Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

help! absolute beginner

111 views
Skip to first unread message

Zachary Turner

unread,
Dec 1, 1998, 3:00:00 AM12/1/98
to
i have never programmed lisp before in my life, so i have no idea what's
really going on here, but i plan on learning it over the next 6 months or
so. what i need is a LISP compiler so that i can get started. I'm looking
into Harlequin LispWorks 4.1 Professional. I've downloaded the Personal
Edition (which is available free for download at their website) but i have
no idea how to get a simple program up and running. Can someone tell me
exactly what i need to do and exactly what code to type just to write an
extremely simple program (the LISP equivalent of the typical C "Hello World"
perhaps?) that prints something on the screen, just so that i can see some
results and know that I'm actually on the right track? I have no idea how
to use LispWorks, and no idea about the language itself, so if someone could
just hold my hand and treat me like a complete moron i'd really appreciate
it. :-) Thanks for any help.

Zachary Turner

Zachary Turner

unread,
Dec 1, 1998, 3:00:00 AM12/1/98
to
by the way, i don't need the code explained to me, i'll learn all of that
soon enough, i just want to see something happen before I go shell out money
on this product. Thanks

Zachary Turner

Zachary Turner wrote in message <741sh7$lje$1...@uuneo.neosoft.com>...

Sunil Mishra

unread,
Dec 1, 1998, 3:00:00 AM12/1/98
to
Have a look at the Association of Lisp Users web site:

http://www.elwoodcorp.com/alu/

You will find a fair bit of information (tutorials as well) to get you up
to speed. Good luck!

Sunil

Toolshed37

unread,
Dec 2, 1998, 3:00:00 AM12/2/98
to
well, i took a brief look at it, and not much help. i dont' really want a
tutorial right now, i just want someone to give me enough code that i can just
cut and paste into the window of LispWorks and then tell me what to do to
execute the program so that i can see some type of output on the screen. i'm
not going to start trying to learn lisp until later, right now i just want to
see something on my screen. if you could just post an _entire_ hello world
program and tell me what to do in LispWorks to get it to run, i'd really
appreciate it.

David B. Lamkins

unread,
Dec 2, 1998, 3:00:00 AM12/2/98
to
In article <19981201224608...@ng146.aol.com> ,
tools...@aol.com (Toolshed37) wrote:

Well, define what you mean by a hello world program... Here are several
that meet your requirements.

Because strings are self-evaluating in Lisp, this is the the simplest
program:
"Hello, world!"

If you want to get closer to the canonical hello world program that does
explicit output, try these:
(write "Hello, world!")
(format t "Hello, world!")
(format nil "Hello, world!")

If you want to pop up a window that says hello world, that kind of output is
platform dependent (and I don't have a copy of LW at hand). Typically, you
can create a window that behaves as a text output stream, and use it
something like this:
(let ((w (make-instance 'window)))
(format w "Hello, world!"))
Again, the window example may not work in LW (if it does, it's a lucky guess
on the class name). Why don't you take a look through the manual and see
whether they have a tutorial?

Finally, if you want to create a standalone application, you're really gonna
have to crack that manual. Every Lisp is different, but it would probably
be fruitful if you searched the index for some hyphenated combination of
dump or save and image or application.

---
David B. Lamkins <http://www.teleport.com/~dlamkins/>

There are many ways to abbreviate something, but only one way not to.

Sunil Mishra

unread,
Dec 2, 1998, 3:00:00 AM12/2/98
to
"Hello World" is a rather futile exercise.

(princ "Hello World")
|Hello World|
"Hello World"

Any of these expressions at the command prompt will do what you want to,
and yet none of these is satisfying as a program.

If you want to write a function that simply does hello world, then

(defun foo ()
(princ "Hello World"))

(foo)

This to me does not quite seem very satisfying as a program either.

I guess what I am trying to say is that Lisp is not very amenable to small
toy programs, simply because they are trivial to put together. Now, if you
want to deal with a large complex problem, Lisp tends to work great. There
are many, many problems that take loads of code in other languages, but end
relatively clean and nice in Lisp. An expressive syntax helps :-)

Going up yet another level, the "right way" to learn a language would be to
figure out what it is good for. Obviously, Lisp is not good for writing
Hello World, since you end up carrying around an environment at least 10MB
in size to get this to work. I don't know what your motives for learning
lisp are, but I would think hard about this question.

Now, if you want to ask what lisp is good for, I would suggest trekking
over to dejanews. This question has been asked enough times. If you have a
problem at hand, there are lots of *very* smart people on this newsgroup
that would be more than happy to help you out. (They've been kind to me,
certainly.)

Sunil

Zachary Turner

unread,
Dec 2, 1998, 3:00:00 AM12/2/98
to
Sunil Mishra wrote in message ...

>Going up yet another level, the "right way" to learn a language would be to
>figure out what it is good for. Obviously, Lisp is not good for writing
>Hello World, since you end up carrying around an environment at least 10MB
>in size to get this to work. I don't know what your motives for learning
>lisp are, but I would think hard about this question.
>
>Now, if you want to ask what lisp is good for, I would suggest trekking
>over to dejanews. This question has been asked enough times. If you have a
>problem at hand, there are lots of *very* smart people on this newsgroup
>that would be more than happy to help you out. (They've been kind to me,
>certainly.)
Oh, I have very valid motives for learning LISP. I'm a university student
and in
the Spring I'm doing an independent study course on Artificial Intelligence,
for which I've decided to learn LISP, since it is perfectly suited for doing
AI
programming. I also realize that it is not explicitly for AI, but that it
does
suit the problem perfectly. I was able to get something to run with the
help
of all the posts, thanks to everyone who posted. I'll probably go shell out
some $$ for LispWorks now, I like the fact that you can write in LISP and
deliver it as a .DLL, which you can link to in C++ :-)

Zachary Turner

Joachim Achtzehnter

unread,
Dec 2, 1998, 3:00:00 AM12/2/98
to
"Zachary Turner" <ztu...@elsitech.com> writes:
>
> Oh, I have very valid motives for learning LISP. I'm a university
> student... I'll probably go shell out some $$ for LispWorks now...

If you just want to learn Lisp then why are you so keen on spending a
large sum of money? Last time I looked, both Harlequin and Franz had
full-featured versions of their Lisp environments free for personal
use. And there are a whole bunch of Lisp implementations that are free
software, in case you ever wanted to look at the inside :-)

Joachim

--
joa...@kraut.bc.ca (http://www.kraut.bc.ca)
joa...@mercury.bc.ca (http://www.mercury.bc.ca)

Zachary Turner

unread,
Dec 2, 1998, 3:00:00 AM12/2/98
to
I'm spoiled I guess? I've been writing games in C/C++ using DirectX and
OpenGL
for some time now and it'd be nice if, after I get good with Lisp, I could
write all
my AI code in LISP, compile it to a DLL, and dynamically link to the DLL in
my
games, which will be written in C/C++.

Joachim Achtzehnter wrote in message ...

Sunil Mishra

unread,
Dec 3, 1998, 3:00:00 AM12/3/98
to
I suspect doing things the other way around - writing your games in lisp,
and using C/C++ routines for specific functions - will work out better. But
then that's just a guess. I'm only familiar with Harlequin Lispworks, and I
know they have an interface for automatically reading header files and
constructing an interface to C. So, as long as you have the header files,
constructing an OpenGL interface *ought* to be straightforward. (I haven't
done this, so I can't be certain.)

Sunil

Zachary Turner

unread,
Dec 3, 1998, 3:00:00 AM12/3/98
to
I'm not sure how practical that would be. All graphics, sound, input, and
network programming I typically do in C++ so it wouldn't make much sense to
write the whole program in Lisp and link in all code relating to graphics,
sound, input, and network programming when I could write it in C++ and link
in only the AI stuff. One thing that's just come to my attention though:
How do you prototype a Lisp function in C? From what I've learned of Lisp
so far (not very much at all) it isn't typed at all. The list (1 2 3 (foo
bar 37)) is quite different in memory size from the list (1). So I guess
the two main things that you would need to know how to declare in C are an
atom and a list, taking into consideration that a list can contain other
lists. Has anyone had any experience doing this?


Sunil Mishra wrote in message ...

Zachary Turner

unread,
Dec 3, 1998, 3:00:00 AM12/3/98
to
On second thought maybe this isn't such a bad idea.. Interesting at the
least... I'll give it a shot in a few months after I get proficient with
Lisp. On another note, does anyone know what the deal is with Franz, Inc.?
I've been trying to get in touch with these people for the last few weeks
and they don't return emails and when I call they say that the lady who
handles sales "is not currently in the office and wont' be in for the
remainder of the day." That's happened twice now and the email thing has
happened twice also. I evaluated Allegro CL 5.0 and I really like it, but
these people are basically sitting here with the "No we don't your money.
Go give it to Harelquin" attitude, which is really frustrating since I want
to purchase Allegro. Has anyone else had better luck with them?

Zach

Vassili Bykov

unread,
Dec 3, 1998, 3:00:00 AM12/3/98
to
For the kind of things you are describing, you might really like Corman Lisp
(http://www.corman.net/CormanLisp.html).

Zachary Turner wrote in message <746dv0$qij$1...@uuneo.neosoft.com>...


>On second thought maybe this isn't such a bad idea.. Interesting at the
>least... I'll give it a shot in a few months after I get proficient with
>Lisp. On another note, does anyone know what the deal is with Franz, Inc.?

[...]


Tim Bradshaw

unread,
Dec 3, 1998, 3:00:00 AM12/3/98
to
* Zachary Turner wrote:
> On second thought maybe this isn't such a bad idea.. Interesting at the
> least... I'll give it a shot in a few months after I get proficient with
> Lisp. On another note, does anyone know what the deal is with Franz, Inc.?
> I've been trying to get in touch with these people for the last few weeks
> and they don't return emails and when I call they say that the lady who
> handles sales "is not currently in the office and wont' be in for the
> remainder of the day."

They've probably been very busy organising a conference (very well, I
thought).

--tim

john....@alcoa.com

unread,
Dec 3, 1998, 3:00:00 AM12/3/98
to
In article <u1z92.2402$5n1.19...@news.magma.ca>,

"Vassili Bykov" <vas...@objectpeople.com> wrote:
> For the kind of things you are describing, you might really like Corman Lisp
> (http://www.corman.net/CormanLisp.html).

Corman Lisp is betaware. I recommend that if you have the money for real
commercial grade software that you stick with Franz or Harlequin. They have
been battle tested, implement the ANSI standards, and you will experience far
fewer fustrations with them. My experience with Corman Lisp involved frequent
crashes, missing features (like macrolet), and far slower performance. And
all I tried on Corman Lisp were a few small benchmarks. I hate to be negative
but you can't really expect a new Lisp environment to be competitive with
folks with a large staff and years behind their product? For a beginner on
Windows I recommend Harlequin Lispworks Personal Edition and when you get a
real project graduate to the Professional Edition (~$500) which costs only a
few hundred more than what Corman wants. If you need even more performance
(and have even more dollars) then consider Franz's Allegro CL(~$3000-4000).
(I have both and prefer ACL over Lispworks but I am biased by having used ACL
on Unix for years and the need for the absolute fastest floating point
speed.)

--
John Watton
Aluminum Company of America

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own

John Atwood

unread,
Dec 3, 1998, 3:00:00 AM12/3/98
to
if you're into games, ya might also be interested in this scheme product
that has an integrated solid modeler:
http://www.schemers.com/3ds20.html


John Atwood
-------------------------------------------

Erik Naggum

unread,
Dec 5, 1998, 3:00:00 AM12/5/98
to
* "Zachary Turner" <ztu...@elsitech.com>

| On another note, does anyone know what the deal is with Franz, Inc.?

well, yes, I do. it appears that you walked into a small restaurant with
Burger King expectations. in so doing, you might have triggered some
strong negative responses based solely in a cultural conflict that you
might not even be aware of. phone numbers and e-mail addresses look very
much alike, in contrast to storefronts, offices, and counters.

you said you know nothing about (Common) Lisp -- one of the things you
will experience is that the community is very different from the Windows
communities, whence it appears you come. for instance, the Common Lisp
market is not marketing-driven, it is not a pyramid game that requires
ever new people nor a bug-and-upgrade scam, and it is not leveraging its
operational costs across a huge volume of sales. rather, it is a pretty
mature market of long-term partnerships with a steady growth. the quick
sale is not unlike a one-night-stand in this setting and you _may_ just
have appeared much less than serious than you believed you were.

| I evaluated Allegro CL 5.0 and I really like it, but these people are
| basically sitting here with the "No we don't your money. Go give it to
| Harelquin" attitude, which is really frustrating since I want to purchase
| Allegro. Has anyone else had better luck with them?

yes, I have. the fact that I have happy Common Lisp clients today is
probably due mostly to the excellent and welcoming attitude at Franz Inc
when I first approached them. they have continued to be very helpful in
making my projects succeed, both for me and for my clients. I think what
you write is grossly unfair, so I have to reiterate my impression that
you have stumbled on a cultural conflict; not all restaurants serve fast
food, some cater to a very different audience and their tastes and needs.

#:Erik
--
The Microsoft Dating Program -- where do you want to crash tonight?

Vassili Bykov

unread,
Dec 5, 1998, 3:00:00 AM12/5/98
to
In article <31218169...@naggum.no>,
Erik Naggum <er...@naggum.no> wrote:
> * "Zachary Turner" <ztu...@elsitech.com>

> | I evaluated Allegro CL 5.0 and I really like it, but these people are
> | basically sitting here with the "No we don't your money. Go give it to
> | Harelquin" attitude, which is really frustrating since I want to purchase
> | Allegro. Has anyone else had better luck with them?
>
> yes, I have. the fact that I have happy Common Lisp clients today is
> probably due mostly to the excellent and welcoming attitude at Franz Inc
> when I first approached them. they have continued to be very helpful in
> making my projects succeed, both for me and for my clients.

I can add that I have had a totally different (from Zachary's) experience in
communicating with Franz, in spite of never even having been their paying
customer. Virtually every e-mail inquiry sent to the sales (out of
curiosity, to keep an eye on the situation in case I do get a chance to
become a paying customer), as well as ordering of ACL 4.3 for Linux, was
followed up by a phone call from them. I have no idea what was the secret of
"looking serious", but I have always had an impression they indeed have a
very enthusiastic and welcoming attitude.

--Vassili

--
Vassili Bykov
The Object People
http://www.objectpeople.com

Vassili Bykov

unread,
Dec 6, 1998, 3:00:00 AM12/6/98
to
In article <7470f8$l5m$1...@nnrp1.dejanews.com>,
john....@alcoa.com wrote:

> Corman Lisp is betaware. I recommend that if you have the money for
> real commercial grade software that you stick with Franz or
> Harlequin. They have been battle tested, implement the ANSI standards,
> and you will experience far fewer fustrations with them. My experience
> with Corman Lisp involved frequent crashes, missing features (like
> macrolet), and far slower performance.

Very true. But I do believe Corman Lisp has its place in the sun,
together with all the big commercial and non-commercial players,
because it scratches some itches other implementations do not.

> I hate to be negative but you can't
> really expect a new Lisp environment to be competitive with folks with
> a large staff and years behind their product?

Perhaps because there does not have to be a competition. You
obviously speak from the viewpoint of a corporate developer who has
the money and expects return of investment. This is great, but what
about the hobbyists? The "Personal" and "Lite" editions of LWW and
ACLW are very different from Corman Lisp, and not always to their
advantage. First thing, they are crippled, second, you don't have the
source, third, you are not allowed to redistribute them as parts of
your application.

> For a beginner on
> Windows I recommend Harlequin Lispworks Personal Edition and when you
> get a real project graduate to the Professional Edition (~$500) which
> costs only a few hundred more than what Corman wants.

What is better, open-source betaware or black box crippleware? Before
answering, it is important to consider what for.

For a beginner, LWW Personal or ACLW Lite are indeed the best choice.

But, there is not necessarily a graduation with getting a real Lisp
project. One may have Lisp-unrelated projects in real life they are
perfectly happy about, or unable to change for the time being. So
what do you use if you want something that is not crippled, is still
free or very low cost, allows you to let others see and use your work
as an application, and allows you to see what is inside so that you
can add what is missing or fix what is broken? None of the free
versions of the big commercial players fit the description.

As for "what Corman wants", you seem to have misread the license
conditions. Roger does not want anything for the compiler and the
runtime system. Corman Lisp is free, and that is uncrippled version
with the right to non-commercially distribute binary applications
which may include the compiler. None of the free versions of the
commercial stuff in Windows come close to this. Add to that full
source code, and there is some weight to compare.

Now, on the technical side, Corman Lisp has a very interesting balance
between low-level and high-level stuff. You can do everything a C
programmer can. This was the reason I said the original poster may be
interested in Corman Lisp, since there was some talk about polygon
pushing and game programming. The FFI in Corman Lisp is very flexible
and transparent. The assembly-level stuff is fully open, to the point
that one can write inline assembly code in Lisp functions. It is,
essentially, Lisp with all the capabilities of C when it comes to
communicating with the outside world.

So, my take on this (besides that I don't think Lisp community
should (or, even, can afford to) kick the underdog) is yes,
Corman Lisp is worse compared to the entrenched big commercial
players, but it just might happen to be the kind of worse that is
better here and there...

--Vassili

P.S. BTW, Roger Corman will soon release an update with improved
thread stability (those crashes are due to Windows thread
implementation "features"), macrolet, and the full source--Lisp and C,
among some other things.

P.P.S. No, I am not affiliated with Roger.

P.^3S. If you are interested in joining an online community of Corman
Lisp users, you are welcome to <http://www.dejanews.com/~cormanlisp>.

Rainer Joswig

unread,
Dec 6, 1998, 3:00:00 AM12/6/98
to
In article <74d0hq$iti$1...@nnrp1.dejanews.com>, Vassili Bykov
<vas...@objectpeople.com> wrote:

>
> As for "what Corman wants", you seem to have misread the license
> conditions. Roger does not want anything for the compiler and the
> runtime system. Corman Lisp is free, and that is uncrippled version
> with the right to non-commercially distribute binary applications
> which may include the compiler. None of the free versions of the
> commercial stuff in Windows come close to this. Add to that full
> source code, and there is some weight to compare.

"Free" nowadays often seems to be:

- you are free to fix the bugs you never wanted to fix.
- you are free to implement the functionality you
never wanted to program yourself
- you are free to wade through large amounts of complicated
source code you never wanted to look at

Some people seem to really underestimate the massive amount of work
that is needed to write a really usable and balanced system. This
week I checked a small piece of code (two pages) in five Lisp systems.
a) one *free* versions I checked was not able to run it
(a class fixnum?) without modifications
b) for the two free systems I gave up waiting for the result
(P266 and SUN E250). My old MacIvory beats them. Ha!

So, people don't get your expectations to high and never underestimate
the work that is needed to develop a Lisp system. I said it some
time ago, I have ***high*** respect for guys like Roger Corman
who are trying to tame the beast.

--
http://www.lavielle.com/~joswig

Pierre Mai

unread,
Dec 6, 1998, 3:00:00 AM12/6/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> Some people seem to really underestimate the massive amount of work
> that is needed to write a really usable and balanced system. This
> week I checked a small piece of code (two pages) in five Lisp systems.
> a) one *free* versions I checked was not able to run it
> (a class fixnum?) without modifications

While the ANSI Standard does not mandate that the type specifier
FIXNUM have a corresponding class (unlike INTEGER), it does IMHO
allow implementations to define classes for other type specifiers
(Section 4.3.7):

"Individual implementations may be extended to define other type
specifiers to have a corresponding class."

CMU CL does define a class for FIXNUM.

> b) for the two free systems I gave up waiting for the result
> (P266 and SUN E250). My old MacIvory beats them. Ha!

If you could give me access to the source code, I´d be very interested
in looking into performance issues in CMU CL on Intel, so that the
quality of implementation increases for one of the free systems out
there ;)

> So, people don't get your expectations to high and never underestimate
> the work that is needed to develop a Lisp system. I said it some

Regs, Pierre.

--
Pierre Mai <pm...@acm.org> http://home.pages.de/~trillian/
"One smaller motivation which, in part, stems from altruism is Microsoft-
bashing." [Microsoft memo, see http://www.opensource.org/halloween1.html]

john....@alcoa.com

unread,
Dec 6, 1998, 3:00:00 AM12/6/98
to
In article <74d0hq$iti$1...@nnrp1.dejanews.com>,
Vassili Bykov <vas...@objectpeople.com> wrote:
>
> > As for "what Corman wants", you seem to have misread the license
> conditions. Roger does not want anything for the compiler and the
> runtime system. Corman Lisp is free

He wants $300 for the development environment if memory serves me. The rest is
"free".

I want to correct something else I said earlier. Allegro CL Professional is
something like $4000 which gives you the ability to distribute runtime
internal to a corporation (external to corporation Franz wants to negotiate a
cut). There is an Allegro Personal which is in the $500-900 range but no
runtime ability which I had not mentioned.

--
John Watton
Aluminum Company of America

-----------== Posted via Deja News, The Discussion Network ==----------

Rainer Joswig

unread,
Dec 6, 1998, 3:00:00 AM12/6/98
to
In article <87soetm...@orion.dent.isdn.cs.tu-berlin.de>, Pierre Mai
<pm...@acm.org> wrote:

> > b) for the two free systems I gave up waiting for the result
> > (P266 and SUN E250). My old MacIvory beats them. Ha!
>

> If you could give me access to the source code, I=B4d be very interested


> in looking into performance issues in CMU CL on Intel, so that the
> quality of implementation increases for one of the free systems out
> there ;)

Nothing special. Just using generic functions instead of
ordinary functions. Calling methods seems to have a huge overhead
(also conses like hell) in CMU CL. The version without methods
ran as expected. I tried this test only
once on my MacIvory and actually methods were faster (!) - on other
implementations there was a 15% speed penalty using
methods.

--
http://www.lavielle.com/~joswig

Rainer Joswig

unread,
Dec 6, 1998, 3:00:00 AM12/6/98
to

> While the ANSI Standard does not mandate that the type specifier
> FIXNUM have a corresponding class (unlike INTEGER), it does IMHO
> allow implementations to define classes for other type specifiers
> (Section 4.3.7):
>
> "Individual implementations may be extended to define other type
> specifiers to have a corresponding class."
>
> CMU CL does define a class for FIXNUM.

You are right.

--
http://www.lavielle.com/~joswig

Rainer Joswig

unread,
Dec 6, 1998, 3:00:00 AM12/6/98
to
In article <74eois$u59$1...@nnrp1.dejanews.com>, john....@alcoa.com wrote:

> In article <74d0hq$iti$1...@nnrp1.dejanews.com>,
> Vassili Bykov <vas...@objectpeople.com> wrote:
> >
> > > As for "what Corman wants", you seem to have misread the license
> > conditions. Roger does not want anything for the compiler and the
> > runtime system. Corman Lisp is free
>
> He wants $300 for the development environment if memory serves me. The rest is
> "free".
>
> I want to correct something else I said earlier. Allegro CL Professional is
> something like $4000 which gives you the ability to distribute runtime
> internal to a corporation (external to corporation Franz wants to negotiate a
> cut). There is an Allegro Personal which is in the $500-900 range but no
> runtime ability which I had not mentioned.

So it seems that LispWorks for Windows is much cheaper?

--
http://www.lavielle.com/~joswig

Georg Bauer

unread,
Dec 6, 1998, 3:00:00 AM12/6/98
to
In article <joswig-0612...@194.163.195.67>, jos...@lavielle.com
(Rainer Joswig) wrote:

Exactly. With commercial applications, this greatly differs:

>- you are free to fix the bugs you never wanted to fix.

You are not free to fix the bugs you never wanted to fix. In fact, often
you don't have the sources to fix it.

>- you are free to implement the functionality you
> never wanted to program yourself

You might be possible to add the functionality you need but the company
never want's to implement. In fact, as soon as you want to fix things in
the lowest levels of the product, you are most often lost. It is better
with Lisp, since Lisp is much better customizeable than C++-Compilers, but
the problem still exists.

>- you are free to wade through large amounts of complicated
> source code you never wanted to look at

Yes, that is definitely much better with commercial products where you
don't get source at all. You might get _some_ source, as with most Lisps,
but that is not a quite common case for commercial software.

No, I definitely prefer high-quality Open Source [TM] software. Makes
living easier, especial in my field of work. Of course, there is much crap
out there. But actually with free software, you don't have to pay for
crap, like is the case with commercial ones.

bye, Georg

Vassili Bykov

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
In article <74eois$u59$1...@nnrp1.dejanews.com>, john....@alcoa.com wrote:
> In article <74d0hq$iti$1...@nnrp1.dejanews.com>,
> Vassili Bykov <vas...@objectpeople.com> wrote:
> >
> > > As for "what Corman wants", you seem to have misread the license
> > conditions. Roger does not want anything for the compiler and the
> > runtime system. Corman Lisp is free
>
> He wants $300 for the development environment if memory serves me. The rest is
> "free".

Right now it's $100. This buys the IDE and the right to sell the applications
you develop.

I don't see the need to quote "free" talking about the rest, the rest has
enough value by itself :-). It's not free in the sense of FSF, it is free in
the sense you don't have to pay. (And in the more humourous senses suggested
by Rainer Joswig). And quite frankly, the IDE is not that great, at least for
now, compared to simple Emacs, so taking it out does not turn the rest into
crippleware.

I never wanted to imply Corman Lisp would give Franz or Harlequin a run for
their money. But, they are players of totally different leagues, and this
works both ways. ACLW Lite and LWW Personal do not have some of the virtues
of Corman Lisp. Those of the hobbyists experienced enough to not be afraid
of rough spots and looking for easy access to foreign and low-level stuff,
can have some great time with Corman Lisp. It all boils down to where you
interests are.

--Vassili

--
Vassili Bykov
The Object People
http://www.objectpeople.com

-----------== Posted via Deja News, The Discussion Network ==----------

Rainer Joswig

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
In article <gb-061298...@hugo.westfalen.de>, g...@hugo.westfalen.de
(Georg Bauer) wrote:

> In article <joswig-0612...@194.163.195.67>, jos...@lavielle.com
> (Rainer Joswig) wrote:
>
> Exactly. With commercial applications, this greatly differs:
>
> >- you are free to fix the bugs you never wanted to fix.
>
> You are not free to fix the bugs you never wanted to fix. In fact, often
> you don't have the sources to fix it.

Often source doesn't help you or access to source
isn't even the problem at all.

Will access to source magically give a Lisp system
a usable GC? Even getting a decent
UI seems not to be possible. Does "Open Source"
and Lisp mean: we will get nowhere?

This stuff is really complicated. It is **naive** to expect
that this will show up soon. What's out there is
light-years away from even the ancient
Xerox InterLisp-D systems.

> >- you are free to wade through large amounts of complicated
> > source code you never wanted to look at
>
> Yes, that is definitely much better with commercial products where you
> don't get source at all. You might get _some_ source,

So you lately have looked into your Genera file system, right?
Looked for the sources of Genera?
Maybe you even have installed the additional source folder
of MCL on your Mac. Did you need it? Much? Would you
have needed it?

> as with most Lisps,
> but that is not a quite common case for commercial software.

I better ask the experts.

> No, I definitely prefer high-quality Open Source [TM]

Now it is a trademark? Religious marketing bullshit.

> software.

--
http://www.lavielle.com/~joswig

Craig Brozefsky

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> Often source doesn't help you or access to source
> isn't even the problem at all.
>
> Will access to source magically give a Lisp system
> a usable GC?

Certainly not, and I don't think that anyone is making that claim. It
has to be written wether the author puts it into public domain or not.
People are motivated by many different things, and not all of them may
be congruent to your motivations and goals, or mine.

I think the previous poster has different criteria for making his
decision about what system to use than you do, and both sides need to
be aware of that. For some a liberal license for distribution is more
important than GC performance, and for others performance and features
are worth whatever price is asked. Luckily, both sides can be
satisfied! It's all good man, it's all lisp.

> > No, I definitely prefer high-quality Open Source [TM]
>
> Now it is a trademark? Religious marketing bullshit.

Yes. Bruce Perens and Eric Raymond have started an organization to do
Open Source license certification, and part of their plan is to
trademark the term that has been in public use since way before they
applied for the trademark, and "protect it" from misuse. It's
actually ideologically free and makes extra effort to espouse the
proper party lines about pragmatism and realism that makes corporate
bottom-feeders salivate. So it's Non-Religious marketing bullshit,
but marketing bullshit none the less. Be sure not to confuse them
with the people who actually writing code.

Tim Bradshaw

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
* Rainer Joswig wrote:

> So you lately have looked into your Genera file system, right?
> Looked for the sources of Genera?

Yes, I do that a lot. I've even fixed bugs in it (not the file
system, the networking stuff). I'm very glad they gave source out
(albeit it wasn't open source), I wish more people did, as it can
really save a bunch of time.

--tim


Rainer Joswig

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
In article <ey3pv9w...@todday.aiai.ed.ac.uk>, Tim Bradshaw
<t...@aiai.ed.ac.uk> wrote:

> * Rainer Joswig wrote:
>
> > So you lately have looked into your Genera file system, right?
> > Looked for the sources of Genera?
>
> Yes, I do that a lot. I've even fixed bugs in it (not the file
> system, the networking stuff). I'm very glad they gave source out
> (albeit it wasn't open source),

Some of this stuff was by accident, AFAIK. ;-) Like the CLIM source. ;-)

> I wish more people did, as it can
> really save a bunch of time.

Open Genera should be shipped with even more source code, as I heard.

--
http://www.lavielle.com/~joswig

Zachary Turner

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to

> you said you know nothing about (Common) Lisp -- one of the things you
> will experience is that the community is very different from the Windows
> communities, whence it appears you come. for instance, the Common Lisp
> market is not marketing-driven, it is not a pyramid game that requires
> ever new people nor a bug-and-upgrade scam, and it is not leveraging its
> operational costs across a huge volume of sales. rather, it is a pretty
> mature market of long-term partnerships with a steady growth. the quick
> sale is not unlike a one-night-stand in this setting and you _may_ just
> have appeared much less than serious than you believed you were.
>
>| I evaluated Allegro CL 5.0 and I really like it, but these people are
>| basically sitting here with the "No we don't your money. Go give it to
>| Harelquin" attitude, which is really frustrating since I want to purchase
>| Allegro. Has anyone else had better luck with them?
>
> yes, I have. the fact that I have happy Common Lisp clients today is
> probably due mostly to the excellent and welcoming attitude at Franz Inc
> when I first approached them. they have continued to be very helpful in
> making my projects succeed, both for me and for my clients. I think what
> you write is grossly unfair, so I have to reiterate my impression that
> you have stumbled on a cultural conflict; not all restaurants serve fast
> food, some cater to a very different audience and their tastes and needs.

How is it unfair? Was there ever a time when you knew *nothing* about
Common Lisp? I'm inclined to believe that there was. Just call it a hunch.
Judging by your other posts to this NG it seems like you want to come across
as the allmighty Lisp programmer who knows all, but you were just a baby
wearing a diaper sucking out of a bottle at one point too. So does that
mean that you deserved to be treated differently than someone who had been
programming in Lisp longer than you? You'll probably say yes for argument's
sake but the _correct_ answer is no. The bottom line is that you have no
idea how serious I am about all my motives behind purchasing a commercial
Lisp implementation. I could be purchasing it so that I can do Lisp on my
spare time, or I could be trying to be the next allmighty Erik Naggum. So
please, don't presume to make claims about the validity of my complaints,
because they were very valid complaints. However, none of this matters
anymore because the issue has since been resolved. I have been contacted by
two different people from Franz, Inc. both of whom have been extremely
helpful. And the bottom line still remains that if they weren't selling
licenses they wouldn't be in business. That's generally how it works.
Whether I want to buy 1 license or 100 licenses doesn't change the situation
any. I'm still trying to give them money.

Zach

Erik Naggum

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
* g...@hugo.westfalen.de (Georg Bauer)

| Exactly. With commercial applications, this greatly differs:
|
| >- you are free to fix the bugs you never wanted to fix.
|
| You are not free to fix the bugs you never wanted to fix. In fact, often
| you don't have the sources to fix it.

I have fixed some thirty-odd problems in Allegro CL with advise before I
got source code, and once I got a load of source with the commercial
support agreement, I fixed yet more problems. it appears that my fixes
are migrating into Franz Inc's own code base. however, it often takes a
whole day to write a good bug report, and only an hour or so to tweak
some behavior in the source. for some reason, navigating in the source
code that came with Allegro CL had only one obstacle: the IF* macro. the
rest was a breeze. :)

| >- you are free to implement the functionality you
| > never wanted to program yourself
|
| You might be possible to add the functionality you need but the company
| never want's to implement. In fact, as soon as you want to fix things in
| the lowest levels of the product, you are most often lost. It is better
| with Lisp, since Lisp is much better customizeable than C++-Compilers,
| but the problem still exists.

well, _my_ Allegro CL 5.0 now does

(format nil "~,,' ,4:B" (get-universal-time))
=> "1011 1010 0001 0110 1000 0111 1110 1001"

because I needed this behavior and didn't feel like writing my own FORMAT
substitute or function to hack numbers and suck. the changes were
actually fairly simple.

| >- you are free to wade through large amounts of complicated
| > source code you never wanted to look at
|
| Yes, that is definitely much better with commercial products where you

| don't get source at all. You might get _some_ source, as with most Lisps,


| but that is not a quite common case for commercial software.

all Allegro CL licensees receive a lot of source code with their (signed)
support agreement. not the internals, of course, but the functions you
would normally call or need to understand.

| No, I definitely prefer high-quality Open Source [TM] software. Makes
| living easier, especial in my field of work. Of course, there is much

| crap out there. But actually with free software, you don't have to pay


| for crap, like is the case with commercial ones.

I don't pay for crappy software to begin with. why do other people? if
you are implying that commercial Common Lisp implementations are crap, I
think you have an attitude problem so big you should be dismissed as a
lunatic, like the other "Free Software" fanatic who posts occasionally to
this newsgroup.

Erik Naggum

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
* "Zachary Turner" <ztu...@elsitech.com>

| Judging by your other posts to this NG it seems like you want to come
| across as the allmighty Lisp programmer who knows all, but you were just
| a baby wearing a diaper sucking out of a bottle at one point too.

newsgroups are not the proper place to discuss your personal problems,
and especially not projecting them onto others in pathetic desperation.

| So does that mean that you deserved to be treated differently than
| someone who had been programming in Lisp longer than you?

I do my homework and work hard to be competent in what I do. that
generally _does_ have a positive effect on other competent people,
actually no matter what they are competent in. I assume you wouldn't
know what I'm talking about.

| You'll probably say yes for argument's sake but the _correct_ answer is
| no.

OK, I won't confuse your mind with facts. it would probably hurt a lot.

| The bottom line is that you have no idea how serious I am about all my
| motives behind purchasing a commercial Lisp implementation.

I never claimed to have. I'm sorry that you have this personal problem
that you have to take out on me, but if you could please go back and read
what I wrote and stop imputing all sorts of insanities to me just because
_you_ would think that way, maybe you'll figure out something important.

| So please, don't presume to make claims about the validity of my
| complaints, because they were very valid complaints.

sure. in the context of your personally experience, but not universally.
since you seem to be the kind of guy who makes universal claims out of
your personal experience, I assume you won't grasp the difference nor the
fact that your fucking stupid "allmighty" crap should tell everybody that
they should stay _far_ away from you.

| I have been contacted by two different people from Franz, Inc. both of
| whom have been extremely helpful. And the bottom line still remains that
| if they weren't selling licenses they wouldn't be in business. That's
| generally how it works. Whether I want to buy 1 license or 100 licenses
| doesn't change the situation any. I'm still trying to give them money.

I wouldn't sell anything to you. nor do I think anybody else should.

Zachary Turner

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
You don't have many friends do you?

rusty craine

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to

Zachary Turner wrote in message <74h5vb$osl$1...@uuneo.neosoft.com>...

>You don't have many friends do you?
>
>
I like the guy. He takes no prisoners and ask for no quarter. There are
several guys in the NG (hmmm no women?) that I try to learn all the lisp I
can from. He's one of them. If ya piss him off just duck your head, wait
till it's over and keep on learning [I'm still a few posting from know it
all :)].

If ya decided to get in a verbable joust with him wear your flack jacket.
He seems to come prepared.

Rusty

Vassili Bykov

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
Erik Naggum wrote in message <31220406...@naggum.no>...

> I don't pay for crappy software to begin with. why do other people? if
> you are implying that commercial Common Lisp implementations are crap, I
> think you have an attitude problem so big you should be dismissed as a
> lunatic, like the other "Free Software" fanatic who posts occasionally to
> this newsgroup.


Hmm, who would that be?


Zachary Turner

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
I dont' want to get in a verbal joust with anyone, and perhaps my rant was
partially unjustified, but I take offense to being made feel like i don't
belong in the Lisp community simply because I am a beginner. If that's not
what he was saying, then I apologize for going off. Otherwise, there's no
reason for me to sit here and have someone tell me that i'm not worthy of
programming in Lisp. In any case, that's how it came across.

Zach

rusty craine wrote in message <74h7lf$p8k$1...@excalibur.flash.net>...

Vassili Bykov

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
Zachary Turner wrote in message <74h8s6$ph4$1...@uuneo.neosoft.com>...

>I dont' want to get in a verbal joust with anyone, and perhaps my rant was
>partially unjustified, but I take offense to being made feel like i don't
>belong in the Lisp community simply because I am a beginner. If that's not
>what he was saying, then I apologize for going off. Otherwise, there's no
>reason for me to sit here and have someone tell me that i'm not worthy of
>programming in Lisp. In any case, that's how it came across.


Zach, I think all you have to do is to give another call to Franz and/or
send an e-mail to the sales. I really have had a very positive experience
with their attitude. They obviously don't have a staff of ten to handle
sales, and you might have just hit the rough patch the first time you tried.

--Vassili


my-las...@mediaone.net

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
Another opinion.

.. On Lisp, Free and Commercial ..

Lisp is a fairly complicated language and environment. It has always been my
opinion that the best lisp is a supported lisp, and that the quality of one's
experience with lisp is proportional to the quality of support.

Some free lisps are supported, but not to the level of a commercial lisp.

Even for my own personal endeavors, I purchase a supported lisp. I don't like
spending the dollars, but for the value I receive it certainly is justified when
compared to things like Visual C++, etc.

And I definitely prefer to write my code, and not spend my time doctoring
someone else's code in some free lisp implementation. I've written GC's, but
that doesn't mean I want to write another one, for no pay, when I have more
pressing code to write, and throw it into some free lisp.

On the other hand, sometimes getting a reproducible bug report to a lisp vendor
is vastly more work without the sources, and I'm not happy when I spend days
essentially acting as a QA resource for a lisp vendor. But these things happen.
At least with a lisp vendor I get (a) the fix, and (b) some degree of
expectation that the fix will be maintained in future versions. That's more
than I can say for bugs I find in the Sun JDK. There it's a case of "maybe I'll
get a fix, some day".

I'm really glad to see new lisps like Corman's lisp come about, simply because I
believe in lisp and want to see it stick around. (I like the .sig I see in some
poster's messages in this group, "those who don't know lisp are destined to
repeat it", or something like that). I'm also glad there are free lisps, and I
don't mean them any injustice. Maybe they're very good. I haven't used them,
except for Kyoto Common Lisp many years ago, which is more or less when I
adopted my motto: "The only good lisp is a supported lisp". KCL was adequate,
but nothing to write home about. That doesn't mean I wouldn't use it, I just
prefer a more supported lisp and am willing to spend the money. But I'm
dedicated to my lisp applications, and they're not just academic toys for me, so
I'm also dedicated to using a well supported lisp.

.. On Franz ..

I've been a Franz customer for years, their support is very good. Sure,
sometimes there are things they don't fix because it's too hard, or they just
can't cost justify it. But I've never been left with a mission critical
application failure by them. If it's critical they fix it, at least in their
mainstream products (Lisp).

Some of their lesser-known products (such as Allegrostore) receive less support
because there is less demand for the product, and less revenue to justify fixes.
It's a chicken-and-egg problem, since they'd sell more if they supported it
better, in my opinion.

That Franz wants to make money is fine with me. They don't give product away
and that's okay, I'd hate to see them go out of business for being
revenue-stupid. I've found that they're open to negotiation if you're
developing non-commercial applications and want a higher powered version of
their product. They'd rather sell you some lisp for less than not sell it at
all, as long as you're not trying to take advantage of them (their good support
has real costs, after all).

I admit that even as a satisfied customer, I have found one or two acts of
pricing by them to be rather predatory, so I'm not always sanguine about the
money, but by and large they're sensible, negotiable, and provide value for your
dollar.

I expect Harlequin is too. I only hope they're not about to go six-feet-under
as a lisp vendor, it'd be terrible for the lisp community both by removing a
source of lisp (threatening the lisp supply), and allowing supply and demand to
raise my costs (threatening *my* lisp supply). Any analogy you draw between
lisp and certain recreational drugs is strictly your own :-)

... On commercial application deployment ...

I strongly dislike the current royalty requirement for commercially deployed
lisp applications. I believe that lisp should be like other languages, allowing
me to deploy my application without owing the lisp vendor a cut of the deployed
application. ACLPC 3.0.2 used to allow this, but ACL5 does not. If Franz is
too demanding in the royalty clause for my current project, it could kill use of
their lisp as the tool to make it happen. It could also kill my project. But
again, they're open to negoation, and I view them as a strategic partner and
enabler, not as someone trying to kill my project.

A free lisp doesn't necessarily have this restriction, but the GNU copyleft can
equally kill use of the free lisp for deployment of COMMERCIAL applications.

Your mileage varies with various free lisps or lisp vendors in this regard,
Franz isn't alone and so I don't mean to single them out. If I used Symbolics
platforms to deliver my app, then I'd either need to ship Ivory machines, or
perhaps license OpenGenera for all the platforms. I'm not sure about the
latter, but it's a good bet they're not going to let you deploy a Genera app
without a Genera license.

Either way, I feel that this particular thing hurts the viability of lisp as a
commercial application deployment vehicle. Tough call though, since it also
helps current lisp vendors remain in business. Ah well...

Apologies for the longwinded reply.
D. Tenny
my-las...@mediaone.net - no spam please

Zachary Turner

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to

>Zach, I think all you have to do is to give another call to Franz and/or
>send an e-mail to the sales. I really have had a very positive experience
>with their attitude. They obviously don't have a staff of ten to handle
>sales, and you might have just hit the rough patch the first time you
tried.


Yeah, like I said a few posts ago, they got back to me and have been very
helpful and accommodating. I am quite happy with them.

Zach

Craig Brozefsky

unread,
Dec 7, 1998, 3:00:00 AM12/7/98
to
my-las...@mediaone.net writes:

> A free lisp doesn't necessarily have this restriction, but the GNU
> copyleft can equally kill use of the free lisp for deployment of
> COMMERCIAL applications.

CMUCL is public domain, you could do whatever you want with it.

The GPL would only kill your application if you did not want to, or
couldn't, distribute the application under the GPL, and you were using
a GPLed lisp. I can think of many applications for which this would
be a project killer, but I can also think of many where it would not
make a difference.

David Steuber The Interloper

unread,
Dec 8, 1998, 3:00:00 AM12/8/98
to
On Mon, 7 Dec 1998 12:15:36 -0000, "Zachary Turner"
<ztu...@elsitech.com> claimed or asked:

% You don't have many friends do you?

Enhance your calm, Zachary and Erik.

While your little flame war is quite entertaining, it is something of
a waste of bandwidth.

Back on topic, I am finding that Graham's "ANSI Common Lisp" is a good
book. It is a bit steep. No running up that hill. But the book is
thin like K&R C and contains exercises relevant to the material that
was introduced. There is always this news group for additional
information if Graham proves too difficult.

BTW, that little remark about Erik seems uncalled for. Erik has never
held back when posting to this group, and I for one am glad of it.

--
David Steuber (ver 1.31.3a)
http://www.david-steuber.com
To reply by e-mail, replace trashcan with david.

May the source be with you...

Erik Naggum

unread,
Dec 8, 1998, 3:00:00 AM12/8/98
to
* "Zachary Turner" <ztu...@elsitech.com>

| I dont' want to get in a verbal joust with anyone, and perhaps my rant
| was partially unjustified, but I take offense to being made feel like i
| don't belong in the Lisp community simply because I am a beginner.

your being a beginner had nothing to do with it. your being arrogant out
of your ignorance does. ignorant people can learn. arrogant ignorant
people can't. I welcome anyone who wants to learn. I'd rather those who
can't just go away.

| If that's not what he was saying, then I apologize for going off.

accepted.

| Otherwise, there's no reason for me to sit here and have someone tell me
| that i'm not worthy of programming in Lisp.

you will find no grounds for his paranoid delusion if you actually read
what people write instead of reacting as if they wrote what you think it
looks like they did.

| In any case, that's how it came across.

... to you. there's no universality in this. don't think there is.

Georg Bauer

unread,
Dec 8, 1998, 3:00:00 AM12/8/98
to
In article <joswig-0712...@194.163.195.67>, Rainer Joswig wrote:

Be warned, this get's long. Hey, _you_ started me on this, so you
deserve what you get ;-)

>Will access to source magically give a Lisp system

>a usable GC? Even getting a decent
>UI seems not to be possible.

Does being commercial help there? I say no. Just because there are
_loads_ of good Open Source software out there. Sure, there are parts
missing - everybody can feel free and add what's missing. The problem I
have with commercial products is that you are tied to what the company
delivers. And there the money plays and not the end user (except where
the end user is where the money is - often enough that is not the case).
If the user would be so important, there would be a CLIM implementation
on every commercial common lisp by now. And there won't be a fight
between SUN and M$ about what Java should be. And there wouldn't be
something like the Halloween Documents.

Ok, granted, the few Lisp companies left are doing quite good. But that
isn't symptomatic for commercial products. For example I would never try
to build an ISP on commercial systems and commercial software. Too much
problems - just look at parts like encryption software, interoperability
etc. No, M$ might try to _sell_ you that there software is
interoperable, but in reality it isn't. They change things like it is
fitting them, even if they break available standards with that.

Sure, sometimes Open Source Software breaks standards, too. But then I
am able to fix that. Not always myself, but it is easier to hire someone
with the right knowledge when the source is available and not a closed
secret. For example I would really like to have sources at least for
parts of Exchange Server, just to get this FPOS to run correctly.

From my point of view of administering Novell and NT servers now for
some years and using in parallel Linux based systems for as many
situations as possible (of course, the important part here is "possible"
- there are lot's of situations where I won't use Linux, just because
there still isn't software for that kind of situation), I would
definitily give Linux the thumbs-up and commercial support the
thumbs-down.

The main problem is, that commercial support is often not up to what
they try to sell. That's the case with every _big_ company (big in the
really big sense) I met. The smaller companies do much better, but it
still is a problem when you want to change things the way _you_ want it.
As soon as you hit the nerve of a programmer when you want to do things
that he doen not like, you _wish_ you had the source.

>So you lately have looked into your Genera file system, right?

You happened to notice that Symbolics is a bit of dead at the moment?
Just asking. No worries :-)

>Maybe you even have installed the additional source folder
>of MCL on your Mac. Did you need it? Much? Would you
>have needed it?

Sure. If not for patching, but for documentation. Because the available
documentation for this commercial product does much leave to be desired.
So I look where I know where the documentation is: in the source. As I
said, Lisp companies doing much better in this problem than others.
Where did you find the sources to the runtime library of Visual Basic?
And how about applying security patches to IIS or Netscape FastTrack
(the latter one might change some time, since Netscape itself joined the
Open Source idea).

>I better ask the experts.

Yes, you definitely should ask them. Hey, you practically _begged_ for
this comment :-)

>Now it is a trademark? Religious marketing bullshit.

Nope. It is just a way to prohibit others to jump a bandwagon without
giving back to the community. There are several ways to solve that
problem. One is the GPL that has many problems (although most of them
are only psychologic problems and not real ones) in the commercial
world. Another one is the Open Source trademark: just a very simple and
easy to follow statement about what software is Open Source compliant
without the "idealistic" hassles of the GPL. So it is much easier to
talk about Open Source in a commercial surrounding than it is about GPL.

You _did_ notice that commercial companies begin to join the Open
Source idea? And some of them really without any force? Even for big
companies there are areas where Open Source pays the rent (like IBM and
Apache to give one example).

Ideologic blindfolding is happening on both sides - those supporting
Open Source and those refusing it. And it is bad on both sides. Open
Source is not the enemy of commercial products. It is just another way
of doing business with different goals.

BTW: Open Source is named in this way to avoid the confusion the "Free
Software" idea had. It is not important that the software is available
for free. It is important what level of freedom in using the available
source you have, to distinguish Open Source from simple giveaways.

To sum it up for me: for servers I prefer Linux. That's why I am still
interested in free Common Lisps, as they would make a perfect base for
CLHTTPD on Linux. On workstations, I still prefer my Mac running MacOS
just because of the IMO superior GUI. There are some promising new
developments for X, but up to now they are far behind what the Mac gives
me.

Sure, this is my personal resume, others might look different at the
situation. But your original posting that I commented in a somewhat
sloppy fashion just cried for some form of correction. Sure, usually I
am a Open Source advocate, but since I have something like a real life
in a real job, I have to look at things differently than RMS for
example. And funny enough, after now over 12 years of work, I think I
like the Open Source idea even more than before. When it wasn't for the
job, Open Source and GPL only gave me free access to good software. Now
I get high-quality tools to get my job done. The "for free" is not
important any more - but the "source included" is.

There are only few standing axioms in the computing business, and those
are not silly things like Moore's law but more things like:

1. All software sucks.
2. All hardware sucks.
3. Sales-Droids are your enemy.
4. There is only one way to get it done: do it yourself.
5. Management will choose the technologically inferior product.
6. Software is only as good as the next guy. And that is a moron.
7. If it is a stupid idea to ride a dead horse, management will do it.

Open Source doesn't solve those, but it helps to solve the problems
arising out of those :-)

And yes, there is a Open Hardware movement, too. And yes, there are
companies that join in. Not many, but it is a start. Make an end to
closed hardware where you don't get a driver for _your_ OS, just because
no one get's a description of the interfaces. And I really believe
that that has nothing to do with religion but everything with good common
sense.

bye, Georg

Erik Naggum

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
* g...@jill.westfalen.de (Georg Bauer)

| The problem I have with commercial products is that you are tied to what
| the company delivers. And there the money plays and not the end user
| (except where the end user is where the money is - often enough that is
| not the case).

I think this is your crucial mistake.

| If the user would be so important, there would be a CLIM implementation
| on every commercial common lisp by now.

and this is an example of where the flaw in your logic becomes visible.
it _is_ because users are important that there is a CLIM implementation
on every commercial Common Lisp system by now, yet it only barely pays
for itself.

| And there won't be a fight between SUN and M$ about what Java should be.
| And there wouldn't be something like the Halloween Documents.

ahem, it's because they fight over who should get more _users_ that this
happens. all this is about market share and mind share. it doesn't
matter whether Bill Gates or James Gosling sit in their offices and think
up evil or good things (respectively) unless they go out there and get
hoardes of _users_ to agree with them with their checkbooks open.

since you're a fellow Germanic European, I can perhaps say that our
cultures are _very_ far from appreciating the customer-orientedness of
American capitalism. the typical "Staatskapitalismus" we find in Europe
is evil because it actively limits the choices of the customers. (and
why did I use the German word for this phenomenon? there is nothing like
in American English. that alone should be worth thinking about.)

| Ok, granted, the few Lisp companies left are doing quite good. But that
| isn't symptomatic for commercial products.

really? perhaps I can compare this to the music industry. most of the
bands and musicians I like have a pretty limited market. (many of them
are German, by the way.) I buy their CD's because I like their music and
know that my music pusher is a good guy who doesn't cheat them or their
labels out of money. however, there's been a huge increase in music in
MP3, and the people who don't want to pay the _musicians_ for their work
are now targeting the increased sensitivity and hostility of the _large_
music distributors and labels towards MP3 distribution of music. not
only do they stand to lose a significant amount of money, they can afford
to fight, so I'm not worried about the large labels and distributors.
I'm worried that if MP3 becomes the preferred medium of distribution, it
will be very much harder for my favorite bands and musicians to buy the
_fantastically_ expensive equipment they use to create their (electronic)
music. it is reasonable to argue that a transition to distribute music
as MP3 will cause a bad case of _commercialism_ in what is available to
the listening audience, which means: I don't get to listen to the music I
like. I might add that spent twenty years thinking I lacked appreciation
for music until I found some labels that sell a couple thousand copies of
each CD _at best_. the CD medium made the music I like possible. MP3
might make it as unavailable as it once was.

the commercial products you whine about are bad because the _customers_
don't care. it's because the f*cking lusers continue to buy shitware
from Microsoft and put up with their fantastic policy of charging the
customers for fixing their own mistakes, whereas the users of free
software differ not in how much they want to pay or what they are
actually buying, but because free software users _care_ more.

| For example I would never try to build an ISP on commercial systems and
| commercial software.

well, I was an ISP back in 1987. (that was when my Y2K drive started and
I argued to the IETF WG on Host Requirements that RFC 822 be updated to
use four-digit years -- RFC 1123 is the result.) back then, if you
wanted something that worked, you got it from companies that actually
cared, and it cost a _lot_ of money. (except the phone company -- the
phone company cares about as much what happens to you as the IRS does.)
getting free software is a _fantastic_ luxury. I _love_ luxuries, of
course, but I don't pretend they aren't luxuries, and I certainly don't
take luxuries for granted. (I do follow the advice of Lazarus Long and
budget luxuries first, though.)

| No, M$ might try to _sell_ you that there software is interoperable, but
| in reality it isn't.

I have a new .signature today. paraphrased, Bill Gates understood early
on that he should not call people who didn't want to understand computers
idiots. he decided to take their money. so of _course_ you've been had
if you buy any of the useless piece of shit he peddles to the ignorant
masses. likewise, do _you_ participate in the Staatslotterie? or do
you, like me, consider lotteries to be an extra tax on stupid people?

if you believe Microsoft's propaganda and you wind up ripped off and
naked, do you become an emperor?

| The main problem is, that commercial support is often not up to what
| they try to sell. That's the case with every _big_ company (big in the
| really big sense) I met. The smaller companies do much better, but it
| still is a problem when you want to change things the way _you_ want it.

well, I'm like this one-person software company who does it because it's
a nice way to avoid real work and still pay all the bills, and my clients
don't get what they say they want, either. if I don't agree with them,
they get to pay somebody else to do it (and come back to me afterwards).
most of the time, we come to terms on what they really want, and that's
what they get. if the client knows much better than me what he wants, he
doesn't need my services. I think this attitude scales very well. as
long as you're free to go elsewhere, I don't see the problem.

now, you'll argue that you _aren't_ free to buy software from whichever
software company you want to, but that's _also_ wrong. it may cost more,
but that means you either operate with a smaller margin, charge more, or
leverage the costs across more volume -- or decide against doing it,
which is no crime. using Microsoft products is also much more expensive
in real terms than using quality software, but it'll take a while before
the suits who don't want to understand computers get sick and tired of
having their money taken away.



| Ideologic blindfolding is happening on both sides - those supporting Open
| Source and those refusing it. And it is bad on both sides. Open Source
| is not the enemy of commercial products. It is just another way of doing
| business with different goals.

I think free software can be the enemy of commercial software the same
way MP3 can be the enemy of commercial music. we don't have to worry
about the high-volume end of the market. it's the nigh invisible, small
players that need to be worried about.

| BTW: Open Source is named in this way to avoid the confusion the "Free
| Software" idea had. It is not important that the software is available
| for free. It is important what level of freedom in using the available
| source you have, to distinguish Open Source from simple giveaways.

amusing. "Free" in Free Software doesn't mean "available for free" the
same way "Free Nelson Mandela" didn't mean you got a piece of him and a
balloon at a parade.

#:Erik
--
don't call people who don't understand statistics idiots. take their money.

Tim Bradshaw

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
* Georg Bauer wrote:
> Does being commercial help there? I say no. Just because there are
> _loads_ of good Open Source software out there. Sure, there are parts
> missing - everybody can feel free and add what's missing. The problem I
> have with commercial products is that you are tied to what the company
> delivers. And there the money plays and not the end user (except where
> the end user is where the money is - often enough that is not the case).
> If the user would be so important, there would be a CLIM implementation
> on every commercial common lisp by now. And there won't be a fight
> between SUN and M$ about what Java should be. And there wouldn't be
> something like the Halloween Documents.

Well, the flip side of this is: has there ever been an OSS Lisp which
has a seriously good GC (comparable with most recent commercial
Lisps)? I'm not aware of one, although I don't know what CMUCL does
on X86 these days.

And, actually, which commercial Lisps *don't* have CLIM?

--tim


Rainer Joswig

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
In article <slrn76r1...@jill.westfalen.de>, g...@jill.westfalen.de
(Georg Bauer) wrote:

> >Will access to source magically give a Lisp system
> >a usable GC? Even getting a decent
> >UI seems not to be possible.
>
> Does being commercial help there? I say no.

Isn't it funny that GCs of commercial systems are lightyears
ahead? That I have yet to see more than a handful
"free" software products with a *usable* user interface?
To create a great product it takes more than
just naming it "open", "free", "open source", ...
What we end up with are geek products. Ugly,
idiosyncratic, complicated, fragile, maintenance
intensive, unsupported, ...

> Just because there are
> _loads_ of good Open Source software out there.

Where? Most of the Open Source stuff is written by geeks
for geeks. Most of the stuff is ***extremely*** user hostile.
A lot of the stuff is of extraordinary poor design.

I'm not saying that there are no good "Open Source"
tools - but saying that this is the path
to the software nirvana is ridiculous.

> Sure, there are parts
> missing - everybody can feel free and add what's missing.

I don't have neither the time, the knowledge nor the will
to need to deal with this stuff.

> The problem I
> have with commercial products is that you are tied to what the company
> delivers.

Come on.

> And there the money plays and not the end user (except where
> the end user is where the money is - often enough that is not the case).
> If the user would be so important, there would be a CLIM implementation
> on every commercial common lisp by now.

We have a CLIM implementation on every commercial Lisp by now.
I can use CLIM on Mac/Unix/Lispm/Windows.

But not that there is some possible user somewhere makes things
magically happen - the demand will drive it. Then
you need to be able to deliver.

> Ok, granted, the few Lisp companies left are doing quite good.

Where are you living? I have the complete opposite view.
How can it be that we look at the same thing - yet I get
a radically different impression. The Lisp market
is silent. Example: How many Lisp projects have you
heard/seen/... about that
were started in Germany in the last two years?
One? Two? How big is Germany?

> For example I would never try
> to build an ISP on commercial systems and commercial software.

We do. We are using SUNs and Solaris, Ascends, Ciscos, ...
Most commercial ISPs I know are built around this with
a lot of commercial and a lot of "free" software.

>Too much problems

With crap like SENDMAIL, etc.? Yes. Thanks for the source.
I would have preferred not to look at the source.
Mayordomo?

> >So you lately have looked into your Genera file system, right?
>
> You happened to notice that Symbolics is a bit of dead at the moment?
> Just asking. No worries :-)

No, they are still alive.
I'm still using my Lisp machine.

> To sum it up for me: for servers I prefer Linux. That's why I am still
> interested in free Common Lisps, as they would make a perfect base for
> CLHTTPD on Linux.

But there are not and they are years away from that.

> Sure, this is my personal resume, others might look different at the
> situation. But your original posting that I commented in a somewhat
> sloppy fashion just cried for some form of correction. Sure, usually I
> am a Open Source advocate, but since I have something like a real life

So your COBOL programs are Open Source? The company
who pays you will release them as Open Source?

> 1. All software sucks.
> 2. All hardware sucks.
> 3. Sales-Droids are your enemy.
> 4. There is only one way to get it done: do it yourself.
> 5. Management will choose the technologically inferior product.
> 6. Software is only as good as the next guy. And that is a moron.
> 7. If it is a stupid idea to ride a dead horse, management will do it.

Come on, this list is plain stupid.

What I'd like to see are more people with a vision, who really
want to move things forward, making using computers more
fun and more productive. I'm also pretty sure that
the Lisp community can provide a lot of ideas to make
that happen.

> Open Source doesn't solve those, but it helps to solve the problems
> arising out of those :-)

Actually the current stuff that is being distributed
as Open Source or whatever are not really part
of the solution. Often they are part of the problem.

Example: Linux sucks. Big time. Stuff like that frustrates me.

> And yes, there is a Open Hardware movement, too.

Another religious marketing bullshit.

--
http://www.lavielle.com/~joswig

rusty craine

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to

Erik Naggum wrote in message <31221713...@naggum.no>...


> don't call people who don't understand statistics idiots. take their
money.

Our little community had a party to celebrate the season last week. There
were 50 or so people there. Computers were the topic of several
conversations during the evening. As I listened to the topics, it became
clear to me that the "ignorant" masses didn't really care that they were
running software per say or bill gates by default. Their collective concern
was only functionality of the total package. ---"I can watch the teley,
play my DVD [sounds like a new brand of underware], play "space blaster",
write my term paper (or download a 'cheater' from the _NET_)". Seemed the
quote of the night.

Movies, games, the internet, a spread sheet, and a little word processing
must be what sells comptures to the masses. This must be the market that
Gates is after. I would say he has done a remarkable job in that market,
Ayn Rand must've smiled.

At their worst Gates and Micorsoft can't be all bad, the masses have voted
with the $$$. To a family at the party, they all had at least one computer.
Many had several, upgarding and giving the old one to "Suzie" for her room.

Alas lisp was not on a single drive save mime, VB was on several ["but i've
not learned to program yet only cost 90 bucks"], java on none, IE on most,
and a few Netscape.

I think Bill Gates is here to stay, guess we should try to adjust. If you
think ya got it bad come over to the 390 for a while. You will use CICS and
that is no shit. Put out by one IBM, as they say, as they do....you and god
will comply.

me and god waiting for IBM to tell us what to do
rusty

Espen Vestre

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
"rusty craine" <rccr...@flash.net> writes:

> I think Bill Gates is here to stay, guess we should try to adjust.

I don't know, Win2000 seems to be such a monster that they *must* be
programming themselves into a corner of an enormous room in a house
they don't longer know how they even got into... MS has made remarkable
turns before (e.g. MSN -> Internet), but will they be able to get
out of the bloatware spiral without damage?

Speaking of your friends, I think they'll get tired of bugs at _some_
point. People are remarkably patient with their PCs, but I think
that has to do with the facts that they don't know better and, most
important, that they're having such fun that they don't mind a few
problems. But at some point, this will change, and they will demand
that the machines do real work for them. Don't forget what happened
to the american car industry, when consumers _understand_ that they
don't get the quality the deserve, they'll do something about it.
And the computer industry will learn that the hard way, just like
the american car industry!

Or so I hope...

--

espen

Larry Hunter

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to

I don't really want to get involved in the war over commercial vs. "open
source" (like most of these X vs. Y battles, my opinion is "both, please")
but I couldn't let this response by Rainer Joswig to Georg Bauer's comment
go by:


Bauer: "To sum it up for me: for servers I prefer Linux. That's why I am


still interested in free Common Lisps, as they would make a perfect base
for CLHTTPD on Linux."

Joswig: "But there are not and they are years away from that."

Rainer, you of all people should know better. The CL-HTTP web page
(http://www.ai.mit.edu/projects/iiip/doc/cl-http/home-page.html) says:

Douglas Thomas Crosher has ported CL-HTTP to CMU Common Lisp which runs on
the 18b release of CMUCL, and is considered beta-test software when
running multi-threaded on Intel X86 hardware under FreeBSD and Linux"

As with many open source projects, it may take some technical effort to get
the desired level of performance, but it is clearly possible to make it
happen today.

Larry

--
Lawrence Hunter, PhD.
National Library of Medicine phone: +1 (301) 496-9303
Bldg. 38A, 9th fl, MS-54 fax: +1 (301) 496-0673
Bethesda. MD 20894 USA email: hun...@nlm.nih.gov

Paul Rudin

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> > Does being commercial help there? I say no.
>

> Isn't it funny that GCs of commercial systems are lightyears
> ahead? That I have yet to see more than a handful
> "free" software products with a *usable* user interface?
> To create a great product it takes more than
> just naming it "open", "free", "open source", ...
> What we end up with are geek products. Ugly,
> idiosyncratic, complicated, fragile, maintenance
> intensive, unsupported, ...
>


Wandering a little off topic here, but a few couterexamples that
spring to mind:

A goodly percentage of software infrastructre of the internet is free,
open source software.

Linux is a more realiable OS than Windows.

Mozzilla is at least as good as Internet Explorer

The best document preperation system for scientific/mathematical
material is (La)TeX by a long way..

A little more on topic, (X)emacs/ilisp is the best lisp editor around
IMO.


Rainer Joswig

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
In article <rbr9u92...@work.nlm.nih.gov>, Larry Hunter
<hun...@nlm.nih.gov> wrote:

> Bauer: "To sum it up for me: for servers I prefer Linux. That's why I am


> still interested in free Common Lisps, as they would make a perfect base
> for CLHTTPD on Linux."

Read: "perfect base"?

> Joswig: "But there are not and they are years away from that."
>
> Rainer, you of all people should know better. The CL-HTTP web page
> (http://www.ai.mit.edu/projects/iiip/doc/cl-http/home-page.html) says:
>
> Douglas Thomas Crosher has ported CL-HTTP to CMU Common Lisp which runs on
> the 18b release of CMUCL, and is considered beta-test software when
> running multi-threaded on Intel X86 hardware under FreeBSD and Linux"

Read: "beta-test software".

> As with many open source projects, it may take some technical effort to get
> the desired level of performance, but it is clearly possible to make it
> happen today.

Linux and CMUCL are not a perfect base for CL-HTTP right now.

More work is needed. It is nice that it is kind of running at all.

--
http://www.lavielle.com/~joswig

Rainer Joswig

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
In article <m31zm9l...@shodan.demon.co.uk>, Paul Rudin
<pa...@shodan.demon.co.uk> wrote:

> A goodly percentage of software infrastructre of the internet is free,
> open source software.

Depending on the definition of "free" the percentage varies.

> Linux is a more realiable OS than Windows.

In general or the individual C program? Does BIND crash
less often or more often than a Windows NT DNS?

> Mozzilla is at least as good as Internet Explorer

I haven't seen a released open source Mozilla yet. Where is it?
The thing in development was highly buggy last time I checked.

> The best document preperation system for scientific/mathematical
> material is (La)TeX by a long way..

Yes.

> A little more on topic, (X)emacs/ilisp is the best lisp editor around
> IMO.

I'm not using it as a Lisp editor. Well, the X-Windows UI
is like a stone tied to the foot. But there are easier to use
alternatives.

--
http://www.lavielle.com/~joswig

Erik Naggum

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
* "rusty craine" <rccr...@flash.net>

| At their worst Gates and Micorsoft can't be all bad, the masses have
| voted with the $$$.

the _masses_ voted for some other things that turned out to remarkably
evil, as well. you can basically count on the masses to do whatever the
best demagogue tells them to do. Microsoft is about marketing and fraud,
and they're very good at it. it has never been about software. had they
spent their energy on the software, it would probably have been great,
but they spend as little as humanly possible on the software to make as
much money as inhumanly possible.

| I think Bill Gates is here to stay, guess we should try to adjust.

lots of things are here to stay. Hinduism, for instance. biggest
religion there is. have you converted, yet? if not, why not?

Craig Brozefsky

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> > Linux is a more realiable OS than Windows.
>
> In general or the individual C program? Does BIND crash
> less often or more often than a Windows NT DNS?

Uhm, both the kernel is more reliable, and for the large part, most of
the programs in the various distributions are more reliable. BIND is
definetly better than MS DNS, I have horror stories about it if you
want.

> > Mozzilla is at least as good as Internet Explorer
>
> I haven't seen a released open source Mozilla yet. Where is it?
> The thing in development was highly buggy last time I checked.

That's because it's in "development". Enter, "Gecko", the result of
the open source effort which is being hailed as an extremely fast and
small layout engine by most major media outlets (outlets, or
sphincters, I can't decide what they are). Mind you that Netscape has
been paying alot of developers of their own to work on the open code
base, so I think this may be a bad example for the argument the poster
was intended to make.

Apache is the obvious example. I've worked with web servers ranigng
from IIS, to Netscape (all generations) to IBM GoServer and all it's
weird Lotus-bound contraptions for torture. Apache is basically
the best hands down, and if you want a particular "type" of server,
"single-process multi-threaded" or "multi-process" or whatever, there
are close to a dozen free servers which fit those models, and kick
much ass.

I guess I would argue that the management of the project, the talent
working on it, and it's ability to capture the interest of users and
developers is a primary determinant wether it is open-source or
commercial. The Lisp vendors are extremely competent, but the same
cannot be said for commercial developers in a large number of other
markets. Thankfully, the Lisp developers working with free software
are extremely competent as well.


Paul Rudin

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> In article <m31zm9l...@shodan.demon.co.uk>, Paul Rudin
> <pa...@shodan.demon.co.uk> wrote:
> Depending on the definition of "free" the percentage varies.
>

> > Linux is a more realiable OS than Windows.
>
> In general or the individual C program? Does BIND crash
> less often or more often than a Windows NT DNS?

I mean the operating system kernel itself. Programs may fall over on
either platform, but on linux you can kill the offending program
and restart it. On windows you often need to turn off the machine and
start again. My experience is that in 7 years of running a number of
different linux distributions on several machines, the OS itself
has crashed once. OTOH rebooting a windows box because the OS has
fallen over happens every couple of days.

>
> > Mozzilla is at least as good as Internet Explorer
>
> I haven't seen a released open source Mozilla yet. Where is it?
> The thing in development was highly buggy last time I checked.

I agree that the latest open source releases are buggy. But the binary
only "netscape" releases are free in any case..

> > A little more on topic, (X)emacs/ilisp is the best lisp editor around
> > IMO.
>
> I'm not using it as a Lisp editor. Well, the X-Windows UI
> is like a stone tied to the foot. But there are easier to use
> alternatives.

Which X windows UI? I've got about 7 or 8 installed on the machine I'm
sitting at just now; and there are plenty of others around. If you
like windows then something like KDE is not that dissimilar in a lot
of respects; and FVWM-95 is a deliberate attempt to emulate some of
the features of the W95 desktop.


"easier to use" is perhaps the key phrase here. The nice thing about
using emacs as an program editor is that you can get more or less the
functionality you want if you're prepared to put the effort in. This
may not be "easy", and the initial learning curve to get up to speed
with using emacs may also not be "easy". But in the long run you get
tools and techniques that you can use across different programming
languages (and other kinds of data).


(And incidentally, I'm reading this newsgroup and preparing this post
inside gnus, an emacs newsreader, and since I have X running on three
machines I have several emacs windows open on three separate monitors,
giving me various presentations of the thread I'm participating in;
the article I'm replying to and the article I'm typing. Each X server
has its own console, and yet the emacs windows are all under the same
emacs process so the underlying data it the same. There isn't a
commercial product available that I'm aware of that would allow
anything close to similar functionality.)


Don't get me wrong: I don't have anything against commercial software
(as of a few weeks ago I'm a full time common lisp programmer, for a
company that sells commercial software). There is also some great
commercial software out there (where would we be without Doom and its
imitators :-) ). But I do think your characterisation of free software
is wide of the mark.


Mike McDonald

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to
In article <rbr9u92...@work.nlm.nih.gov>,
Larry Hunter <hun...@nlm.nih.gov> writes:

> Douglas Thomas Crosher has ported CL-HTTP to CMU Common Lisp which runs on
> the 18b release of CMUCL, and is considered beta-test software when
> running multi-threaded on Intel X86 hardware under FreeBSD and Linux"

But for all practical purposes, there is NO 18B release of CMUCL.

Mike McDonald
mik...@mikemac.com

Steve Gonedes

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to

Paul Rudin <pa...@shodan.demon.co.uk> writes:

< Wandering a little off topic here, but a few couterexamples that
< spring to mind:
<
< A goodly percentage of software infrastructre of the internet is free,
< open source software.
<
< Linux is a more realiable OS than Windows.

Please, don't misinterpret what I'm saying as I like linux very much -
it works great for me but...

What is reliability? Will linux support/do x, y, and z tomorrow?
If you're answer is yes then _that is my point_!

There is no security offered by linux in this and other areas because
_it is a response_ to a `lack of' and was not, and simply cannot
(currently?) be a method to drive and create (partially due to the
attitude which linux thrives on - that is the meeting of a challenge).

Not to say that linux is not doing new things etc. I like it very much
and it works great for me. For example no hardware vendor will create
new hardware due to the perceived flexability that linux offers them
(this is much the same for most every `potential solution' in the
`computer' industry right now, no reason to stop trying or anything
though, IMHO). Even better - no new internet protocols are written
because of the benefits that linux offers - if anything new protocols
are old ideas just clarified to better understand why people keep
failing to solve the same problem (again and again and ...).

Rest assured, Microsoft will not be creating new internet protocols
due to linux. The halloween documents were a joke to demonstrate just
how much control MS has over the media. (Remeber also, microsoft
cannot create software - they never have and never will. They'll soon
have no new software to steal and will being selling people _access_
to their own information, just watch! It's already happening.)

< Mozzilla is at least as good as Internet Explorer

They are both pieces of shit in that neither of them _SOLVE ANY
PROBLEMS_; their total benfit is being derived from their ability to
_create new problems_! For instance, the newer versions of internet
exporer keep getting slower - which demonstrates that it _is not
addressing any problems!_ The right thing to do is _create_ a method
to represent information is a new, efficient (in terms of clarity and
physical space) mannor which will eventually allow for new ways to
express ideas on the internet. Instead we have reoccuring `religious
experiences' guided by reoccuring mistakes...

< The best document preperation system for scientific/mathematical
< material is (La)TeX by a long way..

The worse thing about TeX is how it promotes stagnation (read:
regression) and treats any idea of progression as undesirable and
therefor non-existant. TeX is such a horrendous piece of shit that
people _are scared_ of changing it! There have been no, and there
_NEVER WILL BE_, new ideas or developments in the TeX area of
typesetting. The simple fact that there are numerous versions of latex
proves that the only solution is to dump the cruft and move on
- same goes with X-windows, extract the xfree drivers and dump the
cruft! You see, latex failed to address the problem. The problem, TeX,
was readily recognized and admitted at one point, but people are
beginning to forget this.

< A little more on topic, (X)emacs/ilisp is the best lisp editor
< around IMO.

This is not because it is free software - it is because it solves some
of your problems (whatever they may be)...

[I'm not against free software - in fact I like it and respect it
_very much_, I just get very annoyed when people forget what the
benefits really are and when they criticize other people's beliefs
without valid justification. (and here I am, is this ironic?) It makes
them sound like baptists screaming how evil the moromons are (if you
follow religion at all) :) ]

Also, do you notice how your acceptance of `best' may prevent you from
accepting the new, hence potentially better? You will probably look at
every new editor in terms of features it lacks in terms of xemacs. All
new features that a new editor would introduce to you will most likely
result in the assimilation of new _into the old_.

The danger, you see, is that you may have come to accept some of the
original limitations as features/desirable functionality, and any
attempt to free you from them will be percieved as an attack on your
self due to your familiarity => stability, `_we_ like it', etc etc.

We're all going to die, it's sucks - I know.

[Not trying to sound nasty - just raising some potentially valid points?]


rusty craine

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to

Erik Naggum wrote in message <31222113...@naggum.no>...

> lots of things are here to stay. Hinduism, for instance. biggest
> religion there is. have you converted, yet? if not, why not?

I hope someday we have an opportunity to discuss such topics. It would make
for an interesting evening. Once upon a time I think I found the _answers_
in such a discussion, alas the alcohol blood level was not conducive to
clear recollection.

Malt does more than Milton can to explain God's way to man. (modus tollendo
pones)
Rusty

Jon S Anthony

unread,
Dec 9, 1998, 3:00:00 AM12/9/98
to

Vassili Bykov <vas...@objectpeople.com> writes:

>
> In article <31218169...@naggum.no>,


> Erik Naggum <er...@naggum.no> wrote:
> > * "Zachary Turner" <ztu...@elsitech.com>

> > | I evaluated Allegro CL 5.0 and I really like it, but these people are
> > | basically sitting here with the "No we don't your money. Go give it to
> > | Harelquin" attitude, which is really frustrating since I want to purchase
> > | Allegro. Has anyone else had better luck with them?
> >
> > yes, I have. the fact that I have happy Common Lisp clients today is
> > probably due mostly to the excellent and welcoming attitude at Franz Inc
> > when I first approached them. they have continued to be very helpful in
> > making my projects succeed, both for me and for my clients.
>
> I can add that I have had a totally different (from Zachary's) experience in
> communicating with Franz, in spite of never even having been their paying
> customer. Virtually every e-mail inquiry sent to the sales (out of

I'm a paying customer and (in general) I think Franz has been one of
the top 2 or 3 software companies I've ever had to deal with in terms
of

a) timeliness of responses

b) knowledge and information content of responses

c) proactive behavior to requests (for example, during the various
beta releases of ORBlink I would often fire off several bug
reports and/or requests for changes and often these were
addressed within a day or two with downloadable versions
reflecting the results immediately available.)

I've had some goofs surrounding certain licensing issues and
deliveries of purchased items, but really nothing major. And they
always seem to have an open ear.

Certainly compared to even the best of the ordinary vendors (take your
pick, but I mean folk like Sun, Symantec, Aonix, etc.) Franz is truly
a qualitative jump beyond. No comparison really.

/Jon


--
Jon Anthony
Synquiry Technologies, Ltd. Belmont, MA 02478, 617.484.3383
"Nightmares - Ha! The way my life's been going lately,
Who'd notice?" -- Londo Mollari

Rainer Joswig

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <m3zp8xj...@shodan.demon.co.uk>, Paul Rudin
<pa...@shodan.demon.co.uk> wrote:

> has crashed once. OTOH rebooting a windows box because the OS has
> fallen over happens every couple of days.

whoever compares an OS in terms of stability with Windows
can't be really helped. I mean anybody with some self
respect doesn't do that. "Hey my Hyundai is faster
than your Trabant." Still it doesn't give you a Mercedes
(or whatever these cars will be called now. ;-) ).

> Which X windows UI?

Any of the stuff I have seen so far is not overwhelming.

> If you like windows

No. Not really. Maybe "hate" would more correctly describe
my relationship with Windows.

> then something like KDE is not that dissimilar in a lot
> of respects; and FVWM-95 is a deliberate attempt to emulate some of
> the features of the W95 desktop.

Which already was an emulation of the Mac UI, which already
took elements from Xerox UIs. Doesn't give you the same
sophistication of a Mac UI, though.

> (And incidentally, I'm reading this newsgroup and preparing this post
> inside gnus, an emacs newsreader, and since I have X running on three
> machines I have several emacs windows open on three separate monitors,
> giving me various presentations of the thread I'm participating in;
> the article I'm replying to and the article I'm typing. Each X server
> has its own console, and yet the emacs windows are all under the same
> emacs process so the underlying data it the same. There isn't a
> commercial product available that I'm aware of that would allow
> anything close to similar functionality.)

My Symbolics Lisp machine is accessible the same way. Yet, it provides
a complete environment.

--
http://www.lavielle.com/~joswig

Rainer Joswig

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <871zm9i...@piracy.red-bean.com>, Craig Brozefsky
<cr...@onshore.com> wrote:

> Uhm, both the kernel is more reliable, and for the large part, most of
> the programs in the various distributions are more reliable.

Maybe gradually - but not in principle. Is there
any revolutionary coding practice in effect that
suddenly makes Linux software more reliable?
I mean are the developers switching to Eiffel/Ada/SML and
are really taking reliability into account (runtime assertions,
test suites, checking adherance to specifications,
exception handling infrastructure, ...)?

Or are they just hacking as ever?

> BIND is
> definetly better than MS DNS, I have horror stories about it if you
> want.

I guess there are enough horror stories about BIND, too.
Given the massive amount of work that has gone
into these things, they are still pathetic.



> commercial. The Lisp vendors are extremely competent, but the same
> cannot be said for commercial developers in a large number of other
> markets. Thankfully, the Lisp developers working with free software
> are extremely competent as well.

Yep.

--
http://www.lavielle.com/~joswig

Craig Brozefsky

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> > Uhm, both the kernel is more reliable, and for the large part, most of
> > the programs in the various distributions are more reliable.
>
> Maybe gradually - but not in principle. Is there
> any revolutionary coding practice in effect that
> suddenly makes Linux software more reliable?

Yah, massive peer review, large-scale beta testing, and use of an
educated user base to produce high quality bug reports. It's a
stretch to call these revolutionary tho, but that doesn't change the
fact that they have rather consistently resulted in high quality
software. Does there NEED to be a revolutionary coding practice to
make the reliability worth noting? I mean the issue here after all is
that there is some reliable free software out there, not that free
software is developed with revolutionary coding practices or in
whatever safe programming language is en vogue this year.

The poor sods might be C luzerz and all, but give them credit where
credit is due. At least they're writing what they want rather than
whining about being marginalized and not recognized for their obvious
intellectual superiority, or blasting people for not supporting the
dwindling ranks of the commercial vendors for their favorite
languages. Well fuck, they do that stuff too, but in addition to
writing realiable softwre 8^)

> I guess there are enough horror stories about BIND, too.
> Given the massive amount of work that has gone
> into these things, they are still pathetic.

Yah, I got those too, I got horror stories about any peice of software
I have ever used for more than 15 minutes, that doesn't make them all
pieces of shit. I've seen it randomly die, I've seen exploits to root
it, I've seen it start throwing out random bits of garbage in
responses. I've also see it run name service for universities and
large corporations nearly flawless for years.

Joachim Achtzehnter

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> Craig Brozefsky <cr...@onshore.com> wrote:
>
> > Uhm, both the kernel is more reliable, and for the large part, most
> > of the programs in the various distributions are more reliable.
>
> Maybe gradually - but not in principle. Is there
> any revolutionary coding practice in effect that
> suddenly makes Linux software more reliable?

No, of course not. But neither is the opposite true: The fact that the
source code is kept secret doesn't imply that it is well-designed or
that good coding practices were used. What I am trying to get at
here is that it is futile to compare the quality of "propriatary
software" with that of "free software" in general terms, just doesn't
make sense. We have to realise that the question of whether software is
free or not, is orthogonal to the quality issue.

Note also, that "free software" can be commercial software. There
seems to be this common misconception that somehow free software is
always produced by amateurs, or that proprietary software is always
produced by professionals. Much of the heated arguments and flaming
about this topic would be avoided if people didn't always mixup the
issues of freedom and quality. Free software can be of high or low
quality, and the same is true for proprietary software.

What we can reasonably argue about is whether it is easier to build
and maintain high quality software using the proprietary model or the
free software model. Everything else being equal (most importantly, if
the people doing the development are of comparible caliber) which
model might be more successful for a given project? This is the only
kind of question that makes sense in my mind. So in terms of the Linux
kernel, would it have been easier or more difficult for Linus, Alan
Cox, and a handful of others to build (and maintain!) something like
Linux using a proprietary approach?

And from a user's perspective, you have to weigh the alternatives: put
all eggs in one basket and accept that you depend totally on one
vendor to fix things or extend things? Or have the freedom to hire any
professional software developer to work on freely available source if
the "official" maintainers don't do their job?

And finally, from a vendor's perspective: Would a software company
be more or less successful in generating revenue with either model?
Clearly, only the proprietary model lends itself to the Microsoft
way of making money. But there can only be one (or at most a few)
Microsofts. How about companies like Franz or Harlequin? Is it so
clearcut whether they would be any worse off if they followed the free
software model?

Again, trying to compare "quality" of proprietary software with free
software out of context makes no sense. Let us not forget that most
existing "serious" software was produced as proprietary software,
while a large portion of free software was done as a hobby. Let us not
compare apples with oranges.

Joachim

--
joa...@kraut.bc.ca (http://www.kraut.bc.ca)
joa...@mercury.bc.ca (http://www.mercury.bc.ca)

Alexey Goldin

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> In article <871zm9i...@piracy.red-bean.com>, Craig Brozefsky


> <cr...@onshore.com> wrote:
>
> > Uhm, both the kernel is more reliable, and for the large part, most of
> > the programs in the various distributions are more reliable.
>
> Maybe gradually - but not in principle. Is there
> any revolutionary coding practice in effect that
> suddenly makes Linux software more reliable?

> I mean are the developers switching to Eiffel/Ada/SML and
> are really taking reliability into account (runtime assertions,
> test suites, checking adherance to specifications,
> exception handling infrastructure, ...)?
>
> Or are they just hacking as ever?
>

The only real change is that there is way more testers, all bugs are
readily exposed and easily reported because of source code
availability. And it helps a lot. I am sure if they did more of the
stuff you suggest software would be even more reliable but even first
step helps a lot. Linux is at least two orders of magnitude more
reliable then your favorite MacOS (from personal experience). Some
developers are switching to (yikes!) Perl and Python (which is
actually nice) which eliminate a lot of problems of C (like memory
leaks, bounds checking). I am sure Lisp would make these developers
even more productive because even my favorite Python is way below Lisp
on every account but I afraid they will be put off by attitude of most
competent people on this newsgroup, high cost of commercial products
(How much do you have to pay for CLIM?) and unavailability of meny
libraries they are used to in other languages.

You may want to thing long and hard before you blaiming Linux
developers for not using Lisp. They are usually very open minded and
like new stuff. Redhat standart distribution includes more languages
then any other commercial OS by order of magnitude and many of them
are necessary for functioning of an OS. If they are not using Lisp
this is because Lisp has a problem. And I do not think this problem is
syntax.

P.S. compare progress of free Smalltalk (Squeak ) in last year with
progress of any commercial Lisp and may be you will understand why
open source is important.

P.P.S. I greatly admire your professional skills (which are evident
from your posts) so please do not consider it as personal flame.


Putting on asbestos underwear anyway...

Erik Naggum

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
* "rusty craine" <rccr...@flash.net>

| I hope someday we have an opportunity to discuss such topics. It would
| make for an interesting evening. Once upon a time I think I found the
| _answers_ in such a discussion, alas the alcohol blood level was not
| conducive to clear recollection.

personally, I think most religions look a lot like such recollections.

#:Erik, who is just _begging_ for another "be nice to the humor-impaired"
thread after the Stalin thread was buried.

Rainer Joswig

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <87vhjkh...@piracy.red-bean.com>, Craig Brozefsky
<cr...@onshore.com> wrote:

> jos...@lavielle.com (Rainer Joswig) writes:
>
> > > Uhm, both the kernel is more reliable, and for the large part, most of
> > > the programs in the various distributions are more reliable.
> >
> > Maybe gradually - but not in principle. Is there
> > any revolutionary coding practice in effect that
> > suddenly makes Linux software more reliable?
>

> Yah, massive peer review, large-scale beta testing,

Sometimes I have the feeling I'm living in a beta world.
Everything now is beta. When do I get final stuff?
What once was pre-alpha now is beta.

> and use of an
> educated user base to produce high quality bug reports. It's a
> stretch to call these revolutionary tho,

Yep, Microsoft, Apple, SUN, etc. are doing the same.

> The poor sods might be C luzerz and all, but give them credit where
> credit is due. At least they're writing what they want rather than
> whining about being marginalized and not recognized for their obvious
> intellectual superiority, or blasting people for not supporting the
> dwindling ranks of the commercial vendors for their favorite
> languages.

Just look at CL-HTTP. Only one free Lisp is up to the task
running it. On one platform. With early support for
threads. But it has been used on five commercial
Lisp on different platforms. Hey, I'm not saying
that people should only use commercial software -
just that it seems simply not obvious to some people that things
are not black and white. "Open Source" is not equivalent
to "good" and "proprietary" is not equivalent to "bad".
"User" is not equivalent to "looser". It is simply
neither necessary nor sufficient for "good software"
to be "Open Source".

That people are still hacking with CLisp, CMU CL or now
with Corman Lisp is a good sign. Just don't be constrained
in your mind. There is enough room for improvement
and cool ideas. Look at Corman Lisp. Seems like
it has great potential to reach a lot of users. But this
is the result of the hard work of its developer
with a lot of experience.

--
http://www.lavielle.com/~joswig

Rainer Joswig

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <ucyaog7...@ns.mercury.bc.ca>, Joachim Achtzehnter
<joa...@kraut.bc.ca> wrote:

> What we can reasonably argue about is whether it is easier to build
> and maintain high quality software using the proprietary model or the
> free software model. Everything else being equal (most importantly, if
> the people doing the development are of comparible caliber) which
> model might be more successful for a given project? This is the only
> kind of question that makes sense in my mind. So in terms of the Linux
> kernel, would it have been easier or more difficult for Linus, Alan
> Cox, and a handful of others to build (and maintain!) something like
> Linux using a proprietary approach?

Aren't/weren't they mostly "copying" what is/was already out there
grounded on massive experience and preexisting tools?
This seems to be easier than to develop something completely
new.

> And from a user's perspective, you have to weigh the alternatives: put
> all eggs in one basket and accept that you depend totally on one

> vendor to fix things or extend things?

Or a bunch of vendors competing with each other?

> Microsofts. How about companies like Franz or Harlequin? Is it so
> clearcut whether they would be any worse off if they followed the free
> software model?

No. It would be interesting to explore the idea. Netscape
seems to see it as a way to survive as a company
(note that they are only releasing the browser software,
not the servers). How do you think would an "Open Source"
approach would look like for a Lisp vendor?

Rainer Joswig

--
http://www.lavielle.com/~joswig

Rainer Joswig

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <m1iufky...@flight.uchicago.edu>, Alexey Goldin
<gol...@flight.uchicago.edu> wrote:

> Linux is at least two orders of magnitude more
> reliable then your favorite MacOS (from personal experience).

Depends. But it is clear that the current Mac OS has
inherent design weaknesses (libs not thread safe,
no real preemptive multitasking OS, no memory protection,
static memory allocation for programs, ...) . Apple hasn't delivered
anything that really addresses these issues yet. MacOS X
might be a step forward. When it arrives. I'm
not holding my breath.

> Some
> developers are switching to (yikes!) Perl and Python (which is
> actually nice) which eliminate a lot of problems of C (like memory
> leaks, bounds checking).

Python also seems to be a great step forward from the
usual shell scripts.

> I am sure Lisp would make these developers
> even more productive

I'm not that sure about that.

> because even my favorite Python is way below Lisp
> on every account but I afraid they will be put off by attitude of most
> competent people on this newsgroup,

I don't think the attitude of some Python people is that much
different.

> high cost of commercial products

Some.

> (How much do you have to pay for CLIM?)

For example LWW with CLIM seems to be reasonably priced.
Commercial Unix Lisps are too expensive, IMHO, though.

> and unavailability of meny
> libraries they are used to in other languages.

And vice versa. I'm using libraries in Lisp that are
hardly available in other languages or are much
harder to use in these languages.

> You may want to thing long and hard before you blaiming Linux
> developers for not using Lisp. They are usually very open minded and
> like new stuff. Redhat standart distribution includes more languages
> then any other commercial OS by order of magnitude and many of them
> are necessary for functioning of an OS.

Maybe concentrating on fewer tools might help to remove
bloat and clutter. ;-)

> If they are not using Lisp
> this is because Lisp has a problem. And I do not think this problem is
> syntax.

I see some problems, too. It might be helpful to name some of these
problems and to see if they can be addressed. This
surely depends on the system we would look at. But
I think there are also general guidelines which could
improve the overall user experience.

> P.S. compare progress of free Smalltalk (Squeak ) in last year with
> progress of any commercial Lisp and may be you will understand why
> open source is important.

I don't think this is valid. Some commercial Lisp made good
progress last year and they are starting from a different position.
Btw., remove the support from Disney for Squeak and let's
see what happens. Somebody pays the show.

> ... so please do not consider it as personal flame.


>
>
> Putting on asbestos underwear anyway...

No reason to do that. A "Framethrower" is much more useful
than a "Flame thrower". ;-)

--
http://www.lavielle.com/~joswig

David Cooper

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
Alexey Goldin wrote:
>
> I afraid they will be put off by attitude of most
> competent people on this newsgroup,...
>

I don't want this to become any kind of a flame exchange,
but I'm genuinely interested in exactly what our attitude
problem is and how we might go about adjusting it.

I sense the same kind of perceived attitude problem out in
the field where I work (ICAD/KBE people vs. the various CAD
empires vs. traditional C/C++ and Oracle IT folks). They
seem unwilling/unable to work with us. Part of the problem is
that we are absolutely coding circles around them as far as
getting any kind of practical applications up and running.

(By the way, ICAD is a KBE system based on Allegro CL,
not to be confused with AutoCAD).

I have been in this business for five years now, and the same
drama plays out again and again. The CAD people raise
objections or ``roadblocks,'' we overcome the ``roadblocks,''
and they just find more roadblocks. It becomes apparent that the
roadblocks are not really roadblocks at all, but just excuses.

We roll applications into production despite them. But at greater
cost and lower volume than if we had more cooperation. This keeps
the cost of the technology high, and makes it difficult to get
out of this rut.

This is not the way it is supposed to work.

Note that we are still surviving, because upper-level management
continues to be blown away by the actual results we produce.
But working with the middle-level managers and engineers is
a constant struggle.

Obviously, some of this has to do with protecting of empires -
a manager of a ``Design-Advisor'' department at a certain
of the Big 2 auto companies here in Detroit has about 45 ``C''
programmers in his sweatshop, producing reams of undocumented
unmaintainable ``AI'' applications which run as ``Design Advisors''
in conjunction with their CAD system. He has Ph.D's writing
C programs which do sophisticated feature-recognition against
the CAD database.

Meanwhile, a half-dozen ICAD programmers turn out Generative
KBE applications which make his stuff look like crude school
projects. The ICAD apps generate geometry from Lisp objects
based on rules, obviating the need for this kind of
feature-recognition because you can just reference-chain to
any attribute of any object in the whole model tree. You know
everything about the geometry because your code generated it.

It's called a labor-intensive vs. capital-intensive approach
to doing this stuff. I thought the U.S. is supposed to be a
capital-intensive economy. I guess no one told this guy.

So anyway, how does all this constitute an attitude problem?

Too close to the situation, I'm sure I am one of the primary
instigators of this attitude problem, but it is very difficult
for me to see it in myself because I miss the forest for the trees.

How to fix the attitude problem, and yet stay true to principles?

Personally, I would not be able to live the lie of not calling a
Spade a Spade.

Craig Brozefsky

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> > Yah, massive peer review, large-scale beta testing,
>
> Sometimes I have the feeling I'm living in a beta world.
> Everything now is beta. When do I get final stuff?
> What once was pre-alpha now is beta.

Well, Debian hamm is not beta, but slink is just frozen and in beta
before release. Linux 2.0.X is not beta, but 2.1.X is. Apache 1.3.X
is no longer beta, but there were several beta versions of Apache.
They do early beta releases, and often have a very large number of
people running their beta software, but they do eventually (in most
cases at least) have full non-beta releases.

> > and use of an
> > educated user base to produce high quality bug reports. It's a
> > stretch to call these revolutionary tho,
>
> Yep, Microsoft, Apple, SUN, etc. are doing the same.

Well, by some stretch of word, yes. They just now are getting hang of
releasing betas to the general public for testing. It used to be that
they charged for them. But now we're just mincing words.

> Just look at CL-HTTP. Only one free Lisp is up to the task
> running it. On one platform. With early support for
> threads.

And I've been running it, in CMUCL, on Linux, with multi-processing.
I've also run it in ACL on Linux. For the most part it was
multithreading that was the major hurdle for free implementations to
get over. That is not in the CL standard tho, so I think it's
reasonable, but definetly not desirable, that some implementations
don't support it. There are alot of other CL packages that run in
free lisps, Loom, Ontolingua(I could have been fantasizing), and
other KR apps, the part of speech tagger from Parc, garnet, and
zebu...


Craig Brozefsky

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> > Some
> > developers are switching to (yikes!) Perl and Python (which is
> > actually nice) which eliminate a lot of problems of C (like memory
> > leaks, bounds checking).
>
> Python also seems to be a great step forward from the
> usual shell scripts.

I have a lame solution for CL scripting that I have cooked up in my
secret labratories. I designed a little protocol for a process to
connect to a long running lisp process and do various things, like
setup an environment for itself, load a file in a batch/script mode,
or run a repl. I have a small C program which acts as the front end,
and in effect gives you a startup time for scripting in under oneq
second, persistent data storage in the long running lisp, and supports
the Unix #! convention for scripts.

The protocol is not lisp specific and needs some work to add better
authentication (right now it's just plain-text token based) and I have
to complete the client-server protocol for configuring environments
for the execution of scripts. I use it for shell scripting work when
I just have to have my CL fix, and once I get the environment passing
done, I will be able to do CGI scripts with it.

The "scripting" model really has advantages. You flush your state
every time (most of the time I hate this, but certain tasks are made
easier by it), it integrates with the rest of the host system fairly
well, and it's procedural style supports one-off programming, quickies
if you will.

It's hardly releasable code, as I've been doing a bit of scheme
hacking lately, but if people reinterested I will post my little
description of the protocol, let everyone pick it to shreds, and then
update my client/server code to handle a goodportion of it and release
the code.

The long running process bit might scare some people off, but I really
like it. I can load all my favorite libraries once, and I can keep
data hanging around in it for quick access if I'm dealing with script
that execute frequently, or I have scripts that needto share data. My
present system is CMUCL (multi-proc version only) but should work with
any multi-processing lisp. I'm debating wether I should have some
sort of harness or other mechanism to ensure the long running process
is always available. I want to have it start on boot, and stay around
forever.

> > (How much do you have to pay for CLIM?)
>
> For example LWW with CLIM seems to be reasonably priced.
> Commercial Unix Lisps are too expensive, IMHO, though.

No kidding. I did some pricing awhile back and found the disparity
between windows versions and unix versions dispicable. This has been
gone over before in this newsgroup tho.

> > and unavailability of meny
> > libraries they are used to in other languages.
>
> And vice versa. I'm using libraries in Lisp that are
> hardly available in other languages or are much
> harder to use in these languages.

Perhaps it might be better to say, the perceived monolithic nature of
Lisp is the real detriment there. At the same time every
implementation I have used has had a FFI that rocks. Python and Perl
have so many library interfaces because they are being used as glue
languages, simplifying access to functionality embedded in other
languages, and libraries. Very little if any of the Lisp programming
I do is like this. Often I am working entirely within the realm of
Lisp's capabilities, and would rather spend a day or so re-writing some
functionality, then dickering about with some library interface and
importing alot of cruft.

I'm not sure there is really away to solve this, given the present
user profile of the lisp community, at least as it appears on this
newsgroup. The lack of a standardized FFI makes it a bit difficult to
be implementation agnostic. Perhaps a few flagship library interfaces
for a particular implementation would be enough to spark some interest
in Lisp for this class of programming.

Louis Glassy

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
Rainer Joswig <jos...@lavielle.com> wrote:
; In article <m3zp8xj...@shodan.demon.co.uk>, Paul Rudin
; <pa...@shodan.demon.co.uk> wrote:

[Linux is more stable than MS Windows]

[Rainer replies]
; whoever compares an OS in terms of stability with Windows


; can't be really helped. I mean anybody with some self

; respect doesn't do that. [i.e. MS Windows is hopeless]

True. :-)

;> [Paul] Which X windows UI?

; [Rainer] Any of the stuff I have seen so far is not overwhelming.

[...]

;> [Paul says: I read news with Gnus/Emacs/X, and commercialware
;> doesn't give similar functionality]

[Rainer replies- ]
; My Symbolics Lisp machine is accessible the same way. Yet, it provides
; a complete environment.

I've heard in various newsgroup conversations that LispM's were/are
the cat's meow [=Oregon dialect "the best"] of Lisp environments.
Unfortunately, I only have access to very conventional
Intel-based PC hardware. Are there any software approximations
to a LispM that run under a free Unixy OS?

What is it in particular that makes LispM's special?

(Note- I'm not doubting that they're neat, I just have never used one,
so don't know why people who like them, do like them.)

Without having actually seen a LispM, I'm guessing they have
superior integration of the various pieces (apps, hardware, OS),
relative to other Lisp-oriented software environments.

Be that as it may (and now the head of the LispOS dragon rears its
ugly head!) - suppose you wanted to develop a nice-to-use
Lisp-friendly environment, subject to the following (possibly
crazy) constraints-

* runs on Intel hardware (say, 486/66 or later, 8+ MB RAM)
* runs as a layer on top of a free Unix (*BSD, Linux)

Which apps would you like to see? Alternatively, should
functionality you would like to use be bundled in terms of
what we now think of as "apps"?

If you do think dividing functionality into "apps" is proper,
then here's (roughly) my list... [not in order of importance]

* a way to read/send/manage email.
* a way to transfer data via ftp (upload & download).
* a way to log in to a remote machine via something like telnet.
* a way to read web stuff (e.g., a web browser).
* a way to manage files, or a better way of managing
persistent data, if you know of one.
* a way to read/write/store/manipulate text
(think "text editor").

and of course, I would like these all to be seamlessly
integrated. Of course.

<<Some people have told me before, Emacs/Guile/X/Scsh/foo already
do all this. Why reinvent the wheel? Well, it's a question
of integration. Of the commercial OS's+app bundles, the Mac
stands out in my mind as an example of a good set of ideas,
although [as Rainer points out] there are cracks in the plaster,
and some of the cracks run deep.>>

I've phrased the list of apps above
as "a way to" and not as "ftp client" etc
because maybe there is a better way to do things than what I'm
used to ("what I'm used to" = "Unix, in whatever flavor").

So my question for you is: what apps would you like to see
in a Lisp-friendly environment built on top of a Unix? and
which would you like to see first? (ie. could you put your
short list in order of importance?)

I'm not a kernel hacker, and while I appreciate
(in some spiritual sense :-)
the idea of a completely Lisp-based OS/VM, I don't
see that implementing LispOS is within my current technical
ability or free-time constraints. But if I had a decent
roadmap, I'd like to try my hand at making some small but
useful pieces of the puzzle-- both as a learning exercise
(any excuse to program in Lisp is a good one) and also in
the hopes of making something generally useful...
Kelly Murray had (has) some ideas on this. Do others?

-Lou

--
Louis Glassy (gla...@cs.montana.edu)
Department of Computer Science
Montana State University
Bozeman, Montana 59717 USA

Steve Gonedes

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to

Joachim Achtzehnter <joa...@kraut.bc.ca> writes:

< And from a user's perspective, you have to weigh the alternatives: put
< all eggs in one basket and accept that you depend totally on one

< vendor to fix things or extend things? Or have the freedom to hire any
< professional software developer to work on freely available source if

^^^^^^^^^^^^ ^^^^^^^^ ^^^^^^^^^

No such thing.

Matthias Buelow

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <87vhjkh...@piracy.red-bean.com>
Craig Brozefsky <cr...@onshore.com> writes:

>> Maybe gradually - but not in principle. Is there
>> any revolutionary coding practice in effect that
>> suddenly makes Linux software more reliable?
>

>Yah, massive peer review, large-scale beta testing, and use of an


>educated user base to produce high quality bug reports. It's a

Peer review? Isn't it with linux that a rather dumb mass cheers
and praises every fart that comes from a couple of self-glorifying
"gurus" (even when the same old boring wheel gets reinvented poorly
over and over again)? Educated user base producing high quality
bug reports? That's really a minority, imho.

--
Matthias K. Buelow * Boycott Micro$oft, see http://www.vcnet.com/bms/ *

paul_...@scientia.com

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <joswig-1012...@194.163.195.67>,

jos...@lavielle.com (Rainer Joswig) wrote:
> In article <m3zp8xj...@shodan.demon.co.uk>, Paul Rudin
> <pa...@shodan.demon.co.uk> wrote:
>
> > has crashed once. OTOH rebooting a windows box because the OS has
> > fallen over happens every couple of days.
>
> whoever compares an OS in terms of stability with Windows
> can't be really helped. I mean anybody with some self
> respect doesn't do that. "Hey my Hyundai is faster
> than your Trabant." Still it doesn't give you a Mercedes
> (or whatever these cars will be called now. ;-) ).

well, OK, but the point is that Windows is commercial software and Linux is
free and in many repsects superior.

>
> > Which X windows UI?


>
> Any of the stuff I have seen so far is not overwhelming.

Don't be shy, name names :-)

> My Symbolics Lisp machine is accessible the same way. Yet, it provides
> a complete environment.
>

Sounds good, what does such a device cost?

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own

Tim Bradshaw

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
* Mike McDonald wrote:

> But for all practical purposes, there is NO 18B release of CMUCL.

?. My CMUCL is 18b, and I'm on a sparc/Solaris not intel/Linux box,
and I didn't have to work that hard to find it (I think I just had a
look on cons.org, noticed a new one and installed it). It works.

--tim

Alexey Goldin

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> In article <ucyaog7...@ns.mercury.bc.ca>, Joachim Achtzehnter
> <joa...@kraut.bc.ca> wrote:
> >

> > And from a user's perspective, you have to weigh the alternatives: put
> > all eggs in one basket and accept that you depend totally on one
> > vendor to fix things or extend things?
>

> Or a bunch of vendors competing with each other?

Where many is two? One of whom is "changing focus?" (MCL seems to be
nice --- I've seen demos, but available for Mac only)


>
> No. It would be interesting to explore the idea. Netscape
> seems to see it as a way to survive as a company
> (note that they are only releasing the browser software,
> not the servers). How do you think would an "Open Source"
> approach would look like for a Lisp vendor?
>
> Rainer Joswig

It would be interesting to see what happens with Digital Creations
experiment who released their flagship product as open source last
friday (see http://www.digicool.com, http://www.zope.org). They plan
to charge for consulting as evidently no one knows its internals
better then them.

If Symbolics decided to go this way Stallman would finally feel
vindicated.

Gareth McCaughan

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
Erik Naggum wrote:

> lots of things are here to stay. Hinduism, for instance. biggest
> religion there is. have you converted, yet? if not, why not?

<pedant>
Obviously this doesn't make a scrap of difference to your point;
but both Christianity and Islam are larger than Hinduism according
to the estimates I've seen. Of course this depends on exactly
what you consider makes someone a Christian/Muslim/Hindu.
</pedant>

--
Gareth McCaughan Dept. of Pure Mathematics & Mathematical Statistics,
gj...@dpmms.cam.ac.uk Cambridge University, England.

Lars Lundback

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
I would like to make my small contribution to this - renamed - topic.
The application you (David Cooper) have chosen (CAD) is perfect; it has
all the ingredients to illuminate why Lisp does not make a hit. It is
sad, because Lisp would be the perfect implementation language for eg.
CAD systems.

David Cooper wrote:
>
>
> I don't want this to become any kind of a flame exchange,
> but I'm genuinely interested in exactly what our attitude
> problem is and how we might go about adjusting it.
>
> I sense the same kind of perceived attitude problem out in
> the field where I work (ICAD/KBE people vs. the various CAD
> empires vs. traditional C/C++ and Oracle IT folks). They
> seem unwilling/unable to work with us. Part of the problem is
> that we are absolutely coding circles around them as far as
> getting any kind of practical applications up and running.

I started working in the same field about 25 years ago, and was most
apprehensive when ICAD was introduced. But we found, to my dismay, that
the ICAD approach didn't work, even in the resarch we were conducting
then.

What you say about coding is true for most of the simpler pilot
applications. But part of the overall problem is to make the
user/designer to "think" before designing. ICAD/KBE requires, as I
recall, a very structured design approach, whereas many designers prefer
the draw-something/see-what-you-get loop.

> (By the way, ICAD is a KBE system based on Allegro CL,
> not to be confused with AutoCAD).

Yes, and AutoCAD is fascinating. It doesn't seem that AutoCAD uses
AutoLisp to implement a UI that is powerful, yet easy to understand. Or
if they do, the resulting UI is cumbersome, at least to me.

I note that you say "ICAD programmers", not "designers using ICAD". It
illustrates my point, which is that two cultures fail to meet. Add to
that the fact that large companies have invested heavily in product
designs, using current "standard" CAD systems, and you begin to see the
difficulties in turning this gigantic ship.

> It's called a labor-intensive vs. capital-intensive approach
> to doing this stuff. I thought the U.S. is supposed to be a
> capital-intensive economy. I guess no one told this guy.
>
> So anyway, how does all this constitute an attitude problem?
>
> Too close to the situation, I'm sure I am one of the primary
> instigators of this attitude problem, but it is very difficult
> for me to see it in myself because I miss the forest for the trees.
>
> How to fix the attitude problem, and yet stay true to principles?
>
> Personally, I would not be able to live the lie of not calling a
> Spade a Spade.

The attitude problem?

Well, I mentioned in an earlier posting to this group that I have been
sniffing around Lisp the last 25 eyars. In that time, I have always
worked in either academic or "industrial development" circles. Now, the
"Worse is better" paper by Gabriels identifies what was known to anyone
in my field. I guess you could summarize it with a question:

- Which is the most important to you: Being in the right,
or being the winner?

Bad Boy Bill is certainly the winner, so far. The Lisp community is
certainly losing, so far. Could it be that the attitude problem rests
with the Lisp community, not with the "others"?

Don't get me wrong. I see excellent postings to this group, from people
with deep knowledge in Lisp, or being successful at providing Lisp based
solutions, or good at teaching problem solving with Lisp.

The attitude problem is, I think, that we do not want to be compared to
the "others", since we *know* we are right. But the "others" provide
Java toolkits and numerous small applications that can be learned from,
HTML editors, simple graphic tools, free C compilers, and fairly
standard libraries through which your application can make use of all
those wonderful I/O devices. A would-be Lisper does not see things like
that.

Or rather, he begins to see it, since free, ready-to-use, Common Lisps
with at least integrated screen graphic drawing functions are coming
out. No offence to the free academic Lisps is intended, but they were
never designed for easy use.

And the need for comprehensible Lisp tutorials and related documents
cannot be overstated. I recently found a good "Introduction to
Programming in Java" at my local, very small communal library. To find
anything in Lisp, I would have to look long and hard indeed.

This is the attitude again. Lispers write papers, Java people pour out
the small applications that attract.


Peter Van Eynde

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
On 10 Dec 1998 09:53:08 -0500, Sam Steingold wrote:
> >> ?. My CMUCL is 18b, and I'm on a sparc/Solaris not intel/Linux box,
> >> and I didn't have to work that hard to find it (I think I just had a
> >> look on cons.org, noticed a new one and installed it). It works.
>
>The latest Linux binaries are 6 months old, and no good for compiling
>18b (I tried recently). Therefore, until someone makes linux binaries
>available, "for all practical purposes, there is NO 18B release of
>CMUCL" for Linux.

There is a "no 18B"[1] release for Linux available: 2.4.7 for glibc2-based
systems. It was released on the 9th of November. Download it. Download
the source[2]. Try out the automatic anon-cvs updates of the source. Watch
it recompile itself _automaticly_. Try not to notice the huge
number of patches that are still not integrated in the cvs tree, nor
the bug in clm I found yesterday. A new release (after integrating the
patches into the main cvs tree) for glibc2 and libc5 systems should
happen hopefully before the end of this year...

[1] It's not 18b because the format of 17x and 18x releases are not adapted to
modern Linux packaging methods. And because I don't want to wait forever
before generating a package, so I just call my packages 2.4.7 and 2.4.8 etc.
If you want 18B, do a "cvs -r 18B_RELEASE checkout" or something like that.
Don't expect the result to work. :-(
[2] From your nearest debian mirror, see http://www.debian.org. It's in the
current "frozen" distribution.

Groetjes, Peter

--
It's logic Jim, but not as we know it. | pvan...@debian.org for pleasure,
"God, root, what is difference?",Pitr | pvan...@inthan.be for more pleasure!

rrk137

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
Hi. I have been wanting to begin to learn Lisp programming for a while now
and am now at the point of wanting to begin. I have heard that you need
(perhaps old news?) a Lisp special purpose machine. Has that changed? I have
a few books, but in not actually writing and seeing how one "interacts" it's
a bit dry.
I thank anyone now for their help. Thank you.

rick kelsch

Gareth McCaughan

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
Rick Kelsch wrote:

> Hi. I have been wanting to begin to learn Lisp programming for a while now
> and am now at the point of wanting to begin. I have heard that you need
> (perhaps old news?) a Lisp special purpose machine. Has that changed? I have
> a few books, but in not actually writing and seeing how one "interacts" it's
> a bit dry.

I don't believe there has ever been a time when you needed
a special purpose Lisp machine to run Lisp. There are Lisp
systems of decent quality, both free and commercial, available
for Windows, Linux, Solaris, FreeBSD, MacOS, and probably
N other operating systems I haven't thought of.

If you're running Linux, general opinion is that Allegro
Common Lisp 5.0 for Linux (which is available for free)
is the best around. Some people are getting better performance
for some applications out of CMU Common Lisp (which is not
only available for free, but also comes with complete source
code if you want it -- which you probably don't yet!).

If you're running Windows, Allegro have a free product
(Allegro CL Lite 5.0 for Windows) but it has some restrictions;
Harlequin also do a free product (LispWorks for Windows,
Personal Edition), also with some restrictions. CMU CL
doesn't run under Windows. You might also want to consider
"Corman Lisp", a recently released implementation that's
available for free, with full source code. I think its compiler
isn't all that good, but it allegedly gives good access to
the OS and the hardware.

If you're running MacOS, I don't think there is anything that's
both good and free; but Digitool's "Macintosh Common Lisp"
is by all accounts a very impressive product. There's a demo
version, whose only limitation (but it's a big one) is that
you can't use either it or any application built with it for
more than 15 minutes at a stretch; or (if you get a different
demo version) that it will stop working 4 weeks after you get
it.

If you're running Solaris, CMU CL will run on your machine.
I'm not sure what commercial Lisps (if any) will.

If you're running FreeBSD, CMU CL will run on your machine.
It might be possible to get ACL5.0 to run, too. (The last
I heard -- some time ago -- there was some difficulty because
FreeBSD's Linux emulation doesn't emulate everything.)

On many of those systems (and several others), you can also run
CLISP, which is very economical with memory but compiles only to
an interpreted byte-code (a bit like Java does), not to native
code. It used to be quite severely divergent from the ANSI
standard for Common Lisp, but I think it's getting better.
CLISP is freely available, with complete source code.

There are other dialects of Lisp other than Common Lisp. The
only one you're likely to care about is Scheme. There are a number
of good Scheme implementations for common OSes too, both free and
commercial.

Useful URLs:

Franz (Allegro Common Lisp) http://www.franz.com/
Harlequin (LispWorks) http://www.harlequin.com/
CMU Common Lisp http://www.cons.org/cmucl/
Digitool (Macintosh Common Lisp) http://www.digitool.com/
CLISP http://clisp.cons.org/

Joachim Achtzehnter

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
"rrk137" <rrk...@One.Net> writes:
>
> I have been wanting to begin to learn Lisp programming for a while
> now and am now at the point of wanting to begin. I have heard that
> you need (perhaps old news?) a Lisp special purpose machine.

No, you don't! You can use Lisp on a PC, a Mac, a Unix system, a Palm
Pilot, and who knows what else. You can pay several thousand dollars
for a Lisp environment or you can get one for free. Read the Lisp FAQ
for details about what's available. A good place to start is this URL:

http://http.cs.berkeley.edu/~russell/prog.html

Joachim Achtzehnter

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
jos...@lavielle.com (Rainer Joswig) writes:

> Joachim Achtzehnter wrote:
>
> > would it have been easier or more difficult for Linus, Alan
> > Cox, and a handful of others to build (and maintain!) something
> > like Linux using a proprietary approach?
>
> Aren't/weren't they mostly "copying" what is/was already out
> there grounded on massive experience and preexisting tools?

And isn't this by far the most important of all good engineering
practises? If you want to build a robust system for the real world, as
opposed to exploring new approaches in a research mode, isn't it
almost essential to ground the system on "massive experience and
preexisting tools"?

> This seems to be easier than to develop something completely new.

Yes, and everybody has the option to take this approach, this has
again nothing to do with proprietary versus free. Linus, while still a
student, chose to make use of existing "massive experience" while a
well-known corporation decided to throw away all this knowledge and
re-invent the wheel to produce a "marvel" like NT! :-)

> > And from a user's perspective, you have to weigh the alternatives:
> > put all eggs in one basket and accept that you depend totally on
> > one vendor to fix things or extend things?
>
> Or a bunch of vendors competing with each other?

Yes, competition helps, but sometimes it is not sufficient. About
three years ago we ran into such a problem. We were embarking on a new
project that required tight integration of software developed with
different tools. After some study we decided that CORBA was best
suited to provide this glue. Unfortunately, there was a small problem:
An important component was written in Lisp and none of the "competing"
CORBA vendors supported Lisp. We approached this issue fully intent on
using only "commercially supported" software, but quickly changed our
mind. Not only was it hopeless to expect that a vendor would help us
create the interface to Lisp, the quality of commercial products we
evaluated was often a joke. We now use a free CORBA implementation,
have contributed some effort to get a Lisp mapping working, and are
mightily impressed about the quality of this software.

> > Is it so clearcut whether they would be any worse off if they
> > followed the free software model?
>

> No. It would be interesting to explore the idea. How do you think


> would an "Open Source" approach would look like for a Lisp vendor?

Well, I don't claim to have all the answers. In fact, I work for a
company that develops proprietary software ;-(, the CORBA work we
contributed was an exception. But, let us look at these two
companies. How much of their revenue is from selling shrink-wrapped
boxes? How much is from support and consulting? How many of their
innovations are the result of specific projects funded by outside
sources? I can't be more specific having been expeosed to some inside
information, but I can see them use the free software model and still
survive nicely.

Mike McDonald

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <m3r9u8c...@eho.eaglets.com>,
Sam Steingold <s...@goems.com> writes:
>>>>> In message <ey3emq8...@todday.aiai.ed.ac.uk>
>>>>> On the subject of "Re: help! absolute beginner"
>>>>> Sent on 10 Dec 1998 11:51:44 +0000

>>>>> Honorable Tim Bradshaw <t...@aiai.ed.ac.uk> writes:
> >> * Mike McDonald wrote:
> >>
> >> > But for all practical purposes, there is NO 18B release of CMUCL.

> >>
> >> ?. My CMUCL is 18b, and I'm on a sparc/Solaris not intel/Linux box,
> >> and I didn't have to work that hard to find it (I think I just had a
> >> look on cons.org, noticed a new one and installed it). It works.
>
> The latest Linux binaries are 6 months old, and no good for compiling
> 18b (I tried recently). Therefore, until someone makes linux binaries
> available, "for all practical purposes, there is NO 18B release of
> CMUCL" for Linux.
>

Add HP-UX, Irix, and OSF to the list of unavailable CMUCL 18B ports. Only
Solaris and FreeBSD are available.

Mike McDonald
mik...@mikemac.com

Mike McDonald

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <74nr63$kbi$1...@netra.msu.montana.edu>,
Louis Glassy <gla...@acheta.nervana.montana.edu> writes:
> Rainer Joswig <jos...@lavielle.com> wrote:

> [Rainer replies- ]
> ; My Symbolics Lisp machine is accessible the same way. Yet, it provides
> ; a complete environment.
>

> I've heard in various newsgroup conversations that LispM's were/are
> the cat's meow [=Oregon dialect "the best"] of Lisp environments.

> What is it in particular that makes LispM's special?

Well, if you were still in Oregon, I'd invite you over to take a look at
mine. It's hard to describe how easy LispM's are to use.

> So my question for you is: what apps would you like to see
> in a Lisp-friendly environment built on top of a Unix? and
> which would you like to see first?

Well, obviously, the first should be Erik's CL Emacs. :-) You need an editor
so you can use it to write all of those other apps. Now, the problem arises as
to what you'd use for a windowing system? I don't think you'd want to do it
based on ASCII terminals. That probably means X in some form or another. But
what toolkit? CLIM? Garnet? CLX? All of these have pros and cons.

Next, I'd like to see a CLIM based lisp listener. Then a mail reader and a
news reader. I'd need them both at about the same time. Then a web browser. At
some point, you'd need to thru in the networking support layer so you can
write clients and servers. Then I'd like to see Mahjongg (xmj) ported. :-)

Mike McDonald
mik...@mikemac.com

Erik Naggum

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
* jos...@lavielle.com (Rainer Joswig)

| Aren't/weren't they mostly "copying" what is/was already out there
| grounded on massive experience and preexisting tools? This seems to be

| easier than to develop something completely new.

well, this is a double-edged argument. first, it's positive, because the
GNU/Linux system is becoming a free and very high quality POSIX-compliant
system -- which makes perfect commercial sense, as there is a lot of user
experience and market and mind share there already. second, POSUX isn't
all that great to begin with, so something much better would certainly
have been beneficial in the abstract sense, but how do you get commercial
applications for it, and thus attract end users? it's hard enough as it
is to get vendors to support non-mass-market systems, although Linux is
making important inroads into the managerial mind-set. with the rabidly
anti-commercial attitudes the "Free Software" movement has recently
adopted, it would have been nothing more than a magnificient waste of
time and effort at best.

I don't think another conforming implementation of ANSI Common Lisp would
merit your denigratory comments. the world needs rational critics of the
irrational elements in the free software movement, not irrational critics
of the irrational elements.

#:Erik

Raymond Toy

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
>>>>> "Mike" == Mike McDonald <mik...@mikemac.com> writes:


Mike> Add HP-UX, Irix, and OSF to the list of unavailable CMUCL 18B ports. Only
Mike> Solaris and FreeBSD are available.

Perhaps this is mostly because it's very hard to go from 18a to 18b.
And perhaps because the people working on the x86 ports don't have
easy access to these other machines.

The way things are going I wouldn't be surprised if FreeBSD and/or
Linux will be the only ports making any progress. A sad day.

Ray

--
Why is that we have "cultural diversity" training and sensitivity
training, and second sourcing, but that doesn't apply to computers and
software?

Georg Bauer

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <joswig-0912...@pbg3.lavielle.com>,
jos...@lavielle.com (Rainer Joswig) wrote:

>Another religious marketing bullshit.

Seems it is useless to discuss such a topic with you. Too bad.

bye, Georg

--
http://www.westfalen.de/hugo/

Georg Bauer

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
In article <ey3lnkh...@haystack.aiai.ed.ac.uk>, Tim Bradshaw
<t...@aiai.ed.ac.uk> wrote:

>And, actually, which commercial Lisps *don't* have CLIM?

Eclipse? Golden Common Lisp? Lucid Common Lisp (although I am not sure
about that one)? Actually there is Allegro with a CLIM that is available
for quite some time, MCL with something that might be a full commercial
product some time soon and Harlequin with LispWorks (although the CLIM for
Windows is still beta, as I understand). And for Rainer: yes, I know that
Symbolics has a CLIM.

The "market-leaders" deliver CLIM - _now_. Look about 2 years back and you
will find CLIM only for Unix. I still remember the rant about a missing
CLIM on the MCL-wish-page at Digitool. It was written by Rainer, IIRC.

Sure, it is getting better. Nothing said against that. But if commercial
support would be really that good for the customer, why wasn't there a
CLIM 2 years ago? Actually CLIM isn't that new on the market, isn't it?
It's not as if I say that all commercial products are crap. Only that you
have to look at what it is that you get and what it is that drives
development - and if you find yourself represented in that development.

There is much crap out there and for a lot of it you have to pay for. And
commercial support is often overrated. Of course, there are situations
where the dynamics of OSS is overrated, too. But OSS definitely doesn't
deserve this bashing it received from Rainer. That's why I talked about
ideologic blindfolding ... (Klaus Schilling is the exact counterexample -
ideologic blindfolding from the OSS side)

Actually people often forget what motivated many OSS projects: pragmatism
and pure joy in hacking and sharing. Both are quite good motivations to
get some complex task done. And there are situations where those projects
are more productive than commercial products - just because a company only
can put so much workpower on some job. In some OSS projects, developer
resources are almost unlimited - because people are interested in it and
find fun in working on it. Or are working on it to make it suit some
problem they have (many solutions for Apache were created out of exactly
this reason) - that's what I call pragmatism. Many of the solutions made
in the OSS projects had been impossible if not for the availability of the
sources and the openess of the project.

I think it is a bit shortsighted to put all this under "just products from
geeks for geeks" or disregard it as marketing bullshit. This totally
ignores the existence of high-quality OSS projects that outrule commercial
counterexamples.

To get the turn to Lisp: I don't think that there is a sense in turning
your back to OSS lisps, just because parts of them are not up-to-par with
commercial implementations. Better increase the hacking value of them, so
people get interested in it and deliver the missing parts. Look at what
CLHTTPD achieved. It is a OSS project. Look at scsh - a quite nice system.
Look at guile-scsh, and you geht shorter startup times (the biggest
problem with scsh in my book), although parts of the library are still
missing.

Kelly Murray

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
Georg Bauer wrote:
>
> In article <ey3lnkh...@haystack.aiai.ed.ac.uk>, Tim Bradshaw
> <t...@aiai.ed.ac.uk> wrote:
>
> >And, actually, which commercial Lisps *don't* have CLIM?
>

CLIM was a dismal failure, and in many ways
represents what is wrong with lisp, not the least of which is
CLIM proponents can't acknowledge it was a failure and move on,
and moreover let the short-term-customer-driven vendors move on too.
-k

David Cooper

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
Well, this is getting a bit off-topic now, and I think
the original thread seems to have just about died out,
but I would like to comment on a couple of Lars' points
just for informational purposes.

Lars Lundback wrote:

>
> ... But part of the overall problem is to make the


> user/designer to "think" before designing. ICAD/KBE requires, as I
> recall, a very structured design approach, whereas many designers prefer
> the draw-something/see-what-you-get loop.
>

You're really talking about a couple issues here. Most companies
(one notably exception I believe is Boeing) at least partially
separate ICAD/KBE developers from end-users. Developers have
to take a somewhat structured approach to developing applications,
but all the oil-painting qualities of Lisp (ref: Paul Graham's intro
to ANSI Common Lisp) apply, to an even greater extent than in pure
Lisp. So there is a *very* dynamic ``code-something/see-what-you-get''
loop during development, which makes ICAD the best prototyping
environment I have ever seen.

The end-users of applications are typically engineers, not designers.
Engineers, at least in the Big 2, are not CAD jockeys, and are separate
job functions from CAD designers. In many cases the CAD designers are
like high-paid secretaries to the engineers (though the designers
often end up knowing a heck of a lot more engineering than the
engineers,
and the engineers spend all their time on the phone chasing down
suppliers, prototypes, etc..

Anyway, if engineers can avoid messing around in a CAD system they
generally will. A good ICAD application will let them get in, specify
some requirements, and get some hard answers out, with a minimum of
time and fuss.

So, the best kinds of problems for ICAD are well-defined (but
boring and repetitive) *engineering* problems (which may or may
not be CAD or geometry-related) -- *not* traditional CAD Design
problems with a lot of creative and artistic content, which,
I think, is what you are referring to in your comment.

This is a common misconception. People often mistake engineering
problems for CAD problems, including, probably, the sweatshop
proprietor I mentioned in my earlier post.

>
> I note that you say "ICAD programmers", not "designers using ICAD". It
> illustrates my point, which is that two cultures fail to meet.
>

Well, in that comparison I was comparing ICAD programmers to C
programmers, not CAD designers. The ICAD application end-users
are a different story; in general they are a quite happy bunch,
(at least users of applications I am involved with :)
and they are generally engineers rather than designers. People
come into ICAD development from a lot of different walks of life,
and in my experience ex-CAD designers stand no better or worse a
chance of becoming effective ICAD developers than any other
well-educated folk.

>
> Add to that the fact that large companies have invested heavily in
> product designs, using current "standard" CAD systems, and you begin
> to see the difficulties in turning this gigantic ship.
>

Ah, yes, but it will turn if you find the proper controls
and know how to operate them effectively! Some of these companies
are starting to wake up and figure out that CAD-the-Master
is really an emperor with no clothes (you can draw pretty
pictures and give nice demos, but it does not go very deep
in solving the *engineering* problems which are really the
core thing). Knowledge-the-Master (which is really just a hokey
pretentious way of saying ``capturing product and process definitions
in a high-level dynamic object-oriented declarative language'')
is indeed gaining a foothold.


Hopefully we will get that ship to turn several degrees before it
hits the iceberg!

>
> - Which is the most important to you: Being in the right,
> or being the winner?
>

Um, being in the right. And being the winner. I simply do not
accept that compromise (as much as I think Richard is a very cool
guy).

I understand that a lot of the early Lispers might be a bit weary
of the struggle. But we of the new generation do not have the
same license. We are standing on the shoulders of these giants,
and it is up to us to carry the flag with a renewed energy and
spirit.

I expect most religions start out this way. The pioneers
are seen as members of strange and different cult, and become
weary at the struggle against the mainstream. But after they
have paved the path, the newcomers can come in with renewed
energy and without all the baggage, and virtually stand on the
shoulders of the early giants to achieve great things.

>
> Bad Boy Bill is certainly the winner, so far. The Lisp community is
> certainly losing, so far. Could it be that the attitude problem rests
> with the Lisp community, not with the "others"?
>

Hmm.. that's kind of hard to see from my little enclave over here.
The only evidence I get of that is from this newsgroup. I am,
however, a strong believer in self-fulfilling prophecies for
those who make them.

I have been in the midst of an ever-increasing Lisp activity
here in Detroit for the past five years, and the company I
started over one year ago has never used a Microsoft product.

Call me a radical, I guess. But personally I feel pretty normal.

By the way, how is it that if Bill wins, Lisp loses, and
vice-versa, which seems to be the implication of your comment
here. I just found that a bit puzzling.

Ok, I think the ramble detector is going off... good night.

Erik Naggum

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
* Gareth McCaughan <gj...@dpmms.cam.ac.uk>

| <pedant>
| Obviously this doesn't make a scrap of difference to your point; but both
| Christianity and Islam are larger than Hinduism according to the
| estimates I've seen. Of course this depends on exactly what you consider
| makes someone a Christian/Muslim/Hindu.
| </pedant>

I'm rather unwilling to call people believers in the same religion when
they murder each other for their differences in belief -- it's a good way
to become a target they can agree on. this reduces Christianity to a
huge number of different religions. I think they want it that way, and I
am in no position to argue. ditto for Islam. now, Hinduism is the name
of a huge collective of religions, but at least they aren't at war over
it. that's gotta count for something.

#:Erik, who thinks that if some religion were actually true, it'd all be over

Erik Naggum

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
* g...@hugo.westfalen.de (Georg Bauer)
| But if commercial support would be really that good for the customer, why

| wasn't there a CLIM 2 years ago?

and if free software and open source was so great, why didn't you start
in 1880? think of the greatness that could have been achieved if you
guys had gotten your act together a 100 years sooner. you inconsiderate
bums! just think of all you have _deprived_ us of by waiting 100 years,
and for _what_ good reason, huh? humans have waited over _40,000_ years
for a free CLIM implementation, and you just don't want to give it to us,
do you? death to free software, I say! closed source now!

#:Erik

Sudhir Shenoy

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
In article <86n24w8...@g.pet.cam.ac.uk>, Gareth McCaughan
<gj...@dpmms.cam.ac.uk> wrote:

> Erik Naggum wrote:
>
> > lots of things are here to stay. Hinduism, for instance. biggest
> > religion there is. have you converted, yet? if not, why not?
>

> <pedant>
> Obviously this doesn't make a scrap of difference to your point;
> but both Christianity and Islam are larger than Hinduism according
> to the estimates I've seen. Of course this depends on exactly
> what you consider makes someone a Christian/Muslim/Hindu.
> </pedant>

To add some further pedantry - you cannot actually convert to Hinduism.
Unlike the Semitic religions, it does not expand by proselytisation. Secondly,
it is not a monolithic religion (there is no single religious text, god,
pope etc.)
Since some atheists are Hindus (really!), the actual number of any particular
type of Hindus classified by each ones religious beliefs is probably very small
indeed.

Sorry for the off-topic post but I couldn't resist (since I seem to be one of
the only Hindus reading this ng).

Sudhir
ssh...@gol.com

David Cooper

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
By the way, how many ICAD people are lurking out there?

I really would like to get a comp.lang.lisp.icad (or, using the new
name, comp.lang.lisp.kbo) group going. Pipe dream? Perhaps, but
stranger things have happened on this wondrous spinning sphere.

I think there are a couple hundred active developers worldwide, so I
would probably have to reach about half the active developers to
vote for a newsgroup (you need 100 votes, right?).

Any sour-grapes expatriots would be welcome too, to keep us from
forgetting what idiots we are for still using ICAD.

If I get more than a few responses from this newgroup, I will collect
them and proceed to try gathering people from the IIUG user's group
and from KTI themselves. (any email addresses I get in response to
this post will be used for the purpose of initiating the newsgroup
only).

(hmmm... I'm not really sure what I'm biting off here, am I?)

Rob Warnock

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
Espen Vestre <e...@nextel.no> wrote:
+---------------
| > I think Bill Gates is here to stay, guess we should try to adjust.
|
| I don't know, Win2000 seems to be such a monster that they *must* be
| programming themselves into a corner of an enormous room in a house
| they don't longer know how they even got into...
+---------------

[Yes, I know it's continuing an off-topic thread, but...]

Making the rounds lately:

Did you hear that Windows 2000 slipped again? Yeah,
it's not gonna be out until the *second* half of 1901...


-Rob

-----
Rob Warnock, 8L-855 rp...@sgi.com
Applied Networking http://reality.sgi.com/rpw3/
Silicon Graphics, Inc. Phone: 650-933-1673
2011 N. Shoreline Blvd. FAX: 650-964-0811
Mountain View, CA 94043 PP-ASEL-IA

Erik Naggum

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
* rp...@rigden.engr.sgi.com (Rob Warnock)

| Making the rounds lately:
|
| Did you hear that Windows 2000 slipped again? Yeah,
| it's not gonna be out until the *second* half of 1901...

I heard it was just going to be called Windows 1900.

#:Erik, who looks forward to the naughties.

It is loading more messages.
0 new messages