pitman laid off by harlequin
Kent M Pitman
what prompted the layoff or how many people were affected, so don't
ask me to speculate. Harlequin will presumably make some sort of
announcement on this and I'm as interested to hear what they have to
say on this matter as I'm sure others are. But they HAVE assured me
they continue to be committed to Lisp. And, of course, so am I.
Anyway, this is now my official e-mail address (pit...@world.std.com).
Please don't send mail to me as k...@harlequin.com, since that won't
work any more (sigh).
I'll check out the lisp job boards, but if you happen to have some
lisp-related job that you don't want me to overlook, feel free to
contact me directly by e-mail. And meanwhile I finally have time
to work on that scifi novel I've been writing, and I hope to still
find time to keep posting here, etc.
David Steuber The Interloper
claimed or asked:
% I'll check out the lisp job boards, but if you happen to have some
% lisp-related job that you don't want me to overlook, feel free to
% contact me directly by e-mail. And meanwhile I finally have time
% to work on that scifi novel I've been writing, and I hope to still
% find time to keep posting here, etc.
I'm sure I speak for many when I say I hope you stick around.
Perhaps this is the way of the universe telling you to actually
graduate to professional writer as you have expressed the desire to
be. I wish you the best of luck.
I've always wanted to "write that sci-fi novel." But I have been
cursed with the ability to appreciate art, but not to create it.
Few people can earn a living by writing. But Kent Pitman is famous in
Lisp circles. I think that consulting could pay well. But then you
have all those posts you've made here, the stuff on your web site.
You have the material for several Lisp related books.
If you have some free cash, take a short sabbatical. Everyone needs a
break now and then.
I hear there is this really cool programming language called Java...
--
David Steuber (ver 1.31.2a)
http://www.david-steuber.com
To reply by e-mail, replace trashcan with david.
"Ignore reality there's nothing you can do about it..."
-- Natalie Imbruglia "Don't you think?"
john....@alcoa.com
tras...@david-steuber.com wrote:
> Few people can earn a living by writing. But Kent Pitman is famous in
> Lisp circles. I think that consulting could pay well. But then you
> have all those posts you've made here, the stuff on your web site.
> You have the material for several Lisp related books.
I suggest that Kent edit CLTL3. CLTL2 is too good a reference to let wither
from a few inaccuracies. Why not use the hyperspec you say? Well, I want a
book I can hold - something with some character and lot's of examples. The
hyperspec is too dry and difficult to browse. Besides, the chapter on
Conditions that Kent wrote in CLTL2 is one of the least up to date and he
owes it to us to fix it. I doubt Steele is interested in editing CLTL3 and
Kent is probably the best qualified guy with time on his hands.
How about it Kent?
John Watton
Technical Specialist
Aluminum Company of America
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Kent M Pitman
> I suggest that Kent edit CLTL3. CLTL2 is too good a reference to let wither
> from a few inaccuracies. Why not use the hyperspec you say? Well, I want a
> book I can hold - something with some character and lot's of examples. The
> hyperspec is too dry and difficult to browse.
CLTLn was a nightmare for Steele to maintain and he was happy to get
out from underneath it. I'm certainly not going to get back in that.
CLTL2 was never a document that was intended to describe any
implementation. It was not a committee document. It was an FYI-only
document that was just intended to inform the community about the
shape of things to come. Some vendors attached to it, but that was a
mistake and not one I would repeat.
I recommended to Steele at the time that he not do CLTL2, but for various
reasons he felt he had to. But I see no reason to return to any other
document than the standard document unless some vendor is going to go
back to implementing that instead, and I see no vendor planning to do that.
The whole reasons for standards is so that if you're a vendor and
tired of implementing the language, your customers can feel
comfortable going to someone else who implements the standard. Not
using a standard means convincing people to use an implementation they
can't "second source", which is bad. The key value of ANSI CL is not
that each and every technical decision is optimal, but rather that
each and every technical decision was agreed upon by the community to
be livable and something we would commonly support.
When you say "use CLTL2" vs "use the HyperSpec" you are not talking
about a document, you are talking about a language specification.
These documents specify different languages--neither language is "out
of date". They are simply different languages. People can conform to
one or the other. But since one was designed only by Steele in
isolation and without cooperation from any vendor or standards body
(CLTL2) and one was designed by the community and with cooperation
with vendors (ANSI CL), I know which one I'd recommend people use.
I bristle when people suggest that CLTL2 is a proper document for
ANSI CL. Yes, it's paper. But it describes different semantics than
the right one, and unless you're in on the history, each deviation is
a land mine. I just don't think the benefits outweigh the costs. I
spent a big part of my professional life producing the other document,
and I would abandon it in a moment if I thought it meant the industry
would rally around something else that was better. But CLTL2 is not
"better" in any quantifiable sense that I know of--it is not clearer,
it is not well-structured as a reference document, it is not the
document that people conform to. ANSI CL, though expensive to buy from
ANSI, is available in essentially equivalent form as the HyperSpec,
and is accessible, passably well presented (in particular, it has a
glossary of formal terms, the absence of which was a nightmare for
"language lawyers" trying to interpret even the original CLTL, much
less the weird "diff" format of CLTL2), and widely implemented.
And that's enough.
> Besides, the chapter on
> Conditions that Kent wrote in CLTL2 is one of the least up to date and he
> owes it to us to fix it. I doubt Steele is interested in editing CLTL3 and
> Kent is probably the best qualified guy with time on his hands.
Perhaps I'm just in a bad mood over this layoff, but this particular
remark really rubbed me the wrong way. I don't owe anyone anything.
I have worked on Lisp intensively since 1981. If you divided out the
number of hours I've spent on it by my salary, you would wonder why I
even bothered. But it's because I personally care about the language
and the people who use it. If the output of my 4 years of work as
editor of the HyperSpec left you feeling it was "a little dry", I
doubt you're going to say anything but the same if I write something
else--or, at least, I wouldn't risk it. Telling somebody you don't
like what they spent a big piece of their career doing as an incentive
to get them to do something else is not a good plan.
If I write something else, it will be something I choose to write
because I think it's of value and because I feel others will
appreciate it. I will not write anything out of any sense of
obligation. Now, more than at any other time, that is my absolute
right.
I didn't get a dime for writing that chapter of CLTL2, nor did I write
it for that purpose. I wrote it as part of the x3j13 design work,
Steele asked if he could include it, and I said "sure". The idea that
it is "out of date" is laughable to me because it was never NOT "out of
date". The very nature of CLTL2 as an unsynchronized, unauthorized,
unblessed snapshot of "work in progress" implies that everything that
was done was necessarily out of date as soon as it went to press while
the committee continued voting. Steele documented this in the 2nd
Edition Preface. I told him that wouldn't be enough, but he went
ahead anyway. So it goes.
There was some good done by CLTL2--it helped hold the community
together through a rough time. I'm not saying it was a
single-mindedly bad thing. And I don't mean to say Steele is bad for
having done it; I think it was a bad decision, but he had reasons and
some of those reasons were good ones. It's just that there was a
definite and high negative price for the positives, and that was the
divergence of the community for years to follow as well as the
NUMEROUS confusions created by the presentation style he chose. The
headache it caused for us internally to Lisp companies was truly
enormous--extending far beyond the mere choice of authors or
presentation style. But we are FINALLY past that, and I have no
desire to see us return to it. (I didn't ask him, but I'm 99% sure
that if you ask him he'll agree completely that a CLTL3 is not the
way to go.)
Ugh. Sorry about the strained tone here. It hasn't been the best of
weeks for me. But the points to be made here are important to make,
and if I don't make them I don't know for sure that someone else will.
Well, back to the job hunting thing...
Rainer Joswig
<pit...@world.std.com> wrote:
> Well, back to the job hunting thing...
Good luck!
David Thornley
David Steuber "The Interloper" <tras...@david-steuber.com> wrote:
>On Wed, 4 Nov 1998 07:00:58 GMT, Kent M Pitman <pit...@world.std.com>
>claimed or asked:
>
>% I'll check out the lisp job boards, but if you happen to have some
>% lisp-related job that you don't want me to overlook, feel free to
>% contact me directly by e-mail. And meanwhile I finally have time
>% to work on that scifi novel I've been writing, and I hope to still
>% find time to keep posting here, etc.
>
>I'm sure I speak for many when I say I hope you stick around.
>
as such). I'd like to thank Kent right now for the Hyperspec. I keep
a copy very close to my CL system (on the same hard disk, in fact),
and with that and Graham's Ansi Common Lisp (hey, it's paper and I'm
a bit old-fashioned) I feel that I have excellent reference material.
I have also enjoyed the articles on Kent's web site, and his postings
here.
>Perhaps this is the way of the universe telling you to actually
>graduate to professional writer as you have expressed the desire to
>be. I wish you the best of luck.
>
Me too. Good luck, Kent. Gee, have you considered writing a book
on Lisp? You could call it <thwack!>....oh well, if I see a science
fiction novel by you I'll buy it.
>I hear there is this really cool programming language called Java...
>
I heard that too, but when I looked for a really cool programming
language called Java, I didn't find one. What I found was
statically typed, procedural and object-oriented only, lacked
expandibility, and looked too much like C.
I'm sticking with Macintosh Common Lisp (and, forgive me, C++)
for my personal programming.
Hope to read more from you, Kent!
--
David H. Thornley | These opinions are mine. I
da...@thornley.net | do give them freely to those
http://www.thornley.net/~thornley/david/ | who run too slowly. O-
Patrick A. O'Donnell
Kent M Pitman <pit...@world.std.com> writes:
> john....@alcoa.com writes:
> > Besides, the chapter on
> > Conditions that Kent wrote in CLTL2 is one of the least up to date and he
> > owes it to us to fix it. I doubt Steele is interested in editing CLTL3 and
> > Kent is probably the best qualified guy with time on his hands.
>
> Perhaps I'm just in a bad mood over this layoff, but this particular
> remark really rubbed me the wrong way. I don't owe anyone anything.
I don't think it's just a bad mood. I thought that was over-the-top,
myself.
> If I write something else, it will be something I choose to write
> because I think it's of value and because I feel others will
> appreciate it.
Kent, I have appreciated your postings in this newsgroup, and have
learned a lot, or at least have had vague concepts clarified by
reading your messages. If you're taking requests, a collection of
your writings on Lisp (newsgroup postings, Parenthetically Speaking)
would be of great value to the community (in my opinion, of course).
("Thinking LISP", perhaps?)
But, by all means, work on your SF novel, and good luck to you!
- Pat
David Wild
<john....@alcoa.com> wrote:
> I suggest that Kent edit CLTL3. CLTL2 is too good a reference to let
> wither from a few inaccuracies. Why not use the hyperspec you say? Well,
> I want a book I can hold - something with some character and lot's of
> chapter on Conditions that Kent wrote in CLTL2 is one of the least up to
> date and he owes it to us to fix it. I doubt Steele is interested in
> editing CLTL3 and Kent is probably the best qualified guy with time on
> his hands.
If this happens I would like to ask Kent to include *lots* of worked
examples. If you are trying to learn on your own it is often difficult to
see what the effect of a small change is if you only have one example. Many
of the Lisp books fall down over this.
Good luck with the job search.
--
__ __ __ __ __ ___ _____________________________________________
|__||__)/ __/ \|\ ||_ | / Acorn Risc_PC
| || \\__/\__/| \||__ | /...Internet access for all Acorn RISC machines
___________________________/ dhw...@argonet.co.uk
Kent M Pitman
> For what it's worth, I concur that a CLtL3 is unnecessary.
Phew...
> Kent M Pitman <pit...@world.std.com> writes:
> > john....@alcoa.com writes:
> > > Besides, the chapter on
> > > Conditions that Kent wrote in CLTL2 is one of the least up to date and he
> > > owes it to us to fix it. I doubt Steele is interested in editing CLTL3 and
> > > Kent is probably the best qualified guy with time on his hands.
> >
> > Perhaps I'm just in a bad mood over this layoff, but this particular
> > remark really rubbed me the wrong way. I don't owe anyone anything.
>
> I don't think it's just a bad mood. I thought that was over-the-top,
> myself.
Well, as they say, it's a breakfast cereal AND a floor wax. My mood
has been generally quite upbeat, and maybe John's words weren't chosen
well, but I was feeling later like I probably should have sat on the
reply for longer and not been so snappy--so my apologies to John,
whether it's warranted or not. It's hardly something worth arguing
about--there are plenty enough other things for us to argue about
without second-guessing one another's wording. I managed to
accidentally upset someone fairly recently with some badly chosen
wording of my own so you'd think I'd know better...
Barry Margolin
>reasons he felt he had to. But I see no reason to return to any other
>document than the standard document unless some vendor is going to go
>back to implementing that instead, and I see no vendor planning to do that.
IMHO, the problem with CLtL2 was the timing. It was published before the
standard was finalized, so the language it describes is different from what
J13 agreed on.
If CLtL3 were written now, it would presumably be a pretty accurate
description of the language we agreed on. Such a book would be useful
because the style is much more accessible than the standard (although the
ANSI CL document is much more understandable than many other standards I've
read). And because it would be a mass market book, the hardcopy would be
less expensive than the ANSI document (I'm not sure what ANSI's price is,
but I expect it's more than twice what CLtL3 would cost).
It doesn't seem like it would take too much work to turn CLtL2 into CLtL3.
I don't think the language changed by more than about 5% after CLtL2 was
written. I don't have a printed copy of the standard, and when I want
something I can page through easily to find an answer I generally use
CLtL2, and it's rare that I get information that's contradicted by the
standard.
--
Barry Margolin, bar...@bbnplanet.com
GTE Internetworking, Powered by BBN, Burlington, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Don't bother cc'ing followups to me.
john....@alcoa.com
> When you say "use CLTL2" vs "use the HyperSpec" you are not talking
> about a document, you are talking about a language specification.
> I bristle when people suggest that CLTL2 is a proper document for
> ANSI CL.
I agree it's not a proper document for ANSI CL otherwise there would be no
need to edit it to conform to ANSI CL which was all I suggested. I don't want
another spec just a reference for the little guy. My own copy of CLTL2 is
edited by me to at least note nonANSI parts. Since the release of Graham's
book ANSI CL I use it as my primary hardcopy reference. But sometimes I miss
the examples and humor of Steele.
> > Besides, the chapter on
> > Conditions that Kent wrote in CLTL2 is one of the least up to date and he
> > owes it to us to fix it.
> Perhaps I'm just in a bad mood over this layoff, but this particular
> remark really rubbed me the wrong way. I don't owe anyone anything.
Sorry Kent, my remark was certainly not meant to sound mean spirited. Without
audio the jockular tone I intended didn't make it to print. I doubt one of
those silly smilely faces would have helped. Just let me retract the whole
message and return to my lurker status. Bye!
Martin Rodgers
> CLTLn was a nightmare for Steele to maintain and he was happy to get
> out from underneath it. I'm certainly not going to get back in that.
Phew! I'm not sure I have enough book space already, and there are still
a few existing Lisp books left for me to add to my "Lisp shelf". CLTL3
would be "a book too far" (WWII film reference).
No wonder I prefer technical documents in electronic form.
> Ugh. Sorry about the strained tone here. It hasn't been the best of
> weeks for me. But the points to be made here are important to make,
> and if I don't make them I don't know for sure that someone else will.
Fair enough.
> Well, back to the job hunting thing...
Good luck.
--
Remove insect from address to email me | You can never browse enough
will write code that writes code that writes code for food
j...@rebol.com
> I got laid off by Harlequin today. I don't have any information on
> what prompted the layoff or how many people were affected, so don't
> ask me to speculate. Harlequin will presumably make some sort of
> announcement on this and I'm as interested to hear what they have to
> say on this matter as I'm sure others are. But they HAVE assured me
> they continue to be committed to Lisp. And, of course, so am I.
I hope this isn't too mercenary, but my company is hiring. I'd like to
invite those affected by Harlequin's layoff to get in touch with me
if they are interested. We're not working on or with Lisp, but our product
has a lot of features that lisp people will find familiar (first class
functions and continuations, garbage collection, etc.)
~jrm
David Steuber The Interloper
I don't know for sure how you take requests. I'm not going to make
any. I just now had a thought that I would like to share with you.
The fact that you are a prolific writer on the subject of Lisp (and
the fact that you out right said it) suggests that you really enjoy
the language. The profession that would seem to touch the most people
would be teaching. You seem qualified. Would you enjoy teaching Lisp
at an undergraduate level? The right school should give you time to
write out of pure pleasure as well as publishing academic work.
I know nothing about you beyond your written work that I have read
(and enjoyed). So if this is an unappealing idea, just ignore it.
Kent M Pitman
> The profession that would seem to touch the most people
> would be teaching.
I replied to this in private mail.
It's not a terrible idea, but it's a darned shame teachers aren't paid
what they're worth.
Paolo Amoroso
wrote:
> and the people who use it. If the output of my 4 years of work as
> editor of the HyperSpec left you feeling it was "a little dry", I
> doubt you're going to say anything but the same if I write something
> else--or, at least, I wouldn't risk it. Telling somebody you don't
> like what they spent a big piece of their career doing as an incentive
> to get them to do something else is not a good plan.
From the documentation of Corman Lisp, a recently released Common Lisp
development environment for Windows by Roger Corman
(http://www.corman.net/CormanLisp.html):
"Common Lisp Hyperspec Support. The ANSI Common Lisp standard, along with
much useful accompanying documentation, has been made available in a
package called the Common Lisp Hyperspec. The is courtesy of Kent Pitman
and Harlequin. You may directly browse the entire Hyperspec from within
Corman Lisp, and in addition, all Corman Lisp symbols are linked to the
Hyperspec pages that define them."
I think that this is just the latest tribute to the value of the HyperSpec.
Paolo
--
Paolo Amoroso <amo...@mclink.it>
Tim Bradshaw
[Teaching Lisp]
> It's not a terrible idea, but it's a darned shame teachers aren't paid
> what they're worth.
Well said. In particular, anyone good enough to teach a really good
programming language course is good enough to get so much more money
outside academia that they won't become teachers, thus perpetuating a
vicious circle of poor language teaching.
(I was going to be (co-) teaching and redesigning the MSc Lisp course
here, but I got fed up being treated like dirt by academics -- I'm
just a system person -- and I've accepted a large pay increase in
industry instead. Which is a real shame because I really like
teaching (though not as much as I like being paid well, evidently!))
--tim
Erik Naggum
| And because it would be a mass market book, the hardcopy would be less
| expensive than the ANSI document (I'm not sure what ANSI's price is, but
| I expect it's more than twice what CLtL3 would cost).
FWIW, ANSI X3.226-1994 ships for USD 350.
amazon.com still ships CLTL2 for USD 50.
| It doesn't seem like it would take too much work to turn CLtL2 into
| CLtL3. I don't think the language changed by more than about 5% after
| CLtL2 was written. I don't have a printed copy of the standard, and when
| I want something I can page through easily to find an answer I generally
| use CLtL2, and it's rare that I get information that's contradicted by
| the standard.
uhm, this relates to what we were discussing recently. sometimes, the
differences are so subtle that you would miss them if you even _expect_
similarities. expect differences, and be happy about any similarities
you discover after you have compared them, not looking for them.
it may seem like I'm trying to split hairs, but I have been working with
standards and other specifications for more than a decade, and I also get
commissioned to write specifications -- my observation over these years
is that some people start to rely on what they observe to "work" in some
implementation or another, not through study, but rather through sloppy
acquisition of habit, and then get very upset when some unspecified or
undefined or implementation-defined behavior changes from implementation
to implementation, as it has every right and opportunity to do, and they
feel free to go ahead and _assume_ all kinds of things based on what they
only _think_ works, causing no end of frustrations even when upgrading
the same system and something undocumented was changed. stuff like that
is tremendously difficult to fix once it has slipped below consciousness.
e.g., the infamous Y2K problem is often merely a lack of adherence to
simple specifications. like, back in 1990, I took great pains to specify
that the Oslo Stock Exchange Trade Information Protocol's date format had
a window of 100 years that would be updated with a few months of notice,
like all other changes to it. the window initially covered 1950 through
2049, to be moved in the year 2000 to cover 1960 through 2069, and so
forth. only three of the implementers managed to read the specification
and observe the point that 00 through 49 were 2000 through 2049, and that
the application layer was specifically instructed to use 4-digit years
because the protocol used two-digit years for legacy reasons. too many
people had just assumed that two-digit years meant 20th century to deal
with dates in the year 2000 as early as 1997, so to make it less likely
that the same jerks would implement the next protocol with equal disdain
for specifications, the next revision used explicit four-digit years.
the cost of this change were acceptable. the cost of moving the sliding
window already specified were unacceptably high, even though it would be
less than 5% of the costs of dealing with the Y2K fever. go figure.
#:Erik
--
The Microsoft Dating Program -- where do you want to crash tonight?
Bill Coderre
Well, good luck on the next thing, whatever that may be.
I appreciate all you've done. There's almost no intelligent discourse left
on the Internet (I wish for the days of the old ARPAnet, where if you said
something off topic, you could be kicked off forever, and perhaps even get
kicked out of a job!), and you're one of the sources.
So thanks,
bc
Kent M Pitman
pitman> CLTLn was a nightmare for Steele to maintain and he was happy
pitman> to get out from underneath it. [...]
I forwarded my remarks about CLTL2/CLTL3 to Steele since I didn't
think he read this group regularly. He sent me the following text,
which he said I could forward to the newsgroup:
gls> [...] to whom it may concern: in my opinion, the ANSI standard for
gls> Common Lisp is now the *only* defining document for Common Lisp and
gls> therefore the only one worth using. I am grateful to those who say
gls> they like the style of CLTL and CLTL2, but those documents did have
gls> certain structural problems as defining documents, and in any case
gls> are out of date. CLTL2 was intended solely as a temporary bridge to
gls> the ANSI document, which I undertook only because it appeared to me
gls> that the ANSI process would take several more years (and it did).
gls> I think it would be more work than it would be worth to update CLTL3
gls> to match the standard---the standard itself is a fine defining and
gls> reference document, and furthermore if I, or anyone else, were to
gls> fail in some way to make them match, it would cause worse confusion
gls> than ever. Those people who *really* want the advantages of both
gls> are invited to cut pages 315-349 and 960-971 out of a copy of CLTL2
gls> and paste them onto the back of a copy of the ANSI standard.
gls>
gls> --Guy Steele
Steven Vere
writes:
>
>tras...@david-steuber.com (David Steuber "The Interloper") writes:
>
>> The profession that would seem to touch the most people
>> would be teaching.
>
>I replied to this in private mail.
>
>what they're worth.
This reminds me of the time I was laid off from Lockheed in 1992.
At one point I investigated a position with a small college in San
Francisco in response to an ad in the SF Chronicle. At the end of an
interesting phone conversation, their representative casually mentioned
that the position payed $25,000 per year.
Steven Vere
john....@alcoa.com
gls> [...] to whom it may concern: in my opinion, the ANSI standard for
gls> Common Lisp is now the *only* defining document for Common Lisp and
gls> therefore the only one worth using. I am grateful to those who say
gls> they like the style of CLTL and CLTL2, but those documents did have
gls> certain structural problems as defining documents, and in any case
gls> are out of date. CLTL2 was intended solely as a temporary bridge to
gls> the ANSI document, which I undertook only because it appeared to me
gls> that the ANSI process would take several more years (and it did).
gls> I think it would be more work than it would be worth to update CLTL3
gls> to match the standard---the standard itself is a fine defining and
gls> reference document, and furthermore if I, or anyone else, were to
gls> fail in some way to make them match, it would cause worse confusion
gls> than ever. Those people who *really* want the advantages of both
gls> are invited to cut pages 315-349 and 960-971 out of a copy of CLTL2
gls> and paste them onto the back of a copy of the ANSI standard.
gls>
gls> --Guy Steele
So let me summarize. If you want a comprehensive hardcopy reference to
the ANSI Common Lisp spec you have two options:
1) The complete ANSI Common Lisp standard - cost $350
2) Paul Graham's ANSI Common Lisp which has a complete 91 page
language reference appendix - cost $42
I have choosen option 2 because I am not yet willing to buy or in need
of the $350 reference. (I trust Steele that the standard is a fine
defining and reference document. I have never seen it and probably
never will - never mind performing cut and paste surgery on it as he
flippantly suggests.) I continue to feel that there is something
missing in between these two options. I naively suggested an edit of
CLTL2. Many folks are familiar with it and I continue to believe that
it could be made ANSI compliant (but I never suggested language
implementors latch on to it as the standard). I realize that a formal
CLTL3 will never happen. Lisp is just not a big enough market
anymore. My estimate is that there are only about 2000 active Common
Lisp programmers in the USA today. In true scientific fashion I
extrapolate this from only one data point. My company (ALCOA) has
100,000 employees and 2 Common Lisp programmers (my partner Mike and
I). There are about 100,000,000 employees in the USA and therefore the
estimate of 2000.
Now, I have a proposal for the fraction of you who are not in the burn
all copies of CLTL2 camp. Let's edit it ourselfs. Graham's book ANSI
Common Lisp is a great place to start. He has a 1.5 page summary of
the changes that need to be made (pages 308-309). They are organized
in 14 items. It takes about an hour to use it as a guide to updating
CLTL2. Many of the changes can be made in the margins although a
little more space is required to add the definitions for the 6 new
operators added: allocate-instance, array-displacement, constantly,
define-symbol-macro, ensure-directories-exist, lambda, read-sequence,
and write-sequence. To update the Conditions chapter change all
occurances of simple-condition-format-string to
simple-condition-format-control. And you all thought this was going to
be hard! Don't forget to add new operators to the index. I don't want
to repeat all the changes here but I do volunteer to take sometime
this month to use Graham's book to put together a recipe in the style
of an errata sheet. It will not make CLTL2 100% compliant but it
should be a definite improvement.
John Watton
Technical Specialist
Aluminum Company of America
-----------== Posted via Deja News, The Discussion Network ==----------
Martin Cracauer
>So let me summarize. If you want a comprehensive hardcopy reference to
>the ANSI Common Lisp spec you have two options:
>1) The complete ANSI Common Lisp standard - cost $350
>2) Paul Graham's ANSI Common Lisp which has a complete 91 page
>language reference appendix - cost $42
That may or not be of relevance, but the verbatim ANSI C standard was
allowed to be printed in a normal-priced book.
"The annotated ANSI C standard" is a printed copy of the ANSI document
with some annotations not worth the paper they are printed on (cleanly
seperated from the ANSI pages). You get the real thing for a
reasonable price and it seems the ANSI folks did agree. Probably
because they hoped to get more money out of an often-sold book and
since Lisp is used at many universities, maybe the same approach would
work for Lisp, too.
Martin
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Martin Cracauer. For email address see a Web searcher.
http://www.freebsd.org/ - where you want to go. Today.
Lyman S. Taylor
In article <72c4rb$k35$1...@nnrp1.dejanews.com>, <john....@alcoa.com> wrote:
....
>gls> the ANSI document, which I undertook only because it appeared to me
>gls> that the ANSI process would take several more years (and it did).
>So let me summarize. If you want a comprehensive hardcopy reference to
>the ANSI Common Lisp spec you have two options:
>1) The complete ANSI Common Lisp standard - cost $350
>2) Paul Graham's ANSI Common Lisp which has a complete 91 page
>language reference appendix - cost $42
I think your summary is flawed.
First Graham book is NOT a substitute for the ANSI standard.
It is not intended to be and the appendix doesn't cover lots of
details (that is why it is so short). This not a bad thing.
They address two different audiences. Graham is a tutorial on
a major portion of the language. The standard is a definnition of
the language. Those require disparate approaches.
The better analogy that comes to my mind is from the C++ world.
CLtL is roughly analogous the C++ Annotated Reference Manual (ARM).
The ARM is obsolete. So is CLtL. They were both placeholders until
the standard finished gestating.
Note that the ARM was not the same book as Stroustrup's book about the
language ( tC++PL ) . While Graham's book is not quite as encyclopedic
as tC++PL, neither is a replacement for the corresponding standard.
>flippantly suggests.) I continue to feel that there is something
>missing in between these two options.
I'm not sure which book you want. If you want a substitute for the
CL ANSI standard and aren't commited to have a "dead tree" in front
of you, then you can use the HyperSpec to fill in the details that
Graham doesn't cover. I'm not sure how easy it would be to turn
the HyperSpec into a "dead tree". I think it looses much of its
unique utility as a "dead tree" though.
Between Graham's two books I'm not sure if there are very many
Common Lisp constructs not covered in one way or another (perhaps CLOS).
I don't think tC++PL is the only C++ book you'll every need to refer
to either.
As your "back of the envelope" ( which isn't very "scienfitic", IMHO.
an engineer's "seat of the pants" approach which, with enlightened
intuition, can suffice) calculations imply I don't think a 1,500+
page, encyclopedic, tutorial book would have a large enough run for the
publishers to competitively price it.
>CLTL2. Many folks are familiar with it and I continue to believe that
>it could be made ANSI compliant (but I never suggested language
In IMHO you have to gut CLtL to redo it. In the context of not being
a stopgap transitory document, the "delta" text doesn't make sense.
I also don't see Stroustrup scrambling to update the ARM. "It is dead
Jim".
>Now, I have a proposal for the fraction of you who are not in the burn
>all copies of CLTL2 camp.
I'd don't think anyone needs to "burn" CLtL2. I have a copy at work and
at home as a backup when I can't get to the HyperSpec. Or I'm looking
at things from a archaeological perspective (why'd they do it this way).
Or I need yet another example.
[ I tend to work with a number of folks (students) who prefer to work
with "free" CL implementation. So sometimes I use CLtL2 to
explain why their implementation is "broken" or incorrect.]
Again looking at C++ for an analogy, the C++ standard is $175.00 paper
but only $18.00 in PDF form. Perhaps the committe could conjole ANSI
into doing similar price reduction for Common Lisp (which is the
same price electronic and paper .... sure that's econmically justified.
priced to sell, right).
>Let's edit it ourselfs. Graham's book ANSI
edit and distribute in what format? Electronic? The HyperSpec is
already electronic.
I'm not even going to delve into the copyrights of producing a
derivative work. I don't think CLtL2 is GPL'ed.
>this month to use Graham's book to put together a recipe in the style
>of an errata sheet. It will not make CLTL2 100% compliant but it
>should be a definite improvement.
In some sense CLtL2 is already an errata sheet. It seems to me that
making an errata sheet to an errata sheet means that perhaps it
may be time to move on to another reference. If you like
band-aid hodge podges I suggest taking a gander at C++. ;-)
I don't see significant difference between Steele's "flippant" suggestion
to paste some pages from CLtL2 onto the back of the standard and
adding even more stuff to the "margins" of CLtL3. If your attached
to your dog-earred copy of Steele, cool. But it is not worth another
edition. If you concerned about the standards then you need to
look at them. CLtL2 is "mostly" correct if you don't need to know
the strict letter of the laws then Steele can suffice.
If it is Steele's examples you like then I think a better project would
be to extract ( if doesn't have adverse implications) the examples and
link them back into the HyperSpec. Updating the examples as necessary.
This avoids modifying anyone's documents.
This leaves the formal definitions as defined in the HypeSpec (which I'm
sure Kent has made jive with the standard) so that tons of work need not
be done making sure that the two texts say the same thing. Natural Language
isn't exactly the optimal tool for specification.
--
Lyman S. Taylor "Twinkie Cream; food of the Gods"
(ly...@cc.gatech.edu) Jarod, "The Pretender"
Kent M Pitman
john....@alcoa.com writes:
> So let me summarize. If you want a comprehensive hardcopy reference to
> the ANSI Common Lisp spec you have two options:
> 1) The complete ANSI Common Lisp standard - cost $350
> 2) Paul Graham's ANSI Common Lisp which has a complete 91 page
> language reference appendix - cost $42
Well, Paul Graham's text is not a "reference" and there are probably others
that treat ANSI CL as well--he just happens to have chosen that name for
commercial reasons. He isn't the creator of ANSI CL--he's just someone with
a good book on ANSI CL that happens to be called by the same name.
But your options are, more properly put, to buy the real standard or to buy
a commercial teaching text. In general, MANY other programming language
definitions are not even readable--some are not even human-readable. They
are not intended as the text you should buy.
I have commonly made an analogy to laws about driving. I spent an
afternoon at the library once when a friend was ticketed for "running
a red light" and read the Massachusetts General Law Annotated (MGLA)
on the subject. It is NOT a pretty piece of work. MGLA is very long
and complicated and full of special cases evolved by case law. It is
NOT the thing you learn in driving class and though I like to think
the ANSI CL spec is better written, the two share the sense that there
is supposed to be a teaching text that really is your reference. Also,
driving lessons are almost comic book form, are very small, and yet
drivers are held to all the rules of the document they never see, so
the situation is quite analogous. My understanding has always been
that ANSI intends its customers to be people buying the language or
with a strong need for a highly reliable reference--like those writing
teaching texts. Beyond that, I think they assume the free market will
produce books like K&R for C, which are smaller and more approachable
than the formal specifications such as the ANSI C specification, which
I'd be surprised if you have ever seen.
> I have choosen option 2 because I am not yet willing to buy or in need
> of the $350 reference.
NO ONE -- certainly not me -- is recommending you buy this unless you
have a special commercial need for it. It is not intended an end user
item. You probably assumed because you asked about the price and
someone told you that this was the recommended solution.
However, the Common Lisp HyperSpec is a fine alternative to buying the
$350 version since it IS a derivative work of the spec an it
more-or-less faithfully uses the ACTUAL text of the standard--modulo
some typography problems for missing characters in HTML and a few
"glue" sections for the hypertext not needed in the hardcopy and of
course the everpresent possibility that the TeX->HTML processor I
wrote to convert it had some bug. If you want the 100% right version,
you must get it from ANSI. If you want something that is essentially the
same, get CLHS.
> (I trust Steele that the standard is a fine
> defining and reference document. I have never seen it and probably
> never will
Except for the typography, see
http://www.harlequin.com/education/books/HyperSpec/FrontMatter/index.html
Download instructions are at
http://www.harlequin.com/education/books/HyperSpec/
>- never mind performing cut and paste surgery on it as he
> flippantly suggests.)
I doubt anyone will do that.
> I continue to feel that there is something
> missing in between these two options. I naively suggested an edit of
> CLTL2.
This document is owned by Steele. It is his personal work. It is not
community owned. If he doesn't want to make another version, that's
all there is to be said. You're welcome to suggest your interest, of
course, but he knows there is interest. I think he's made it clear
that it's *he* who doesn't have the interest, and he has some sound
technical reasons for not trying.
> Many folks are familiar with it and I continue to believe that
> it could be made ANSI compliant (but I never suggested language
> implementors latch on to it as the standard).
(As ambassador Kosh on Babylon 5 would so coldly say: Irrelevant.)
Other books may indeed happen. But they will have other names and be
written from scratch exactly not to infringe Steele's copyright.
> I realize that a formal
> CLTL3 will never happen. Lisp is just not a big enough market
> anymore. My estimate is that there are only about 2000 active Common
> Lisp programmers in the USA today. In true scientific fashion I
> extrapolate this from only one data point. My company (ALCOA) has
> 100,000 employees and 2 Common Lisp programmers (my partner Mike and
> I). There are about 100,000,000 employees in the USA and therefore the
> estimate of 2000.
>
> Now, I have a proposal for the fraction of you who are not in the burn
> all copies of CLTL2 camp. Let's edit it ourselfs.
You are proposing something that is will get you sued. See my web page
http://world.std.com/~pitman/law.html
if you want some pointers to references on copyright law. Statutory damages
for properly registered copyrighted works (independent of any show of damages)
can run as high as $100,000 per infringement if I remember right. You must
start from scratch or obtain authorization from the copyright owner or you're
setting yourself up for trouble. I know we all like to talk about "Lisp"
like it's all "ours" but individual pieces of people's investments are owned,
and that's yet another reason for not using CLTL2. The fact that even the
original CLTL was copyrighted turned into a source of problem a while back
and was one of many contributing reasons for going to another forum and format
to avoid the legal complications caused by that in trying to extend it, etc.
> Graham's book ANSI
> Common Lisp is a great place to start. He has a 1.5 page summary of
> the changes that need to be made (pages 308-309). They are organized
> in 14 items. It takes about an hour to use it as a guide to updating
> CLTL2.
Graham's work is also copyrighted. You are risking doubling your liability
with this approach.
> Many of the changes can be made in the margins [...] And you all
> thought this was going to be hard!
I don't know who said it was hard. Getting sued really doesn't
have to take a lot of work.
Barry Margolin
>> I continue to feel that there is something
>> missing in between these two options. I naively suggested an edit of
>> CLTL2.
>
>This document is owned by Steele. It is his personal work.
It may be his personal work (with contributions by another dozen, according
to the front matter), but the copyright notice says Digital Equipment
Corporation. It seems pretty common for authors to assign copyright to the
publisher (all the technical books on my shelf seem to do it, but "The
Extended Phenotype" in my backpack is copyrighted by the author).
Kent M Pitman
> >[CLTL2] is owned by Steele. It is his personal work.
>
> It may be his personal work (with contributions by another dozen, according
> to the front matter), but the copyright notice says Digital Equipment
> Corporation. It seems pretty common for authors to assign copyright to the
> publisher (all the technical books on my shelf seem to do it, but "The
> Extended Phenotype" in my backpack is copyrighted by the author).
You're right it's Digital's. I dont know the nature of his agreement
with them, though. My real point had been that it was "owned" and not
GPL'd ... it's not community property. permissions are required.
I further hope we see new books, not retreads of old ones. New points
of view help. New code examples help. So do new jokes.
john....@alcoa.com
ly...@cc.gatech.edu (Lyman S. Taylor) wrote:
> First Graham book is NOT a substitute for the ANSI standard.
Never said it was. I used his words - "Language Reference" The first sentence
of which if I can still quote without being sued - "This appendix describes
every operator in ANSI Common Lisp."
> As your "back of the envelope" ( which isn't very "scienfitic", IMHO.
> an engineer's "seat of the pants" approach which, with enlightened
> intuition, can suffice) calculations imply I don't think a 1,500+
> page, encyclopedic, tutorial book would have a large enough run for the
> publishers to competitively price it.
My point I believe. Doesn't anyone have a sense of humor anymore?
> >Let's edit it ourselfs. Graham's book ANSI
> edit and distribute in what format? Electronic? The HyperSpec is
> already electronic.
> I'm not even going to delve into the copyrights of producing a
> derivative work. I don't think CLtL2 is GPL'ed.
Not a derivative work in my mind. Just an errata style sheet: Example:
ANSI deleted "enclose": Cross out bottom 5 lines on page 207. Cross out bottom
10 lines page 214. Delete index entries on pages 991, 1020.
Let the lawsuits begin!
john....@alcoa.com
> Well, Paul Graham's text is not a "reference" and there are probably others
> that treat ANSI CL as well--he just happens to have chosen that name for
> commercial reasons. He isn't the creator of ANSI CL--he's just someone with
> a good book on ANSI CL that happens to be called by the same name.
His appendix D is titled - "Language Reference". As for other writings on
ANSI CL I'd appreciate a reference. The only other Lisp book I have bought
recently is Slade's Object Oriented Common Lisp published in 1997. He
pretends ANSI never happened. In fact he refers to CLTL2 as the "bible" and
"blueprint." I'd provide a full quote but fear being sued. Actually I'm at
home and the book's at work.
> However, the Common Lisp HyperSpec is a fine alternative
I agree. I just have a nostalgia for books.
> > Graham's book ANSI
> > Common Lisp is a great place to start. He has a 1.5 page summary of
> > the changes that need to be made (pages 308-309). They are organized
> > in 14 items. It takes about an hour to use it as a guide to updating
> > CLTL2.
Sued for using a book as a reference? In my errata sheet I wasn't even
planning on using any direct quotes of Graham's book. My instructions were
going to be of the form:
On page 1010 of CLTL2 change index entry simple-condition-format-string to
simple-condition-format-control. Did I learn that from Graham or from the
appendix of X3J13 votes in the Franz ACL4.3 user manual? Maybe they'll both
sue!
Lyman S. Taylor
In article <72er4f$sqc$1...@nnrp1.dejanews.com>, <john....@alcoa.com> wrote:
>In article <72cuvj$f...@pravda.cc.gatech.edu>,
> ly...@cc.gatech.edu (Lyman S. Taylor) wrote:
>Never said it was
What???? Here's your orignal text which you choose to delete:
----------------------------------------------------------------------------
>So let me summarize. If you want a comprehensive hardcopy reference to
>the ANSI Common Lisp spec you have two options:
-------------------------------------------------------------------------
The critical part being "want a comprehensive hardcoypy reference to"
If there are only "two" alternatives what is that indicative of?
"ANSI Common Lisp" is not a substitute for CLtL2, either. The
"purpose" of CLtL2 is to function as a comprehensive reference manual.
To bring it up to date you have to make it equivalent to the standard.
Graham's book is not in this classification. Or we violently
disagree on what the word comprehensive means.
>of which if I can still quote without being sued - "This appendix describes
>every operator in ANSI Common Lisp."
As I said in the quoted message, this decription is incomplete in
some cases. A list of the names of all the operators isn't a
reference. Or at least a reference of any usefulness. For example
there are a couple of destructive operators that Appendix D fails
to mention are destructive. If you need your code to have
referential transparency or want to avoid consing that is a crucial
point to leave out.
As a "pocket guide" reference, yeah it passes. However, if I truely
needed CLtL2 to answer a question then the modern reference would be
the standard (or HyperSpec).
>> >Let's edit it ourselfs.
.....
>Not a derivative work in my mind. Just an errata style sheet: Example:
"edit it" to me means "edit the CLtL2". You make no explicit mention
of creating a seperate document. If you meant to
create a seperate errata document, then you should not say things like
CLtL3. There would be *no* CLtL3. There would be a CLtL2 with lots of
scribbling in it. "Let's create an errata document for CLtL2" would be
much more specific.
>ANSI deleted "enclose": Cross out bottom 5 lines on page 207. Cross out bottom
>10 lines page 214. Delete index entries on pages 991, 1020.
If there are changes in semantics that require non trival
proofreader marks and/or substantial textual additions to correct this
can quicly become extremely unwieldy. As I stated before CLtL2 is
already an errata of CLtL1. So layering even more corrections on top
becomes dubious if extensive.
If this were solely an exercise in deletions this might be a tractable
task. It is the "new" material that is extremely likely to be
problematical.
Secondly, this would probably have to be a peer reviewed (or some
sort of comittee based) exercise. In some sense each "editor" would
have to translate the ANSI standard into "Steele". If you don't
use some peer review process and folks who know the ANSI standard
"backwards and forwards" you can very well introduce mistakes or
omissions.
Perhaps there is some utility in a marked up CLtL2 that is "less" in
error than the raw version. I'd perfer just to use the
"raw" version with full knowledge that it is out of date and that
if I need the nitty gritty details I need to go refer to the "real
deal". That way I don't get a false sense of security.
Lyman S. Taylor
Lyman S. Taylor <ly...@cc.gatech.edu> wrote:
...
> .... For example
> there are a couple of destructive operators that Appendix D fails
> to mention are destructive.
Someone pointed out the notation that is suppose to illustrate this
in the appendix ( angle bracket around the argument means it can be
side effected). I missed that in Appendix's introduction. Perhaps a better
example would be the absence of what errors are invoked when.
In appendix D, when he lists an argument as being a proper list, it would
probably be prudent to only pass proper lists to those operators. However,
there are speical cases in the standard where perhaps the argument
doesn't have to a proper list. Or it only matters if it isn't a proper
list in certain contexts and the saftety it turned on to the highest
limits.
For a reference directed at "users", a comprehensive listing of all of
the "special cases" aren't critical. However, it is difficult to
believe that with different formatting and deletion of examples and
supportive material that a 1000+ page reference can be condensed down
to less than 100 pages without omission of something of substance.
I have trouble believing the standard is full of that much "fluff".
You can make an arugment that substance doesn't matter to most
people though. However, that would a document that is a satisfacory
substitute for, but not equivalent to the standard.
john....@alcoa.com
> >Never said it was
>
> What???? Here's your orignal text which you choose to delete:
> ----------------------------------------------------------------------------
> >the ANSI Common Lisp spec you have two options:
After posting the original message I expected quibbles with the word
"comprehensive" but I stand by my claim that I never claimed Graham's book as
a substitute for the standard. In my mind we have the ANSI standard (God the
Father, beyond our view on His throne in heaven, if you will) and the
Hyperspec (God the Holy Spirit, the noncorporal breath of the Father). What
we don't have is a value priced incarnation of the ANSI standard (God the
Son, the perfect image of the Father).
> "edit it" to me means "edit the CLtL2". You make no explicit mention
> of creating a seperate document. If you meant to
> create a seperate errata document, then you should not say things like
> CLtL3. There would be *no* CLtL3. There would be a CLtL2 with lots of
> scribbling in it. "Let's create an errata document for CLtL2" would be
> much more specific.
Okay, let's create an errata document for CLtL2. Sorry for the
misunderstanding. By "edit it" I meant an individual with his pencil, his copy
of CLTL2, and the errata sheet would edit (mark up) his copy of CLTL2.
> Secondly, this would probably have to be a peer reviewed (or some
> sort of comittee based) exercise. In some sense each "editor" would
> have to translate the ANSI standard into "Steele". If you don't
> use some peer review process and folks who know the ANSI standard
> "backwards and forwards" you can very well introduce mistakes or
> omissions.
The errata sheet would be posted periodically like a FAQ. I think my peers
would quickly point out mistakes, ommisions, etc. Before long the errata sheet
would be well worked over.
John Watton
my-las...@mediaone.net
I don't have any requisitions at the moment, but my company is building a
lisp&java based webcentric system for collaborative development and deployment
of web content. It has some unique change-management capabilities (apropos the
version control topic which was also floating around in this group). All UI
access (which isn't deliberately command line access) is done through the web
browser.
I expect I might have some reqs soon (Q1), so if any of the affected Harlequin
folks might be interested in some pretty head's down product development in the
area of webcentric change management, you might find my project interesting.
If someone is really interested, it wouldn't be the first time I've
"manufactured" reqs where none existed before. It took me 9 months of lobbying
to get this project off the ground, this is the first quarter it's been
officially funded, so we're just getting started.
Email me at:
(concatenate 'string "dtenny"
"@" ; anti-spam device...
"truesoft.com")
if you're interested, or at the .sig address.
D. Tenny
my-las...@mediaone.net - no spam please
Paul Wallich
Vere) wrote:
>In <sfwk918...@world.std.com> Kent M Pitman <pit...@world.std.com>
>writes:
>>
>>tras...@david-steuber.com (David Steuber "The Interloper") writes:
>>
>>> The profession that would seem to touch the most people
>>> would be teaching.
>>
>>I replied to this in private mail.
>>
>>It's not a terrible idea, but it's a darned shame teachers aren't paid
>>what they're worth.
>
> This reminds me of the time I was laid off from Lockheed in 1992.
(I remember that. It was right after they vociferously denied reports
that they were cutting back their AI funding...)
>At one point I investigated a position with a small college in San
>Francisco in response to an ad in the SF Chronicle. At the end of an
>interesting phone conversation, their representative casually mentioned
>that the position payed $25,000 per year.
That's good money for entry-level teaching, alas...
You could ask retired lisp programmers to donate their time,
but I doubt there are any...
paul
Kent M Pitman
people and the little statuette, you perhaps see a parallel to this
thread.
john....@alcoa.com
The statuette is CLTL2 and you want to snap it in two, in anger? And then
someone has to go away for two years?
As for me, I am not devoted to CLTL2. It's too thick for me. I prefer thinner
books like Graham's ANSI Common Lisp. I have two copies each with a spiral
binding so that it will lay flat. Will you at least agree that it's appendix
makes a handy pocket "reference?"
Reginald S. Perry
You know, if you post your address here, I'm sure we can round up a
few people to hang out outside your house with Kent statuettes. :-)
-Reggie
Kent M Pitman
Well, I was thinking Steele, actually. With CLTL2 being held up as the
essential book of wisdom and all that.
Christopher B. Browne
posted:
I heard it the way you said it.
One difference is that I don't think there's anyone who's going to try to
assassinate Guy Steele because he won't return to the Narn and write
CLTL3...
Definitely some good B5 episodes as it winds down to the end. Garibaldi has
returned to being eminently entertaining, which is a very good thing.
--
Those who do not understand Unix are condemned to reinvent it, poorly.
-- Henry Spencer <http://www.hex.net/~cbbrowne/lsf.html>
cbbr...@hex.net - "What have you contributed to Linux today?..."