool you. He knows tons about every Lisp I've ever heard of, and
has used most quite a bit (ACL, LispWorks, Lucid (LCL), Genera, ...) and
whose work overlapped a bit with mine. He also has many Lisp machines
(20+) and is just a fun guy to ask questions to, since he knows so
much.

There were many talks...so many that I couldn't do justice. But I'll
mention some things that I was interested in.

First off, there was an incredible talk about a way to do constraint
logic programming with CL (called ConS/Lisp). It was presented by
Matthias Holzl, who designed a system, influenced by Screamer (Siskind),
in which many search-based CLP problems are fun and efficient to code.
He made intuitive and effective use of CLOS, and shows you how you can
define a langauge to decouple of the problem from strategies to solve
it. His paper should be coming soon, and I hope to read it.

I went to Duane Rettig's tutorial on simple streams, and it helped me to
see how they worked, though it didn't attempt so much to explain why a
redesign of streams was necessary, and why Grey streams was a failure
(other than saying that simple streams were much faster). However,
there wasn't much time, and all that is well documented in the simple
streams documentation available on Franz's site.

A couple of the guys from Digitool/HotDispatch were there, and spoke
quite elaborately on the HotDispatch success story. I think it's
amazing, because these guys built a really complete,
transactionally-intensive website almost 100% in CL. The HotDispatch
site is amazing, and I've been using it for years now for getting help
with simple jobs. It surprised me to see just how much of their system
was done in CL. They develop under MCL and deliver with LW. They have
their own webserver (in CL), some sort of O/R mapper, tons of services,
asynchronous notifications, a sophisticated caching layer, and
more...and it's all done in CL. An interesting story they told was that
after securing VC, they were instructed to port their system from CL to
Java, which was unusually difficult, and which made showed relative
productivity levels of a factor of 10 to 15 in favor of CL (of course,
YMMV if you try this in your company, but I'm not surprised). At the
same time, I'm working part-time on a site in the J2EE framework using
Jython primarily, and it's coming along faster than I would have
imagined (I guess reiterating Norvig's point that CL's got some stiff
competition -- at least for that sort of work). Oddly, my dream idea is
very closely related, in theory, to what the HotDispatch guys did, and
so I was naturally very interested, and glad to see their success. I
strongly recommend that people create profiles on HotDispatch if they're
looking for work (including work you can do from home, etc.)

On that note (of Lisp success stories) was ITA Software (used by
Orbitz). There was an interesting discussion of how their software
worked by one of the guys there. He discussed a bit of how they used
CL, though not exclusively CL, and how paramount speed was to their
system. The conclusion of that talk did include Lisp playing a key role
in their ongoing success, and ability to cope with the changing
environment, and with the complexity of their problem.

If you like CL, but your thing is more signals processing (forcing you
to use Matlab or something like it instead), then you might want to know
about an extremely multi-talented guy named John Amuedo, who's got a
crazy signals processing and graphics (plotting) environment in MCL. He
does all sorts of stuff with music, and has a rich background in AI.

I was glad to see that several works related and somehow overlapped with
my ongoing work in statistical pattern recognition. Jans Aasman's
recommendation engine, as well as several others made some use of
pattern recognition.

dave

Henrik Motakef

unread,
Oct 31, 2002, 4:55:34 PM10/31/02
to
Dave Bakhash <ca...@alum.mit.edu> writes:

> I just got back home, missing the last day of this year's International
> Lisp Conference. But for the people who were unable to make it there
> this year, I'd like to say a few things about it, in the hopes that
> people will find one another, and good things will result.

Thanks for your summary.

Scince a lot of the topics seem to be very interesting, I wonder if
some of them will be made available on the conference web site, or
anywhere else. Of course, if someone remebered the photo/divx idea,
pointers would be appreciated, too :-)

Regards
Henrik

Dave Bakhash

unread,
Nov 1, 2002, 1:51:13 AM11/1/02
to
Henrik Motakef <henrik....@web.de> writes:

> Thanks for your summary.

sure thing. I guess I left out that there were some opportunities for
CL programmers. I didn't get such a good look, but they were posted.
I'm sure they'll be posted here soon.

> Scince a lot of the topics seem to be very interesting, I wonder if
> some of them will be made available on the conference web site, or
> anywhere else. Of course, if someone remebered the photo/divx idea,
> pointers would be appreciated, too :-)

As I said, there should be some proceedings within the next few months.
If you go to http://www.international-lisp-conference.org and click on
speakers pages, you should be able to link to the abstracts of at least
some of the talks. If there's something you're interested in, then you
can probably contact the author directly, or (if you have trouble) post
that you're interested, and someone will get you the information you're
looking for.

Also, I did notice that Ray was taping some of the talks (including the
keynote speakers) so that might become available.

dave

Fred Gilham

unread,
Nov 1, 2002, 3:04:36 PM11/1/02
to

Dave Bakhash wrote:
> I just got back home, missing the last day of this year's
> International Lisp Conference.

Same here --- I'm skipping the wine-tasting tour in favor of work and
a home-life.... :-(

I went to the 1999 Lisp conference and gave a paper there. I thought
that one was good. This one was better, both qualitatively and
quantitatively. There were four solid days of presentations, with two
tracks in the morning on three of the days. There was one day of
tutorials with two tracks.

Some comments.

I agree with Dave regarding the organization, and particularly the
role Franz played. I personally found it transparent --- I just did
conference. I didn't have to think about anything else except which
presentations I wanted to hear. It was also fun. I probably
shouldn't say this, but the two registration helpers from Franz were
more of an attraction than some of the talks. :-)

The tutorials were a little disappointing --- the Advanced Lisp
tutorial in particular was too advanced for beginners (a beginner was
sitting right next to me and left about half way through it) and too
elementary for experienced lisp programmers. I ended up wishing I had
gone to Duane's talk on simple streams. Kent's tutorial was
interesting but not focused.

I found that many talks were sort of paired off with anti-talks (not
intentionally I'm sure). For example, Richard Stallman talked about
free software, and Kent Pitman talked about why he was not sure free
software was the right thing. Both talks were interesting and both
had points to argue with. Also, Roger Corman's excellent talk was
about how to implement OS-level threads with preemption, and it was
followed by another talk, also excellent, by Manuel Serrano, the
author of Bigloo, about why OS-level threads with preemption were a
bad idea. His code was in scheme, and I ended up calling his idea for
threads `hygenic threads' though I don't know if anyone would agree
with me on that. It did seem to be the intent, though: impose certain
disciplines on your threading mechanisms to avoid the nasty dark
corners.

One theme of the conference that I didn't necessarily like was what I
called `flying under the radar'. There was a sort of spectrum of
versions of this idea.

One talk, which I thought was going to be very different, was about
what the presenter called `ubiquitous Lisp'. Unfortunately his idea
was something like the following: Lisp is really just closures,
Javascript has closures, Javascript is everywhere, let's call
Javascript "Lisp" and declare victory. Perhaps coincidentally, the
speaker left town immediately after the talk. :-)

Another talk was called `sharpening the parentheses', and the idea was
to make Lisp into XML, then have a plugin that interpreted the
Lisp-XML in the browser. This, at least, was recognizably Lispy.

Most people were more positive about things. The slogan for these
people was "On the web nobody knows you're written in Lisp." True,
but what happened to "If you've got it, flaunt it." But I agree that
at least in these situations you can work in Lisp.

Peter Norvig, I think, has pretty much jumped ship. He's good to
listen to as someone who has gone through it, but he really isn't
pushing Lisp any more. One point he made was that at Google they
don't really need a Lisp toplevel or on-the-fly patching, because
they've got a highly redundant server farm with 10000 servers. When
they need to upgrade, they take the servers down one at a time and
upgrade them. When a server crashes, they have 9999 more to take up
the slack. What I took from his comments about this was that you
could choose between using Lisp or getting 10000 servers. :-)

Over and over again people at the conference said that they couldn't
do what they were doing without Lisp. I was glad to hear this. But
they'd also say that they either didn't emphasize Lisp or had trouble
even when successful with people complaining about their use of Lisp.

There were a few talks that I thought were incredibly good or
incredibly interesting for some reason that I want to focus on here.
An example of this was Doug Lenat speaking about Cyc. I've heard of
Lenat for years but never heard him speak. He is lucid and
entertaining as a speaker. He has devoted 20 or so years to something
that couldn't be done (according to many), namely building a huge
common-sense database that machines can use to reason about the world,
and is making a commercial success. He's hiring Lisp people ---
though the qualifications are that you have to `dream in Lisp'.
That's not so bad, but you also have to live in Texas. Oh well. Some
people reminded me that the unit of bogosity is the `micro-lenat' but
it did seem like he was accomplishing something pretty amazing.

Another speaker was Richard Greenblatt of MIT & LIM fame. Let me
preface my comments by saying that I enjoyed his talk hugely and found
it fascinating.

He started off on what I considered the wrong foot by saying that CLOS
was a bad idea because it had multiple inheritance and multiple
dispatch, and Lisp would be better off with something more like the
Objective-C object model. Having alienated most of his audience to an
extent that was audible, he then went on to speak about the origins of
life, and particularly the idea that some parts of the genetic code
are `precisely universally conserved', that is, everything alive on
the earth has these particular codes for particular proteins. The
point is that if a mutation occurs in this area, the organism isn't
viable. He used this as an argument for intelligent design. Not what
I expected at a Lisp conference, but certainly interesting. Something
I learned is that ribosomes are interpreters of the genetic code ---
they take messenger RNA and interpret it to produce the proteins it
codes for.

The conference ended with a talk by Harold Cohen, who is the author of
AARON, which is a program that he has been writing for 30 years to
create art. I think it was a bit unfortunate that he gave his talk at
the end, since only the `survivors' got to hear it. He talked about
how he had been writing this program for a long time, starting with
Fortran and then C, and had wound up in Lisp and CLOS when he needed
to get the program to work with color. The guy is a kind of genius.
He has written his program, which he says is 2mb of Lisp code, and
he's also built machines to render the paintings. He's also a painter
in his own right. He is an excellent speaker even though he's quite
old. He was obviously giving a prepared talk, but he was describing
something that was basically his life work and it showed. One amusing
aspect of his experience with the interface between technology and art
was the tendency for people to focus on the mechanisms he built to
render the art rather than the art itself. Early on he had a `turtle'
drawing machine and he had to stop using it and go to a flat-bed
plotter because people liked looking at the turtle too much. He also
had a painting machine and he said that people went bananas when the
machine started washing its brushes and paint cups when it was done.

As is always the case, one of the best things about the conference was
meeting fellow Lispers in the flesh. I got to spend a fair amount of
time talking with Rob Warnock, and I met J. P. Massar, both of whom
are (I think) pretty incredible talents.

I also met Marco Antoniotti, who in his spare time from posting to CLL
and doing other net-related stuff is actually doing world-class
bioinformatics. He's another guy that seems to be something of a
genius.

Tim Moore gave a good presentation of CLIM and McCLIM as well as
allowing me to corner him and pump him about CMUCL multiprocessing.

Daniel Barlow talked about CLiki. One problem with this conference is
that I can see that all my spare time for the next month or so is
going to be soaked up looking into all the stuff I heard people talk
about, such as CLiki.

Robert Strandh talked about GSharp, a new, usable version of which is
supposed to be coming out around Christmas. I know what I want in my
stocking! :-) GSharp is something like the TeX of music formatting,
and it's written using McCLIM.

Apart from that I met several people who were just curious about Lisp,
including a few who paid their own way to the conference just to find
out about it.

Numerous people asked me about Garnet. :-) I even brought it to show
people on my laptop. One old-time Lisper was surprised to see that it
still existed. It was fun to be recognized for something I'd done.

The conference will be held again next year, probably the second or
third week of October, in New York. I am planning on going again and
I hope to have a tutorial or presentation to give.

--
Fred Gilham gil...@csl.sri.com || His word is a creative word, and
when he speaks the good exists as good. God is neither arbitrary nor
tyrannical. He is love, and when he expresses his will it is a will
of love. Hence the good given by God is good for us.-- Jacques Ellul

Pratibha

unread,
Nov 1, 2002, 9:37:23 PM11/1/02
to
> I am a big fan of the sdf public access UNIX system, and had the
> pleasure of meeting Stephen Jones, who made this project happen. ...

> http://sdf.lonestar.org
> Stephen is looking into providing Open Genera to sdf users, since he has
> a license, and since he hosts with Alphas. ...

>
> On that note I learned some history about Symbolics, and what is now
> Symbolics. From several people who have used Open Genera, it seems like
> the ideal Lisp environment. I don't know if it's for me, but if it goes
> up on sdf, I'm there. For those who don't know, Open Genera supports

> X, and so it pops up an X window on the Alpha into a emulated Lisp
> machine (at least, this is my understanding, as someone who's never used
> it). As far as you know, you're really on a Lisp machine...and don't
> worry that it's an emulated environment -- it's supposedly faster than
> ACL on the Alpha for several benchmarks.

I assume the way this would work is to run opengenera on the hosting
Alpha with display to one's X server over the Internet? One would
need to somehow configure one's Internet connection to allow incoming
requests from an sdf X client (the hosting Alpha)? Or maybe
start a vncserver on the hosting Alpha and run a vncviewer on one's
machine to view the vncserver over the Internet? I wonder what the
interactive performance of either of these options would be,
assuming that this is the intended usage mode...

Also, the environment would be Symbolics Common Lisp/Flavors/Zmacs,
not ANSI-CL/CLOS/emacs, right?

Kenny Tilton

unread,
Nov 1, 2002, 11:49:08 PM11/1/02
to

Dave Bakhash wrote:

<a nice summary, including...>

> There's a lot to say, so first off, I'll start by saying that it was
> very well organized.

Hear! Hear! This was a serious "contrib" to CL.

> For me, it was amazing to finally meet people I've been talking to
> directly or indirectly for years.

Including John McCarthy in a cameo role!!

>
> Anther big speaker (for me) was Peter Norvig. ... I shared his feelings that Common Lisp,


> though having many advantages over other programming languages, was no
> longer alone, and that many of the key features that CL had that were

> not found in other languages have been adopted...

That bit bothered me. PN listed 8 cool features of CL and said of those
6 had been replicated by /some/ language. But!! No /one/ new language
has more than a few. Break it down by new language. All of a sudden we
discover that, just to get to the 80% point of "replicating" CL you need
to drag in four languages, each one contributing at most 30% of CL. (All
preceding numbers fabricated but close enough for government work.)

I would further whine that the check list did not include little things
like "fast" or "standardized". Granted, CL did not invent speed, but
none of the so-called Lisp competitors were fast, and that is one of the
huge objections to them. Not on Norvig's list.

btw, McCarthy nailed Norvig on another missing item from the list, viz,
code as data.

All in all a shabby presentation. Is Python fast? Does it have code as
data? Decent (not ref counting) GC? Macros? chya.

> Digitool's MCL is coming soon to OS X

woo-hoo!

> Even though we lived close by for many years, I never met Kent Pitman
> until this past week.

Kent was great, but I just don't get his anti-free thing. Jim Croce said
it: you don't spit into the wind. And his example of his http server
being made unmarketable by AllegroServe... well, there is also cl-http.
Not that that is "free", but it certainly seems like it might be a
commodity in that it is easy to produce. Potential investors demand of
our little startup: can't someone else replicate your work, eliminating
your competitive advantage? They are not saying the replicators will
give their work away, they just don't want us to be delivering a
commodity which anyone (esp. a groilla like Oracle) can churn out. They
want to hear that what we did is hard or patented (fwiw).

>
> There were many talks...so many that I couldn't do justice. But I'll
> mention some things that I was interested in.

[all good ones]

I got a big kick out of LinJ, the Lisp-to-Java source translator. I have
visions of porting Cells to Java via LinJ, which IIRC they said they
would share.

Interesting double-standard: Gabriel denounced OO unchallenged,
but Greenblatt was actually hissed (me included!) for dissing multiple
inheritance.

Back to Gabriel, I was fascinated to learn how hard it was to get a
degree in poetry, which boiled down to writing and reading a ton.
gabriel bemoaned how little we ask of comp sci grads, but it occurred to
me that since we do not yet know how to program (hey, biggies like
Gabriel, Greenblatt, and Graham are wrong about OO) we might be excused
for not yet knowing how to teach it.

Wake-up call: Imran Shah (I believe it was) of U of Colorado, Boulder
ended a nice talk by saying one of the only things wrong with CL was
that you cannot find CLers. We did a sanity check and discovered he had
advertised only locally. No posting on the Franz site, nothing here on
c.l.l., I imagine nothing on monster or dice or hotjobs... Shah was
thrilled to learn that he had just been looking in the wrong place, and
I think we need to think about how we can prevent this int he future.

which brings me to the ALU, which is getting ready to step things up,
and maybe one thing they (we!) can do is host a members-only job hunting
area on the soon-to-be reinvigorated web site, and make the hell sure
people know about it.

btw, IIRC the ALU voted fer sher to hold the next conference (next year)
in my adopted home town, the Big "if you can make here, you'll make it
anywhere" Apple, the city so great they named it twice, New Yawk, New
Yawk. y'all can stay at my place.

more as i think of it, but all in all it was an astonishing event. ray,
the alu and franz deserve (and got) a lot of credit. the coordination
and content were simply over the top excellent. me, i was there just to
see the great roster of lispers, but in the end the best part for me was
rubbing elbows at last with other trench-diggers such as myself.

now i gotta get to work on browsing the half-dozen or more cool sites I
learned about this week.

--

kenny tilton
clinisys, inc
---------------------------------------------------------------
""Well, I've wrestled with reality for thirty-five years, Doctor,
and I'm happy to state I finally won out over it.""
Elwood P. Dowd

Abhijit Rao

unread,
Nov 2, 2002, 12:46:36 AM11/2/02
to
On 31 Oct 2002 22:55:34 +0100, Henrik Motakef <henrik....@web.de>
wrote:

>Dave Bakhash <ca...@alum.mit.edu> writes:
>
>> I just got back home, missing the last day of this year's International
>> Lisp Conference. But for the people who were unable to make it there
>> this year, I'd like to say a few things about it, in the hopes that
>> people will find one another, and good things will result.
>
>Thanks for your summary.

Thanks! Yessir!

>Scince a lot of the topics seem to be very interesting, I wonder if
>some of them will be made available on the conference web site, or
>anywhere else. Of course, if someone remebered the photo/divx idea,
>pointers would be appreciated, too :-)

aye aye !! Please some photos! I am reeeeeealy curious to /see/ all
these people.

--
quasi
http://abhijit-rao.tripod.com/digital/lisp.html

"I slept with Faith, and found a corpse in my arms on awaking; I drank and danced all night with Doubt, and found her a virgin in the morning."
~ A. Crowley

Marc Spitzer

unread,
Nov 2, 2002, 12:50:19 AM11/2/02
to
Kenny Tilton <kti...@nyc.rr.com> wrote in
news:3DC35AB2...@nyc.rr.com:

> btw, IIRC the ALU voted fer sher to hold the next conference (next
> year) in my adopted home town, the Big "if you can make here, you'll
> make it anywhere" Apple, the city so great they named it twice, New
> Yawk, New Yawk. y'all can stay at my place.

Cool, I can pay for the conference and skip the hotel and airfair.

Now since there apears to be at least 3 people doing, or in my case trying
to do, lisp in NYC area any chance of having a meeting and looking into
starting a users group? I did not see an chapters listed on the website.

marc

Marc Spitzer

unread,
Nov 2, 2002, 12:52:08 AM11/2/02
to
Abhijit Rao <quas...@yahoo.com> wrote in
news:6kp6sucum8k1k87di...@4ax.com:

Do not forget the sign in helpers. ;->

marc

Rob Warnock

unread,
Nov 2, 2002, 3:12:06 AM11/2/02
to
Marc Spitzer <mspi...@optonline.net> wrote:
+---------------
| Kenny Tilton <kti...@nyc.rr.com> wrote:
| > ...New Yawk. y'all can stay at my place.

|
| Cool, I can pay for the conference and skip the hotel and airfair.
+---------------

Speaking of hotels: this just in from the RISKS Digest:

<URL:http://catless.ncl.ac.uk/Risks/22.33.html#subj6.1>
...
If you stayed at a Holiday Inn, Holiday Inn Express, or Crowne
Plaza hotel and checked out between 24 Oct and 26 Oct 2002,
you are likely to have been one of 26,000 people who were
charged 100 times what they owed, such as $6,500 to $21,000
per night. A credit-processing error resulted in the decimal
points being dropped. Most of the charges were later reversed,
although many people discovered that their credit limits had
been exhausted. Overcharged guests will get two free nights
at any of those hotels.

Any attendees who stayed in the conference hotel [a Holiday Inn],
left early [between 10/24 & 10/26], and didn't yet look at their
bill should check your hotel bill carefully... :-(


-Rob

-----
Rob Warnock, PP-ASEL-IA <rp...@rpw3.org>
627 26th Avenue <URL:http://www.rpw3.org/>
San Mateo, CA 94403 (650)572-2607

Arthur Lemmens

unread,
Nov 2, 2002, 4:56:12 AM11/2/02
to

Kenny Tilton wrote:

> now i gotta get to work on browsing the half-dozen or more cool sites I
> learned about this week.

Can you give us some URL's?
Thanks.

Arthur Lemmens

Rob Warnock

unread,
Nov 2, 2002, 5:28:39 AM11/2/02
to
Marc Spitzer <mspi...@optonline.net> wrote:
+---------------
| Kenny Tilton <kti...@nyc.rr.com> wrote:
| > ...New Yawk. y'all can stay at my place.

|
| Cool, I can pay for the conference and skip the hotel and airfair.
+---------------

Speaking of hotels, this just in from the RISKS Digest:

<URL:http://catless.ncl.ac.uk/Risks/22.33.html#subj6.1>
...
If you stayed at a Holiday Inn, Holiday Inn Express, or Crowne
Plaza hotel and checked out between 24 Oct and 26 Oct 2002,
you are likely to have been one of 26,000 people who were
charged 100 times what they owed, such as $6,500 to $21,000
per night. A credit-processing error resulted in the decimal
points being dropped. Most of the charges were later reversed,
although many people discovered that their credit limits had
been exhausted. Overcharged guests will get two free nights
at any of those hotels.

The conference hotel was a Holiday Inn, but given that ILC started on
the 27th, looks like we just missed getting bitten by this one... ;-}

Abhijit Rao

unread,
Nov 2, 2002, 6:56:08 AM11/2/02
to
On Sat, 02 Nov 2002 04:49:08 GMT, Kenny Tilton <kti...@nyc.rr.com>
wrote:

>btw, IIRC the ALU voted fer sher to hold the next conference (next year)
>in my adopted home town, the Big "if you can make here, you'll make it
>anywhere" Apple, the city so great they named it twice, New Yawk, New
>Yawk. y'all can stay at my place.

Unashamedly extending that invitation to meself, I have to make
preparations to start immediately in order to reach Nook Yawk in time.
But I am prepared to stay in a tent on your lawn/terrace if you
promise to extend a cat5 cable to it.

Paolo Amoroso

unread,
Nov 2, 2002, 7:25:41 AM11/2/02
to
On 31 Oct 2002 22:55:34 +0100, Henrik Motakef <henrik....@web.de>
wrote:

> Scince a lot of the topics seem to be very interesting, I wonder if


> some of them will be made available on the conference web site, or

A couple of papers are already available:

A Free Implementation of CLIM
http://www.bricoworks.com/~moore/clim-paper.pdf

CLiki: collaborative content management for community web sites
http://ww.telent.net/ilc-slides/paper.tar


Paolo
--
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README

Kenny Tilton

unread,
Nov 2, 2002, 8:50:27 AM11/2/02
to

Yep, I am thinking we call ourselves the International Lisp Users Group,
New York chapter. ILUG-NY (pronounced to the tune of "I Love NY", the
tourism promotion jingle.)

The format should be a big table with pitchers of beer in the middle.
McSorley's?

kenny
clinisys

Dave Bakhash

unread,
Nov 2, 2002, 9:58:35 AM11/2/02
to
Marc Spitzer <mspi...@optonline.net> wrote in message news:

> > btw, IIRC the ALU voted fer sher to hold the next conference (next
> > year) in my adopted home town, the Big "if you can make here, you'll
> > make it anywhere" Apple, the city so great they named it twice, New
> > Yawk, New Yawk. y'all can stay at my place.
>
> Cool, I can pay for the conference and skip the hotel and airfair.
>
> Now since there apears to be at least 3 people doing, or in my case trying
> to do, lisp in NYC area any chance of having a meeting and looking into
> starting a users group? I did not see an chapters listed on the website.

I'm all for this. I'll call an open meeting, and will reserve a
conference room available for it right in midtown manhattan, 5 minutes
from Penn Station. This first one in NYC should just be to bring the
people together, such that we can get to know each other, and what
we're all doing, and possibly if we can help each other.

I know for sure that there are several Lispers in Long Island
(including Carl Shapiro). There's also Marco, Kenny, and of course
Raymond in the city proper. People can bring their laptops if they
want, and demo stuff. I'll be able to demo lots of stuff, will fire
up a transaction system, show POS interfacing stuff, IVR, and some of
my pet projects (e.g. XStrokes).

What I might do as well is to let some people know at NYU, so students
there who might know a little about CL, but don't have course
offerings can come too. We can start organizing some volutary
instructional sessions, etc. I tought a course in CL at BU,
voluntarily (the CS department donated a room to me for a couple of
hours once a week), and it was successful with over 20 students,
starting with about 30.

It's fascinating to ask students to bring their problem sets to class,
and show them how what they regularly spend 6 hours on in C or C++
usually takes no more than 15-30 minutes in CL. I used to do that,
but then stopped, because I would rather teach CL in its own right
than comparitively. I know that at least two of those people
continued to use CL afterwards, and fired up ACL regularly thereafter.

I'll try to gather a mailing list of interested people. In the
meantime, people can let me know what their schedules look like. Feel
free to send me email.

dave

Dave Bakhash

unread,
Nov 2, 2002, 10:07:17 AM11/2/02
to
Kenny Tilton <kti...@nyc.rr.com> writes:

> > Now since there apears to be at least 3 people doing, or in my case
> > trying to do, lisp in NYC area any chance of having a meeting and
> > looking into starting a users group?
>
> Yep, I am thinking we call ourselves the International Lisp Users
> Group, New York chapter. ILUG-NY (pronounced to the tune of "I Love
> NY", the tourism promotion jingle.)

I'm all for it. I love the name ILUG, and ILUG-NY is perfect, though a
quick check shows that we won't get the name ilug.org. I posted already
that I will reserve a conference room in midtown, 5 min. from Penn
Station, and all are invited.

> The format should be a big table with pitchers of beer in the
> middle. McSorley's?

I guess Java would be better (i.e. "IJUG" [of beer].)

I would like to round up as many of the NYC metro Lispers, Long
Islanders, etc. I only know of a few, though: Kenny, Carl, Ray, Marc,
Marco...I know I'm missing some for sure.

I'll write up a formal posting about it.

dave

Kenny Tilton

unread,
Nov 2, 2002, 10:44:44 AM11/2/02
to
Arthur Lemmens wrote:
>
> Kenny Tilton wrote:
>
>
>>now i gotta get to work on browsing the half-dozen or more cool sites I
>>learned about this week.
>
>
> Can you give us some URL's?

Good idea. (btw, if later the ALU ends up sharing the conference
proceedings, everyone should dive into that so they also find the stuff
which happened not to light my personal fire.)

caveat: this will be a little embarrassing for me since I will be saying
stuff like "OK, one cool thing I discovered was this place called
Google"... well, not /that/ bad, but close enough. I kinda live in this
extremely isolated techno-niche (and am only now thinking about checking
slashdot each day) so I have some wicked blind spots.

The last (speaker) shall be first: I did not know the painting program
Aaron was available as shareware:

http://www.kurzweilcyberart.com/KCATaaron/DOWNLOADbasic

Harold Cohen, the author, gave a nice talk to end the conference.

Also, I can't say I had not seen this place before, but it was nice to
get reminded of:

http://www.hotdispatch.com/home.html

I am thinking about at long last kicking off a project to produce a
portable gui (X, Win32 and Mac) and HotDispatch might be a neat way to
get help with the effort if I do not want to go the open route.

More to come.

Kenny Tilton

unread,
Nov 2, 2002, 11:07:09 AM11/2/02
to
Abhijit Rao wrote:
> On Sat, 02 Nov 2002 04:49:08 GMT, Kenny Tilton <kti...@nyc.rr.com>
> wrote:
>
>
>> y'all can stay at my place.
>
> But I am prepared to stay in a tent on your lawn/terrace if you
> promise to extend a cat5 cable to it.

"We have a pool and a pond. The pond would be good for you."
-- Chevy Chase to Bill Murray, Caddy Shack

:)

the tent might work. Central Park is nearby, and if you set up in
Strawberry Field you would /not/ be the only one there overnight.

seriously, I have here a copy of "Sleep Cheap in NY", but when I offered
it to a visitor he said he could find all that on-line. But one
recommended site picked at random:

http://www.centralparkhostel.com/

Abhijit Rao

unread,
Nov 2, 2002, 1:22:31 PM11/2/02
to
On Sat, 02 Nov 2002 16:07:09 GMT, Kenny Tilton <kti...@nyc.rr.com>
wrote:

>"We have a pool and a pond. The pond would be good for you."


>-- Chevy Chase to Bill Murray, Caddy Shack

It will be a new an enriching experience, I am sure - this pond.

>
>:)
>
>the tent might work. Central Park is nearby, and if you set up in
>Strawberry Field you would /not/ be the only one there overnight.

What a nice (polite :very) way of going about it. :) you need not have
worried though - it will take me years before I reach Noo York (if I
get there at all). They tell me I will have to cross several oceans
to get there - not to mention unnamed dangers in the desert lands that
lie in between. But no fear! I have my trusty Victorinox by my side
- we both shall face the dangers together and unflinchingly!

>seriously, I have here a copy of "Sleep Cheap in NY", but when I offered
>it to a visitor he said he could find all that on-line. But one
>recommended site picked at random:
>
> http://www.centralparkhostel.com/

Your kindness has inspired me to extend an open invitation to all
lispers who may be passing through my city - it is the least I can do.
Onwards to Utopia!

Kenny Tilton

unread,
Nov 2, 2002, 2:45:43 PM11/2/02
to

Abhijit Rao wrote:
> On Sat, 02 Nov 2002 16:07:09 GMT, Kenny Tilton <kti...@nyc.rr.com>
> wrote:
>
>
>>"We have a pool and a pond. The pond would be good for you."
>>-- Chevy Chase to Bill Murray, Caddy Shack
>
>
> It will be a new an enriching experience, I am sure - this pond.

No chlorine, for sure. You know, W and I visited Alcatraz on Tutorial
Day, and I have an idea.

Alcatraz was meant for the criminals other prisons did not want. Maybe I
should offer my roof just to those who have been keel-hauled on c.l.l.,
such as Quasi, Ilias, JB, and Mel. Maybe a lifetime achievement spot for
Xah, as well. We could run an anti-conference on my front stoop...

Hmmmm....

JB

unread,
Nov 2, 2002, 3:17:41 PM11/2/02
to
Kenny Tilton wrote:
> Alcatraz was meant for the criminals other prisons did not
> want. Maybe I should offer my roof just to those who have
> been keel-hauled on c.l.l., such as Quasi, Ilias, JB, and
> Mel. Maybe a lifetime achievement spot for Xah, as well.
> We could run an anti-conference on my front stoop...

I have been treated well here. I do not know what you mean.
--
JB

Kenny Tilton

unread,
Nov 2, 2002, 3:50:07 PM11/2/02
to

OK, I'll see if I can get IRT instead.

Actually, everyone in the list has been well-treated here. The
"keel-haul" analogy was selected after several milliseconds of careful
thought.

Bill Clementson

unread,
Nov 2, 2002, 4:29:20 PM11/2/02
to
Abhijit Rao <quas...@yahoo.com> writes:

> On 31 Oct 2002 22:55:34 +0100, Henrik Motakef <henrik....@web.de>
> wrote:
> >Scince a lot of the topics seem to be very interesting, I wonder if
> >some of them will be made available on the conference web site, or
> >anywhere else. Of course, if someone remebered the photo/divx idea,
> >pointers would be appreciated, too :-)
>
> aye aye !! Please some photos! I am reeeeeealy curious to /see/ all
> these people.

Some pictures that I took are at:
http://lisp.home.attbi.com/ilc_2002.htm

I believe that ALU will be making available videos & a CD at a later
date.
--
Bill Clementson

Jim White

unread,
Nov 2, 2002, 4:36:29 PM11/2/02
to
Dave Bakhash wrote:

> Kenny Tilton <kti...@nyc.rr.com> writes:
>>Yep, I am thinking we call ourselves the International Lisp Users
>>Group, New York chapter. ILUG-NY (pronounced to the tune of "I Love
>>NY", the tourism promotion jingle.)
>
> I'm all for it. I love the name ILUG, and ILUG-NY is perfect, though a
> quick check shows that we won't get the name ilug.org. I posted already
> that I will reserve a conference room in midtown, 5 min. from Penn
> Station, and all are invited.

Well, thanks to Mr. Newman, we've got a whole song for ILUG-LA! ;-)

Jim

Will Deakin

unread,
Nov 2, 2002, 5:04:59 PM11/2/02
to
Jim White wrote:

> Well, thanks to Mr. Newman, we've got a whole song for ILUG-LA! ;-)

I presume this is Randy rather than, say, Cardinal...

;)w

Vlad S.

unread,
Nov 2, 2002, 7:51:20 PM11/2/02
to
Dave Bakhash <ca...@alum.mit.edu> wrote in message news:<c29fzum...@no-knife.mit.edu>...


> On that note I learned some history about Symbolics, and what is now
> Symbolics.

Pray tell, what did David Schmidt say about the status of Symbolics?

Abhijit Rao

unread,
Nov 3, 2002, 12:37:05 AM11/3/02
to
On Sat, 02 Nov 2002 19:45:43 GMT, Kenny Tilton <kti...@nyc.rr.com>
wrote:

>Alcatraz was meant for the criminals other prisons did not want. Maybe I

>should offer my roof just to those who have been keel-hauled on c.l.l.,
>such as Quasi, Ilias, JB, and Mel.

I was not aware I was part of any such prestigious group. I sincerely
thank you for opening my eyes to membership. It is such pleasure to
see that one's light hearted attempts at humours also can actually
land him into the esoteric league. I must be lucky or what?

> Maybe a lifetime achievement spot for
>Xah, as well. We could run an anti-conference on my front stoop...

anti-conference? Unfortunately having fought hard to get to do Lisp I
would have to politely decline. But I /do/ get the point.

I would also request one more thing from you (what? another one?? die
of shame you mosquito!). I would be honored by my entry into your
killfile. I would not like you to have to associate with criminals.

Matthew X. Economou

unread,
Nov 3, 2002, 12:19:49 AM11/3/02
to
>>>>> "Pratibha" == Pratibha <ibpra...@yahoo.com> writes:

Pratibha> I assume the way this would work is to run opengenera on
Pratibha> the hosting Alpha with display to one's X server over
Pratibha> the Internet? One would need to somehow configure one's
Pratibha> Internet connection to allow incoming requests from an
Pratibha> sdf X client (the hosting Alpha)?

VNC probably isn't a good idea, as it is relantively insecure, not to
mention slow (even compared to X11). I've seen decent performance
using X11 connections forwarded over SSH tunnels, especially with a
fast cipher selected (e.g. Blowfish) and gzip-compression enabled. Of
course, this is over a broadband Internet connection, not dialup.

That said, I'd gladly slog through a slow-ass X11 connection to play
with Genera. I would pay money to support the server on the other
end, too.

--
Matthew X. Economou <xeno...@irtnog.org> - Unsafe at any clock speed!
I'm proud of my Northern Tibetian heritage! (http://www.subgenius.com)
Max told his friend that he'd just as soon not go hiking in the hills.
Said he, "I'm an anti-climb Max." [So is that punchline.]

Kenny Tilton

unread,
Nov 3, 2002, 1:12:58 AM11/3/02
to

Abhijit Rao wrote:
> I would also request one more thing from you (what? another one?? die
> of shame you mosquito!). I would be honored by my entry into your
> killfile. I would not like you to have to associate with criminals.

That does it, the anti-conference is off. What a prickly, humorless lot!

JB

unread,
Nov 3, 2002, 5:30:20 AM11/3/02
to
Kenny Tilton wrote:

> OK, I'll see if I can get IRT instead.
>
> Actually, everyone in the list has been well-treated here.
> The "keel-haul" analogy was selected after several
> milliseconds of careful thought.

Hey Kenny!
There are those who are good and those who are evil. Those
who disclose their heretic views are always evil.

It must be good for you to belong to those who are good!

--
JB

"Die töricht genug ihr volles Herz nicht wahrten,
Dem Pöbel ihr Gefühl, ihr Schauen offenbarten,
Hat man von je gekreuzigt und verbrannt."

(Goethe: Faust I)

Abhijit Rao

unread,
Nov 3, 2002, 6:27:15 AM11/3/02
to
On Sun, 03 Nov 2002 06:12:58 GMT, Kenny Tilton <kti...@nyc.rr.com>
wrote:

>That does it, the anti-conference is off. What a prickly, humorless lot!

Ha. What a humour. The same when people find it so funny to corner a
junior in the bathroom and say "hey weener you are sooo short - you
cant even kiss and do it at the same time." The same as when people
find it so humorous to go bomb every tom dick & harry. "Fakin evil
people - lets bomb them & see them roast. hahaha."

Unfortunately this sub-thread has skewed away badly. *sigh*

Tim Bradshaw

unread,
Nov 3, 2002, 8:27:12 AM11/3/02
to
* Matthew X Economou wrote:

> VNC probably isn't a good idea, as it is relantively insecure, not to
> mention slow (even compared to X11).

VNC tunnelled over ssh is as secure as ssh. I've used VNC over a 64k
line, and it's basically OK. For many modern X applications it's
actually considerably faster than X, because it only sends stuff which
actually involves a visual change on the screen, unlike X applications
which will happily send 3 billion futile requests to the server before
they do anything at all in the way of making a window appear (trace
netscape or xemacs sometime, and I bet really modern things are much
worse than this).

--tim

Will Deakin

unread,
Nov 3, 2002, 8:36:26 AM11/3/02
to
Abhijit Rao wrote:

> Ha. What a humour. The same when people find it so funny to corner a
> junior in the bathroom and say "hey weener you are sooo short - you
> cant even kiss and do it at the same time." The same as when people
> find it so humorous to go bomb every tom dick & harry. "Fakin evil
> people - lets bomb them & see them roast. hahaha."

Please could you translate.

> Unfortunately this sub-thread has skewed away badly.

There is a (relatively) popular humourous radio show in this country
called `just a minute' in which a contestant must talk on a subject for
60 seconds without hesitation, repetition or deviation. Other contestant
can challenge on the basis of this. I think the above would probably
fall under the category of `deviation from the english language...'

;)w

Abhijit Rao

unread,
Nov 3, 2002, 9:12:32 AM11/3/02
to
On Sun, 3 Nov 2002 13:36:26 +0000 (UTC), Will Deakin
<aniso...@hotmail.com> wrote:
>Abhijit Rao wrote:
>
>> Ha. What a humour. The same when people find it so funny to corner a
>> junior in the bathroom and say "hey weener you are sooo short - you
>> cant even kiss and do it at the same time." The same as when people
>> find it so humorous to go bomb every tom dick & harry. "Fakin evil
>> people - lets bomb them & see them roast. hahaha."
>
>Please could you translate.

Could you please tell me what part you want translated and into which
language? '-) The above is correct English I suppose (except for the
quoted bits which are exempt from grammar). :-)

>> Unfortunately this sub-thread has skewed away badly.
>
>There is a (relatively) popular humourous radio show in this country
>called `just a minute' in which a contestant must talk on a subject for
>60 seconds without hesitation, repetition or deviation. Other contestant
>can challenge on the basis of this. I think the above would probably
>fall under the category of `deviation from the english language...'

I fear I have to admit that I don't see the 'deviation...' bit here.
You may have to be more elaborate if I have to understand. ;) BTW
which country?

>;)w

Paolo Amoroso

unread,
Nov 3, 2002, 10:06:40 AM11/3/02
to
On Sat, 02 Nov 2002 13:25:41 +0100, Paolo Amoroso <amo...@mclink.it>
wrote:

> A Free Implementation of CLIM
> http://www.bricoworks.com/~moore/clim-paper.pdf

Note that initial support for graph drawing was implemented since the paper
was written.

Dave Bakhash

unread,
Nov 3, 2002, 6:25:48 PM11/3/02
to
voodo...@hotmail.com (Vlad S.) writes:

I didn't hear anything specific about Symbolics, except that what's left
are three people, and that they maintain Open Genera. I have never used
their Lisp, so I know very little about it.

Still, from the way people who actually use Open Genera talk about it,
it seems like the ideal Lisp platform to develop under.

dave

Greg Menke

unread,
Nov 4, 2002, 1:22:31 AM11/4/02
to

They're a pretty small operation now, but still have some long-term
customers. Due to contract constraints, their price for Genera is
fixed- pretty much right out of the hobbyist range unfortunately.
They have an emulator that runs on Alpha hardware if you don't have
native hardware handy. They have a good bit of their own hardware on
hand, parts & some facilities to repair them. They try to keep track
of their existing hardware, attempting to keep it out of the
landfills. Software-wise, they bring in previous Symbolics people
from time to time to handle updates & bug fixes, etc- however they are
not in a position for large scale development. Unless you're willing
to pay for it I imagine...

They've been toying with ideas related to either augmenting a
commodity 64 bit processor (PowerPC, Itanic...) to be somewhat close
to "Lisp-ready", or to pick up funding to roll a custom processor.
Personally, I doubt there's much of a market for a custom processor
these days- but what do I know. I think it might be easier to sell a
PCI based coprocessor board with some suitably tweaked Sparc/PowerPC,
all loaded up w/ ram and custom fpga's to make the thing Lispy. Some
of the bio-informatics presentations sure sounded as if something like
that might be handy.

Mr. Schmidt gave a fascinating presentation, I'm glad I made it. He
even had a drawing for one of their boards and a copy of Genera.
Unfortunately I had neither business cards or a handy piece of paper
on which to scratch my name...

Gregm

Reini Urban

unread,
Nov 4, 2002, 5:28:59 AM11/4/02
to
Bill Clementson schrieb:

Uh. Roger Corman cut his hair!

--
Reini Urban - Programmer - http://inode.at

Frode Vatvedt Fjeld

unread,
Nov 4, 2002, 6:42:18 AM11/4/02
to
Greg Menke <gregm...@toadmail.com> writes:

> [..] They've been toying with ideas related to either augmenting a


> commodity 64 bit processor (PowerPC, Itanic...) to be somewhat close
> to "Lisp-ready", or to pick up funding to roll a custom processor.
> Personally, I doubt there's much of a market for a custom processor
> these days- but what do I know. I think it might be easier to sell
> a PCI based coprocessor board with some suitably tweaked
> Sparc/PowerPC, all loaded up w/ ram and custom fpga's to make the
> thing Lispy. Some of the bio-informatics presentations sure sounded
> as if something like that might be handy.

Is there really still a perceived need for such "lispy" hardware? The
current mass-produced CPUs are amazingly fast, and if there is a 25%
(or even more) overhead of mapping some run-time environment to these
CPUs, that's still 75% of something that's constantly getting faster,
at no effort on "our" part. Other than slightly reduced code-size and
somewhat simpler compilers, what would be the expected gain of such
hardware?

--
Frode Vatvedt Fjeld

Greg Menke

unread,
Nov 4, 2002, 8:14:04 AM11/4/02
to

I think the theory is to the extent Lispy hardware offers some useful
performance increment, someone will be willing to buy it. Personally,
I doubt that any significant number of people would be particularly
interested in a custom processor, but there might be some interested
in a really fast Lisp coprocessor. I think code size and simpler
compilers don't really enter into it these days. On the other hand,
one of the bio-informatics guys said he'd be glad to buy some kind of
board that would give a reasonable performance boost. I think the key
is the machine instruction level implementation of tag bits and/or
other low level Lisp features that are typically emulated via higher
level abstractions on general purpose hardware. Even if this made a
huge speed improvement, the board would have to be cheap to build
because not too many people would be interested. If Java could be run
better/faster on the board too, it might reach a larger market. The
idea of running full safety code as fast or faster than no safety code
on conventional processors is an appealing argument.

Gregm

Frode Vatvedt Fjeld

unread,
Nov 4, 2002, 8:38:55 AM11/4/02
to
Greg Menke <gregm...@toadmail.com> writes:

> The idea of running full safety code as fast or faster than no
> safety code on conventional processors is an appealing argument.

It's certainly appealing, but not very realistic.

--
Frode Vatvedt Fjeld

Kenny Tilton

unread,
Nov 4, 2002, 10:54:06 AM11/4/02
to

I think in the long run (we're talking macro, not micro) what we are
learning is that tag bits are handy, so someday we'll have them, if only
to support Java. That's OK, as long as Lisp compilers can leverage them.
But if I am right, we'll see that only when Intel or Motorola decides
to put them in. I wonder if Motorola is looking for a way to grow share? :)

Raymond Toy

unread,
Nov 4, 2002, 11:48:17 AM11/4/02
to
>>>>> "Kenny" == Kenny Tilton <kti...@nyc.rr.com> writes:

Kenny> Frode Vatvedt Fjeld wrote:
>> Greg Menke <gregm...@toadmail.com> writes:
>>

>>> The idea of running full safety code as fast or faster than no
>>> safety code on conventional processors is an appealing argument.
>> It's certainly appealing, but not very realistic.

>>


Kenny> I think in the long run (we're talking macro, not micro) what we are
Kenny> learning is that tag bits are handy, so someday we'll have them, if
Kenny> only to support Java. That's OK, as long as Lisp compilers can
Kenny> leverage them. But if I am right, we'll see that only when Intel or
Kenny> Motorola decides to put them in. I wonder if Motorola is looking for a
Kenny> way to grow share? :)

FWIW, Sparcs have instructions to support tag bits. Two of those
instructions have been deprecated in the Sparc V9. And they didn't
extend the instruction for 64-bit support.

Ray

Michael Hudson

unread,
Nov 4, 2002, 1:06:40 PM11/4/02
to
Will Deakin <aniso...@hotmail.com> writes:

Or William Harold...

Cheers,
M.

--
I'm about to search Google for contract assassins to go to Iomega
and HP's programming groups and kill everyone there with some kind
of electrically charged rusty barbed thing.
-- http://bofhcam.org/journal/journal.html, 2002-01-08

Eric Thorsen

unread,
Nov 4, 2002, 2:00:24 PM11/4/02
to
I'd be very interested in an NYC lisp user group. I'm one of 2 people
I know in Westchester county using Lisp and would like to do whatever
I can to spread the word. My schedule is fairly open but evenings
Mon-Fri are best.

Thanks,
Eric

ca...@alum.mit.edu (Dave Bakhash) wrote in message news:<8a3667a0.0211...@posting.google.com>...
> Marc Spitzer <mspi...@optonline.net> wrote in message news:
> > > btw, IIRC the ALU voted fer sher to hold the next conference (next
> > > year) in my adopted home town, the Big "if you can make here, you'll
> > > make it anywhere" Apple, the city so great they named it twice, New
> > > Yawk, New Yawk. y'all can stay at my place.
> >
> > Cool, I can pay for the conference and skip the hotel and airfair.
> >
> > Now since there apears to be at least 3 people doing, or in my case trying
> > to do, lisp in NYC area any chance of having a meeting and looking into
> > starting a users group? I did not see an chapters listed on the website.
>
> I'm all for this. I'll call an open meeting, and will reserve a
> conference room available for it right in midtown manhattan, 5 minutes
> from Penn Station. This first one in NYC should just be to bring the
> people together, such that we can get to know each other, and what
> we're all doing, and possibly if we can help each other.
>
> I know for sure that there are several Lispers in Long Island
> (including Carl Shapiro). There's also Marco, Kenny, and of course
> Raymond in the city proper. People can bring their laptops if they
> want, and demo stuff. I'll be able to demo lots of stuff, will fire
> up a transaction system, show POS interfacing stuff, IVR, and some of
> my pet projects (e.g. XStrokes).
>
> What I might do as well is to let some people know at NYU, so students
> there who might know a little about CL, but don't have course
> offerings can come too. We can start organizing some volutary
> instructional sessions, etc. I tought a course in CL at BU,
> voluntarily (the CS department donated a room to me for a couple of
> hours once a week), and it was successful with over 20 students,
> starting with about 30.
>
> It's fascinating to ask students to bring their problem sets to class,
> and show them how what they regularly spend 6 hours on in C or C++
> usually takes no more than 15-30 minutes in CL. I used to do that,
> but then stopped, because I would rather teach CL in its own right
> than comparitively. I know that at least two of those people
> continued to use CL afterwards, and fired up ACL regularly thereafter.
>
> I'll try to gather a mailing list of interested people. In the
> meantime, people can let me know what their schedules look like. Feel
> free to send me email.
>
> dave

Christopher C. Stacy

unread,
Nov 5, 2002, 1:21:53 AM11/5/02
to
If this thing happens, I'd like to hear about it well in advance.

Roger Corman

unread,
Nov 5, 2002, 3:22:11 AM11/5/02
to
On Mon, 04 Nov 2002 11:28:59 +0100, Reini Urban <rur...@inode.at> wrote:

>Uh. Roger Corman cut his hair!

Guilty as charged! About a year ago. I updated my photo on the cormanlisp
web site quite back then as well.

I am sorry you weren't at the conference, Chris. Your name came up in a number
of conversations. I know it would have been a very long trip for you.

Roger

vsync

unread,
Nov 5, 2002, 11:53:39 AM11/5/02
to
Fred Gilham <gil...@snapdragon.csl.sri.com> writes:

> Most people were more positive about things. The slogan for these
> people was "On the web nobody knows you're written in Lisp." True,
> but what happened to "If you've got it, flaunt it." But I agree that
> at least in these situations you can work in Lisp.

[...]

> Over and over again people at the conference said that they couldn't
> do what they were doing without Lisp. I was glad to hear this. But
> they'd also say that they either didn't emphasize Lisp or had trouble
> even when successful with people complaining about their use of Lisp.

I've had the same problem. I'm currently working on a Lisp-based site
and content management suite thing (can't really come up with a more
concise definition than that ATM) using mod_lisp. (This has curtailed
my Usenet postings for some months now, as I finally realized that
it's far more productive to do nothing but hack Lisp than to do
nothing but talk about Lisp incessantly. I tend to be fairly
obsessive, so for me it's one or the other in the early phases.) My
primary goal is to have something I like for myself to use, but
secondarily I hope to either show it off as part of my portfolio for
jobs requiring knowledge of Lisp-based or similar languages, or to use
the codebase for consulting, charge for support, modifications, etc.

Anyway, I've got it to a point where it works tolerably fast for a
small site, it doesn't break down, and it's admin friendly and stable
enough to start deploying to my test box. I've been showing it to a
friend of mine, and he's been impressed with how it works, how it's
organized, and the cleanliness of the HTML output (we both worked for
a "vertical portal company" which produced the most broken (both in
organization and in syntax) HTML I've ever seen).

There's nothing that I couldn't have done in another language, of
course, (all hail Turing) but Common Lisp has been extremely helpful,
from CLOS's multiple-inheritance and method combination, to the macro
facilities I was able to use to cleanly and easily generate the HTML,
to the condition system and extremely helpful "little things" such as
pathnames. How do I explain this? Most people seem to think that if
it can be done in something besides a Lisp, that it necessarily
_should_ be done in anything besides a Lisp.

My friend knows that I like Lisp, and from some Slashdot postings that
I wrote, complete with code snippets, he's come to understand why I
like it over, say, Java (he already despises C++). He also knows that
I've been writing this primarily so that I can easily create
Web-accessible resources using a language and libraries I enjoy. And
I'm purposefully setting things up so that it will be as easy to
install (or easier) than any other Apache plugin, and that no
knowledge of Lisp, or even that my software is written in Lisp, is
necessary. But it always comes back to the same thing: "Cool! Now
that you've got it fleshed out, how long will it take you to translate
it from the godly metalanguage Lisp [his words] into something more
normal?"

I can never really explain that the need for the concept expressed by
his question isn't really relevant to me, nor that I can't really see
the point in rewriting something that works fine, and that works as
robustly as it does because the environment I used made it acceptable
and easy to write more robust code than seems normal. Then the
conversation always moves to a tiresome discussion of things like
"marketing" and "mass acceptance" and stuff I can't bring myself to
care about.

--
vsync
http://quadium.net/
"If MS could only work this hard to make quality software. But they
never do. If they can't play dirty tricks they get bored and go out in
the corridor to play paint-ball."
-- http://yro.slashdot.org/comments.pl?sid=43916&cid=4576582

Chris Beggy

unread,
Nov 5, 2002, 1:32:06 PM11/5/02
to
vsync <vs...@quadium.net> writes:

> I've had the same problem. I'm currently working on a Lisp-based site
> and content management suite thing (can't really come up with a more
> concise definition than that ATM) using mod_lisp.

That's great!


Craig Brozefsky

unread,
Nov 5, 2002, 2:43:14 PM11/5/02
to
vsync <vs...@quadium.net> writes:

> necessary. But it always comes back to the same thing: "Cool! Now
> that you've got it fleshed out, how long will it take you to translate
> it from the godly metalanguage Lisp [his words] into something more
> normal?"

How about, "When its done, and that won't happen for another couple
years, but it would go faster if you helped me write the lisp."

> I can never really explain that the need for the concept expressed by
> his question isn't really relevant to me, nor that I can't really see
> the point in rewriting something that works fine, and that works as
> robustly as it does because the environment I used made it acceptable
> and easy to write more robust code than seems normal. Then the
> conversation always moves to a tiresome discussion of things like
> "marketing" and "mass acceptance" and stuff I can't bring myself to
> care about.

The site configuration and content mgmt market is so full of shite
that I think it's obvious that doing it over in a "more accpetbale"
language is in itself a step backwards, because for any acceptable
language there are an ungodly number of similarly small-time site mgmt
toolkits just like it. How would it compete for users and developers
in such a saturated market? By sticking with lisp you are aiming for
a different level of user and developer which can help your toolkit
advance with the state of the art. If you scrunch it into another
language you are likely to lose the flexibility and ease of
development and experementation that will differentiate your toolkit
from the bajillion others.

--
Sincerely,
Craig Brozefsky <cr...@red-bean.com>
Free Scheme/Lisp Software http://www.red-bean.com/~craig

Tim Bradshaw

unread,
Nov 5, 2002, 5:19:12 PM11/5/02
to
* Craig Brozefsky wrote:
> The site configuration and content mgmt market is so full of shite
> that I think it's obvious that doing it over in a "more accpetbale"
> language is in itself a step backwards, because for any acceptable
> language there are an ungodly number of similarly small-time site mgmt
> toolkits just like it. How would it compete for users and developers
> in such a saturated market? By sticking with lisp you are aiming for
> a different level of user and developer which can help your toolkit
> advance with the state of the art. If you scrunch it into another
> language you are likely to lose the flexibility and ease of
> development and experementation that will differentiate your toolkit
> from the bajillion others.

This is always the dilemma. In a related area, Sun face the same kind
of issues. The baying hordes are crying for them to convert to Linux
and x86, but if they do that, what do they have to offer? Well-made
and maintained machines? I dunno.

--tim

Heow

unread,
Nov 5, 2002, 7:20:34 PM11/5/02
to
There are more out here than I expected as I ran across a Lisper in a
random bar here in the East Village! Count me in, and I'll pester the
other guy too.

- Heow


cst...@dtpq.com (Christopher C. Stacy) wrote in message news:<uu1iwa...@dtpq.com>...

Paolo Amoroso

unread,
Nov 7, 2002, 4:39:24 AM11/7/02
to
On Sat, 02 Nov 2002 04:49:08 GMT, Kenny Tilton <kti...@nyc.rr.com> wrote:

> Dave Bakhash wrote:
[...]
> > Anther big speaker (for me) was Peter Norvig. ... I shared his feelings
> > that Common Lisp,
> > though having many advantages over other programming languages, was no
> > longer alone, and that many of the key features that CL had that were
> > not found in other languages have been adopted...
>
> That bit bothered me. PN listed 8 cool features of CL and said of those
> 6 had been replicated by /some/ language. But!! No /one/ new language
> has more than a few. Break it down by new language. All of a sudden we
> discover that, just to get to the 80% point of "replicating" CL you need
> to drag in four languages, each one contributing at most 30% of CL. (All
> preceding numbers fabricated but close enough for government work.)

Do you mean that we finally have a mathematical proof of Greenspun's Tenth
Rule?


> Wake-up call: Imran Shah (I believe it was) of U of Colorado, Boulder
> ended a nice talk by saying one of the only things wrong with CL was
> that you cannot find CLers. We did a sanity check and discovered he had
> advertised only locally. No posting on the Franz site, nothing here on
> c.l.l., I imagine nothing on monster or dice or hotjobs... Shah was

Let's not forget the ai+lisp-jobs mailing list.

Mark Dalgarno

unread,
Nov 7, 2002, 5:24:22 AM11/7/02
to
Paolo Amoroso <amo...@mclink.it> writes:

Just one thought on this. Five out of the six Lisp programmers I've
recruited in the last three years have come via postings I made on
c.l.l.

However, in a larger organisation recruitment may be handled by a
personnel department which does things in a rigid manner and may not
have the flexibility to advertise on usenet or the wider internet (or
to conduct the follow-up telephone interviews).

Moral - if you're in this position make sure c.l.l is one of the
recruitment channels your personnel department uses - it may also save
your organisation money in agency and advertising fees.

Mark

Dave Bakhash

unread,
Nov 7, 2002, 11:57:25 AM11/7/02
to
spam...@alphaGeeksInc.com (Heow) writes:

> There are more out here than I expected as I ran across a Lisper in a
> random bar here in the East Village! Count me in, and I'll pester the
> other guy too.

I'll add you to the list if you want...send me your actual email.

dave

Marc Battyani

unread,
Nov 7, 2002, 12:33:00 PM11/7/02
to

"Paolo Amoroso" <amo...@mclink.it> wrote

> On Sat, 02 Nov 2002 04:49:08 GMT, Kenny Tilton <kti...@nyc.rr.com> wrote:
>
> > Dave Bakhash wrote:
> [...]
> > > Anther big speaker (for me) was Peter Norvig. ... I shared his
feelings
> > > that Common Lisp,
> > > though having many advantages over other programming languages, was no
> > > longer alone, and that many of the key features that CL had that were
> > > not found in other languages have been adopted...
> >
> > That bit bothered me. PN listed 8 cool features of CL and said of those
> > 6 had been replicated by /some/ language. But!! No /one/ new language
> > has more than a few. Break it down by new language. All of a sudden we
> > discover that, just to get to the 80% point of "replicating" CL you need
> > to drag in four languages, each one contributing at most 30% of CL. (All
> > preceding numbers fabricated but close enough for government work.)
>
> Do you mean that we finally have a mathematical proof of Greenspun's Tenth
> Rule?

Almost. It's just a money problem. Basically what he said is that it costs
less to spend a lot of development time for really optimizing some parts of
the code when you have more than 10000 servers. A 20% decrease in
performance would result in 2000 servers, with their associated costs, being
added.
That bothered me as well because I don't have 10000 servers but only 3 and I
don't have the 200 talented C/C++ programmers of Google to write the macros
and compiler I get for free with a Common Lisp. So at the break I asked him
if he though his arguments were applicable to small companies or what
language he would use in that case. "No, I would use Lisp" he replied.

Marc


Wade Humeniuk

unread,
Nov 7, 2002, 3:42:49 PM11/7/02
to

"Marc Battyani" <Marc.B...@fractalconcept.com> wrote in message
news:279F7DE621B48C70.A0450637...@lp.airnews.net...

> Almost. It's just a money problem. Basically what he said is that it costs
> less to spend a lot of development time for really optimizing some parts of
> the code when you have more than 10000 servers. A 20% decrease in
> performance would result in 2000 servers, with their associated costs, being
> added.

That argument is not quite valid, as machine performance increases, in 18 months
those 10,000 servers may be down to 5,000. Now you have diminishing returns
on that 20%, until they almost diminish to nothing. Maintenence and administrative costs
are probably the largest cost after development, was anything said about that?

That argument has more sway for things like cell phones where many 100,000's of
phones are made and saving $1 per phone is a big deal.

> That bothered me as well because I don't have 10000 servers but only 3 and I
> don't have the 200 talented C/C++ programmers of Google to write the macros
> and compiler I get for free with a Common Lisp. So at the break I asked him
> if he though his arguments were applicable to small companies or what
> language he would use in that case. "No, I would use Lisp" he replied.


--
Wade

(format t "Email: ~A"
(map 'string
'code-char
'(119 104 117 109 101 110 105 117 64
116 101 108 117 115 46 110 101 116)))

Daniel Barlow

unread,
Nov 7, 2002, 7:12:29 PM11/7/02