Lisp Machines vs. Commercial Lisps ???
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
for there Lisp Environment when you can get a Lispm for around 1K.
BTW, the Lispm blows most Commercial Lisps out of the water in terms of
features.
Allegro CL $3K with student discount. (The ed. that makes exe. files under
windows.) no wonder the price of Allegro CL is not on there web site.
KnowlegeWorks $2.5K Commercial
LispWorks $950 Commercial
GoldWorks III < $2K educational.
MCL around $650
PS
Why is Allegro so pricey?
PPS
The only good deal besides free Lisps is Corman Lisp about
$125 student price.
I refuse to pay more for a compiler than for a used car.
Another benefit of buying a Lispm vs. a Commercial Lisp is this--the price
of the Lispm includes the computer.
Expect a computer for Lisp to cost <$2,000 if you want a good one.
or around $500 if you want an E-Machine.
Note: this is not to invite flames but to point people who'd like to learn
Lisp and are not made of $$$ in the right direction.
If I had enough $$$ I prob. would buy Allegro--but I am a student and don't
have alot of spare $$$ to feed my hobby.
One sad fact is this: the price for used Lispms is so reasonable because
Lispm companys priced themselves out of existance in the first place--if
they provides 1-2K machines they'd sell a lot to hobbyists. Esp. since Lisp
Compilers for Windoz are so costly.
The Lisp Machine is ANSI CL, and includes all source code--so you can hack
Lisp to your hearts content.
Tim Bradshaw
[Drivel. Endless repetitions of this kind of thing over the years are
why so many people on cll seem so unfriendly. The two rational
responses I can see are to completely ignore the fools who produce
them, or to be so unpleasant to them that they go away for ever. I'm
failing to do either here, I'm afraid.]
> BTW, the Lispm blows most Commercial Lisps out of the water in terms of
> features.
Like `running natively on all major platforms', `delivering
executables on all major platforms', `supporting 64bit platforms',
`being actively supported', `competing in performance with C on
current systems, `supporting CORBA', and a long and growing list of
other features.
> The Lisp Machine is ANSI CL, and includes all source code--so you
> can hack Lisp to your hearts content.
No, it's not (assuming you mean Genera 8.x). It's close but not as
close as most of the current commercial CLs are.
--tim (who used a LispM before you knew what Lisp was I expect, and
has undoubtedly owned more of them than you)
Kent M Pitman
@ hotmail . com> writes:
> How can Commercial Lisp Vendors expect people to pay 2K or more
Because people will pay that. It doesn't have to be you.
Stop complaining about people making a profit. That's a good thing.
You don't have to buy from them if you don't like.
> for there Lisp Environment when you can get a Lispm for around 1K.
Well, depending on the model you get, the amount of direct power (in
elec) and indirect power (in elec for A/C) you use may drive the price
up after-the-fact in hidden ways.
And there are a finite number of such machines, so for many applications,
especially those that must be deployed on standard hardware, this is
not an acceptable solution.
Also, Lispms are not continually improving in speed. Some people like
to get a new processor every 6 months or a year that is faster. Even
if you buy the software emulation on the Alpha, that's no longer a
maintained piece of hardware. (Also, I think obtaining a legal
software license costs you, in addition to obtaining the hardware to
run it--or that may be only for new hardware, I'm not sure.)
These are not criticisms of the LispM environment per se, but since the
only sources of LispM environments are presently not actively producing
a new stream of releases on modern hardware, there are some definite
limitations to this approach.
I certainly still use my LispM (Macivory Model 3) from time to time...
but not daily.
> Why is Allegro so pricey?
I have no special knowledge of this, but I suspect the reason is that
they do not want to make their money on Lisp sales, but rather on Lisp
support. Implicitly, by screening out people who think that this is
"pricey" they are eliminating people who don't have money for support,
and streamlining their commercial dialog to be established companies
with adequate cash flow to both be able to buy support and appreciate
its need. From the tone of the your message, it sounds to me like if
this is their plan, it's working.
(Note that phone companies do the same thing with charging more for
daytime calls, and airlines with charging more for people who book at
the last minute. People protest that they won't do these things, and
they make their calls at night or their trips with lots of planning,
but businesses often can't afford to and end up paying the high
prices.)
The real question is, if you are comfortable with the other offerings,
why do you care about the price of a particular implementation that
comes from a vendor that seems to be economically successful? They
must know what they are doing because people continue to buy them,
perhaps because they offer a product of sufficient quality that it
commands the price they ask in enough cases to make it worth their
while.
> I refuse to pay more for a compiler than for a used car.
So don't. An offering that is not something a person in the market
wants to afford is not a sign of a sick market. I refuse to pay more
for a car than the cost of a car, but that doesn't keep Mercedes-Benz,
Porsche, etc. from making really expensive cars, nor should it, it
just means everyone has a different notion of what "car"...
> One sad fact is this: the price for used Lispms is so reasonable
> because Lispm companys priced themselves out of existance in the
> first place--if they provides 1-2K machines they'd sell a lot to
> hobbyists. Esp. since Lisp Compilers for Windoz are so costly.
There are no new ones being made. They can't make them at these
costs. They couldn't even make them at this cost when they were being
made. They cost about $30K at the time the company was going out of
business, and even then they knew the prices needed to about $5K to
satisfy the customer base, they just couldn't reach that price point
and break even.
It's well-known in the computer industry (and most industries) that if
you drop prices, you get a larger base. The problem is that dropping
prices drops price-per-box and you have to survive on less income
until the larger base kicks in. Lots of time, the money just isn't
there. Forget the LispM, look at Apple--nearly everything for the Mac
is more expensive than one would wish, almost surely because they
can't afford to drop the price enough to get the volume that would
justify the drop--and as a consequence of not dropping the price, they
can't get the market either. Catch-22.
Pascal Costanza
"Franz Kafka" <Symbolics _ XL1201 _ Sebek _ Budo _ Kafka @ hotmail .
> How can Commercial Lisp Vendors expect people to pay 2K or more
> for there Lisp Environment when you can get a Lispm for around 1K.
They are still in business, so obviously they don't need to expect it,
they just get it.
> Allegro CL $3K with student discount. (The ed. that makes exe. files under
> windows.) no wonder the price of Allegro CL is not on there web site.
You get ACL for free for private use. Why do you want to create exe
files? In the Lisp world, it's quite safe to assume that you want to be
more than a hobbyist when you want to create exe files.
> MCL around $650
Digitool still has the introductory offer of MCL 5.0 for $495.
> The only good deal besides free Lisps is Corman Lisp about
> $125 student price.
>
> I refuse to pay more for a compiler than for a used car.
...so this limits your choices.
> If I had enough $$$ I prob. would buy Allegro--but I am a student and don't
> have alot of spare $$$ to feed my hobby.
There are lots of good Common Lisp implementations out there. As a
student you usually get discounts on commercial CL systems - LispWorks
costs $540 and MCL currently $297.
> One sad fact is this: the price for used Lispms is so reasonable because
> Lispm companys priced themselves out of existance in the first place--if
> they provides 1-2K machines they'd sell a lot to hobbyists. Esp. since Lisp
> Compilers for Windoz are so costly.
You have to take into account that Common Lisp is not exactly a
mainstream language. So the group of supporters is considerably smaller
than for other languages, and there are currently no big companies
involved in Lisp. The commercial vendors need to have good ideas how to
ensure their existence in the long run. What you are currently
perceiving is that their ideas of how to achieve this don't match yours.
Pascal
Gareth McCaughan
> How can Commercial Lisp Vendors expect people to pay 2K or more
> for there Lisp Environment when you can get a Lispm for around 1K.
Because their Lisp environments allow you to do things
that you cannot do with a Lisp machine bought for around 1K.
Such as taking advantage of the absurd speed of modern
PC hardware, making executables everyone else can use,
not requiring you to buy an extra computer, not screwing
you completely if your hardware suffers a failure, and
so on.
And because they offer commercial support, which some people
value highly.
> BTW, the Lispm blows most Commercial Lisps out of the water in terms of
> features.
>
> Allegro CL $3K with student discount. (The ed. that makes exe. files under
> windows.) no wonder the price of Allegro CL is not on there web site.
Will a Lisp Machine bought for $1K make EXE files that
run in Windows? :-)
> Why is Allegro so pricey?
Because they have found that enough customers are
willing to pay that price. It's a free market in
operation. I wouldn't pay $3K for Allegro myself,
but fortunately I don't have to because there are
cheaper systems around. If I ever find myself
needing what Franz offers badly enough to pay $3K
for it (or, presumably, more, since it's a while
since I was a student), I know where to find it.
> I refuse to pay more for a compiler than for a used car.
Uh-huh. And apparently you also refuse to pay more than
$125 for a used car. That's your choice, of course.
For my part, I don't see why there should be the
slightest relationship between the price of a compiler
and the price of a used car. But, since it seems to
matter to you, there are plenty of used cars for sale
that cost a lot more than the student edition of Allegro CL.
> Another benefit of buying a Lispm vs. a Commercial Lisp is this--the price
> of the Lispm includes the computer.
Another benefit of buying a Lisp that runs on stock hardware
is that you don't need to find space for another computer.
> Expect a computer for Lisp to cost <$2,000 if you want a good one.
If you want one that merely runs Lisp programs as fast
as that $1000 Lisp Machine, then you do not need to pay
anything like $2000 for it. (Did you mean ">" rather than
"<", by the way?)
> If I had enough $$$ I prob. would buy Allegro--but I am a student and don't
> have alot of spare $$$ to feed my hobby.
That's fine. Just as well there are good free CL implementations,
and cheap commercial ones, as well as the pricey Allegro CL,
isn't it?
> One sad fact is this: the price for used Lispms is so reasonable because
> Lispm companys priced themselves out of existance in the first place--if
> they provides 1-2K machines they'd sell a lot to hobbyists. Esp. since Lisp
> Compilers for Windoz are so costly.
They would also have made a loss on every one they shipped.
As for me, I'd *love* to have a Lisp Machine, but that
love doesn't go far enough to make me either buy old,
power-hungry, flaky, space-consuming hardware and try
to squeeze it into my house, or to shell out the money
to buy an Alpha machine and OpenGenera (which would be,
I believe, somewhat more than the $3K you quoted for
Allegro CL). I choose not to whine about this.
--
Gareth McCaughan Gareth.M...@pobox.com
.sig under construc
OCID
"Franz Kafka" <Symbolics _ XL1201 _ Sebek _ Budo _ Kafka @ hotmail . com>
> How can Commercial Lisp Vendors expect people to pay 2K or more
> for there Lisp Environment when you can get a Lispm for around 1K.
Their product, their price
> BTW, the Lispm blows most Commercial Lisps out of the water in terms of
> features.
>
> Allegro CL $3K with student discount. (The ed. that makes exe. files under
> windows.) no wonder the price of Allegro CL is not on there web site.
>
> KnowlegeWorks $2.5K Commercial
>
> LispWorks $950 Commercial
>
> GoldWorks III < $2K educational.
>
> MCL around $650
Clisp, CMUCL, SBCL, OpenMCL, LUSH, newLisp etc free. Clisp works fine
under cygwin on windoze as well. And then there are various scheme dialects
as
well ... some that compile to executable binaries.
Allegro also has an eval version which expires every 60 days but can be
renewed
at the click of a button and is a decent learning environment.
> PS
>
> Why is Allegro so pricey?
>
> PPS
>
> The only good deal besides free Lisps is Corman Lisp about
> $125 student price.
>
> I refuse to pay more for a compiler than for a used car.
Don't have to
> Another benefit of buying a Lispm vs. a Commercial Lisp is this--the price
> of the Lispm includes the computer.
>
> Expect a computer for Lisp to cost <$2,000 if you want a good one.
>
> or around $500 if you want an E-Machine.
>
> Note: this is not to invite flames but to point people who'd like to learn
> Lisp and are not made of $$$ in the right direction.
>
> If I had enough $$$ I prob. would buy Allegro--but I am a student and
don't
> have alot of spare $$$ to feed my hobby.
Take a look at the free lisps. They are pretty high quality.
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
One benefit of my Lispm: no coredump(ala buggy C++/Java), blue screen of
death, or sigsegv 'memory protection' errors. :)
I'd like Windows/Linux better if someone could find a way to fix those
errors--they have a strange way of creaping in during mission critical code
or @ random times.
Use of Lisp=No Buffer Overruns :) sorry Rober Morris.
PS
This is not saying anything bad about Lisp Compilers on MS Windows, just the
excellent job Microsoft did on there OS.
BK
> How can Commercial Lisp Vendors expect people to pay 2K or more
> for there Lisp Environment when you can get a Lispm for around 1K.
Interesting. You are not talking about that Lispm Nubus card for Mac
68K, are you?!
> BTW, the Lispm blows most Commercial Lisps out of the water in terms of
> features.
What about performance?
> Allegro CL $3K with student discount.
(:-o)
> MCL around $650
They used to have a student license for $80 per seat. This was when
Roger Corman was still working on PowerLisp on the Mac. When I
mentioned the MCL $80 student lisense he commented that even as a
shareware author, he could hardly compete with that offer.
> The only good deal besides free Lisps is Corman Lisp about
> $125 student price.
God bless him!
> I refuse to pay more for a compiler than for a used car.
Seems reasonable, but I guess for corporates 2 or 3 grand per seat is
not much of an issue. Check out the client lists of those Lisp vendors
and you will probably find that it's mostly big multinational
corporations.
>
> Another benefit of buying a Lispm vs. a Commercial Lisp is this--the price
> of the Lispm includes the computer.
>
> Expect a computer for Lisp to cost <$2,000 if you want a good one.
>
> or around $500 if you want an E-Machine.
>
> Note: this is not to invite flames but to point people who'd like to learn
> Lisp and are not made of $$$ in the right direction.
How about an eMac with education discount ($778) and an MCL student
license? That should be about the same price than what you quoted for
those second hand Lisp Machines and I'd think you get better overall
value with the eMac, considering that most people will be doing more
than just Lisp.
> If I had enough $$$ I prob. would buy Allegro--but I am a student and don't
> have alot of spare $$$ to feed my hobby.
So what's wrong with Corman Lisp and a cheap PC, or a second hand CRT
iMac and OpenMCL? Shouldn't cost you more than $500-$600.
> One sad fact is this: the price for used Lispms is so reasonable because
> Lispm companys priced themselves out of existance in the first place--if
> they provides 1-2K machines they'd sell a lot to hobbyists.
I doubt that it would have been possible *at the time* to produce
those Lisp Machines at that price.
> The Lisp Machine is ANSI CL, and includes all source code--so you can hack
> Lisp to your hearts content.
So is OpenMCL (and possibly other open source CLs) and it's free. So
the situation doesn't seem to be all that bad, or is it?!
rgds
bk
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
>Why do you want to create exe files?
I want to create some freeware lisp apps in Windows so that people can see
how useful Lisp is--and maybe encourage more people to use Lisp to develop
applications.
If I have to pay for the tools I am using I can not afford to give my apps
away.
If I don't give my apps away, then I can not shout surprize that cool app
you used was written in Lisp--and could not have be written as easily in
C++/Java -- hopefully this will convince more people to try Lisp.
Wanting to write AI code drove me to choose Lisp over C++/Java, Lisp has
better support for Symbolic Computation that either C++/Java and also has an
ANSI standard.
Some benefits right off the bat:
No memory leaks.
No datatype conflicts.
No stack based bugs (I might have got the term wrong) like Robert Morris's
worm used.
Uniform syntax for data and program.
No stack overflows--in Lisp I got
Error: Stack full
:retry with a larger stack :abort to top level :contiune enter a new value
in C++/Java I got:
Stack overflow: Core Dumped.
Some other benfits of Lisp to people wanting to learn how to program:
1.) no pointers (I hate pointers--they tend to crash systems, and are harder
for people to learn than Lisp.)
2.) no datatyping. (Why do I need seventeen functions in C++/Java to do what
one function in Lisp can do. I hate typing the same function over and over
and over and over and over .... and over.
3.) no memory management (The computer does a better job that I do; so, I'll
let the computer do it.)
4.) real macros. (Not the crappy excuse for macros C provides.
5.) I can write generalized functions easily. (I don't have to torture
myself with C++ style templates.)
6.) The ability to represent symbols easily. (In C++/Java it is a pain.)
7.) Multiple Inheritance, MultiMethods. MOP. etc. (C++ is more messy.)
8.) Run-time, on the fly coding. (It is easier to fix bugs.)
9.) I can compile on Lisp function at a time. (It takes a long time to
recompile my C++ programs.)
10.) I don't have to remember which lines require a ';' and which do not.
I picked up Lisp on my own in about 6 months from Dr. David T's book--than I
tried to learn C++/Java at college, it's been about a year and I still
understand Lisp better.
"The C++ Compiler the Sadist; The C++ Programmer the Masochist." -- An Anon.
C++ programmer trying to fix a memory leak @ my school.
PS
If you get more people intrested in AI you'll get more people intrested in
Lisp. My intrest in AI drove me to chose Lisp over Pascal, Java,
VisualBasic, C, C++, and Java.
MicroSoft Products sucking so badly drove me away from VisualBasic,
VisualStudio, and VisualC++ but that's another story.
The only other Lang. besides Common Lisp I'll even consider are: Scheme, and
Prolog. I'll look at Prolog because of my intrest in AI, and NLP.
I am trying to become self taught in AI programming.
"Trying to do AI in C++ is as painful as getting a root canal without
novacain; trust me I've tried once--that drove me to Lisp faster that Killer
Bees would drive me to water. :)" -- Misc. Philosopher.
Tj
While you're getting semi-flamed here, I think that's an interesting
question. To understand this odd-seeming price, you need to get rid
of old conceptions of pricing strategies. We're trained to think in
terms of two strategies: cost+profit and "kill profit for
marketshare!" However it often happens these are ruinous strategies.
For one thing, price often has an effect on perceived value. If
Allegro CL priced at $300 instead of $3000, they wouldn't get 10X the
customers. They'd probably even lose customers since they'd be seen
as competing against lower-cost products. They're not aiming at those
customers.
Also, businesses are less price-sensitive. You're an individual, and
therefore more sensitive. Do you complain about the cost of a
McDonald's grill? No, because it's almost never sold to individuals.
However, companies are in the business of investing money to get more.
Their accountants are happy to pay thousands to get a lot of
perceived benefit.
Here's a fun article:
http://www.fastcompany.com/online/68/pricing.html
I've got a nasty cold, so I'll leave it at that. I'm sure there's a
good book on Amazon that I can't recall right now. Incidentally, I
like to think that Free Software is extremely important because there
are some needs that can't be serviced at sustainable pricing.
Therefore, it must be taken care of outside financial systems.
Tj
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
> > features.
>
> Like `running natively on all major platforms', `delivering
> executables on all major platforms', `supporting 64bit platforms',
> `being actively supported', `competing in performance with C on
> current systems, `supporting CORBA', and a long and growing list of
> other features.
Try:
An Object Oriented Database System (Statice)
An Expert System Development Tool (Joshua, KEE)
Source Code to an Operating System in Lisp (Genera)
The price of these for Windows would add several thousand
to the price of Allegro--all included with the Lispms for the
same price now--when they were sold everything was
prob. an add on.
The two benefits of a Lispm are:
1.) Loads of Lisp source code--just fun to Look at. I esp. like the
OS is written in Lisp--Lispms don't crash often.
2.) An Integrated Developement Environment.
For students who want to learn about Lisp, there neat.
The best benefit I found: no sigsegvs or blue screens of death :)
Andrew Wolven
"Franz Kafka" <Symbolics _ XL1201 _ Sebek _ Budo _ Kafka @ hotmail . com>
wrote in message news:pjyta.4513$pn7...@news02.roc.ny.frontiernet.net...
> for there Lisp Environment when you can get a Lispm for around 1K.
My Lispm is broken. :(
I already paid to have it fixed once and I just can't afford to keep it
running. It is 13 or 14 years old after all. So my answer to your question
is reliability.
AKW
Tj
http://www.amazon.com/exec/obidos/tg/detail/-/013026248X/002-1380390-0472843
BTW, I notice you ignore Corman Lisp, which IIRC is priced low.
Tj
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
"Andrew Wolven" <awo...@nospam.net> wrote:
> My Lispm is broken. :(
At work we have kept Lispms working for longer than Intel or Solaris
workstations. & we are currntly trying to write a Lispm emulator so
that Genera 8.x can be run under Mac OS/X, Window, Linux, yada yada
yada.
If you got a newer Lispm you should be able to replace the SCSI drives
yourself for a fraction of the cost of a new system--look on e-bay.
If it is a 36xx or earlier--most parts a not replaceable.
I can't wait for the day when Lispm companies release new
machines--concidering how advanced there arch. was for the mid 80's. (40
bit--right now most processors at 32bit.)
What chips becides the Alpha no longer supported are 64bit & will run either
Mac OS/X+Virtual PC+Windows or Windows. Please enlighten me on this.
Kunle Odutola
> What chips becides the Alpha no longer supported are 64bit & will run
> either Mac OS/X+Virtual PC+Windows or Windows. Please enlighten me on
> this.
Intel Itanium/IA-64
AMD Hammer/Athlon64/Opteron
IBM Power4
Motorola PowerPC
Kunle
Boethius
> for there Lisp Environment when you can get a Lispm for around 1K.
>
> BTW, the Lispm blows most Commercial Lisps out of the water in terms of
> features.
Fine, then buy yourself one.
> Note: this is not to invite flames
Of course not. };-)
> but to point people who'd like to learn
> Lisp and are not made of $$$ in the right direction.
If only you were as concerned of not wasting bandwidth as you
obviously are of not 'wasting' (sic) your money...
Gareth McCaughan
> At work we have kept Lispms working for longer than Intel or Solaris
> workstations. & we are currntly trying to write a Lispm emulator so
> that Genera 8.x can be run under Mac OS/X, Window, Linux, yada yada
> yada.
I'm puzzled by something. You seem to alternate between
being a student who can't afford to pay much for a
commercial Lisp system, and having a job programming
Lisp Machines. I'm sure it's none of my business, but
I'm curious: what gives? Are you doing a degree in your
spare time, or something?
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
"Gareth McCaughan" <Gareth.M...@pobox.com> wrote in message
news:slrnbbg90k.21k2....@g.local...
Degree on Line at SUNY Empire State.
I can make sure I only get Lisp courses so I don't have to learn an other
language.
When I say at work--I meant I knew some people at RIT who used Lispms and
loved them.
I have to avoid learning about pointers, and memory managment--Lisp does it
for me. :)
I'm happy that way.
I can just code with out designing the program logic easier in Lisp or maybe
Prolog ;) than
in any other Language.
I might even go for an Eng. Lit. degree--I have an A.S. Comp. Sci.
degree--very hard to
get work with it. I learned Lisp on my own. I wanted to do some NLP, Text
Generation and
Lisp was very well suited for that kind of project.
I work from home doing this or that. BTW, I have yet to make money--maybe
after I get a
B.S./B.A. in what I haven't yet decided. (Could be a dual degree.)
U of R's NeuralLingistics looks good to me after my BS/BA but it is only a
PHD program.
If will fill my love of English Lit. Comp. Sci. and Persuasive Rhetoric. I'm
sure to find use
for Lisp in NeuralLinguistics but it will be at least three years before I
can start.
Kunle Odutola
> What chips becides the Alpha no longer supported are 64bit & will run
> either Mac OS/X+Virtual PC+Windows or Windows. Please enlighten me on
> this.
Intel Itanium/IA-64
AMD Hammer/Athlon64/Opteron
IBM Power4
Motorola/IBM PowerPC
Kunle
Ray Blaak
> I can make sure I only get Lisp courses so I don't have to learn an other
> language.
Learn as many languages as you can. Even you still only want to use Lisp,
having knowledge of other ways of doing things makes you a wiser programmer.
Cheers,
Ray Blaak
Gareth McCaughan
> Degree on Line at SUNY Empire State.
>
> I can make sure I only get Lisp courses so I don't have to learn an other
> language.
I hope you don't mind my saying this, but: What a
silly constraint to put yourself under! Lisp is
a fantastic (and my favourite) programming language,
but if you refuse to learn anything else then you're
impoverishing yourself. Open your mind!
(If you already know, let's say, Prolog, ML, C++,
Perl and an assembly language[1], then you may consider
yourself excused. But in that case, there's no need
to "only get Lisp courses".)
[1] There's nothing special about that list. It would
be about as good to know Eiffel, Forth, C, Ruby
and an assembly language. I must insist on the
assembly language, though.
MechaDragon X
"Ray Blaak" <bl...@telus.net> wrote in message
news:u3cjrr...@telus.net...
And looks better on the resume...
David Steuber
> Pascal writes:
> >Why do you want to create exe files?
>
> I want to create some freeware lisp apps in Windows so that people can see
> how useful Lisp is--and maybe encourage more people to use Lisp to develop
> applications.
That's all well and good. But real Free ware comes with source.
> If I have to pay for the tools I am using I can not afford to give my apps
> away.
>
> If I don't give my apps away, then I can not shout surprize that cool app
> you used was written in Lisp--and could not have be written as easily in
> C++/Java -- hopefully this will convince more people to try Lisp.
Again, providing the source would go a long way.
> If you get more people intrested in AI you'll get more people intrested in
> Lisp. My intrest in AI drove me to chose Lisp over Pascal, Java,
> VisualBasic, C, C++, and Java.
That is a debatable point. I imagine plenty of AI work is done in C++
or Java. There is a lot of mind share in these languages.
You have the sources for your LispM code. Here is a thought. See if
you can create a Lisp program that will take Lisp code and produce an
excecutable image for Windows, Linux, Mac, or whatever. Then you can
port the LispM stuff you have over. If the licensing allows it, you
can even share your work.
In the meantime, I will solve my cash problems by sticking with free
Lisp implimentations and not complaigning when they don't measure up
to the commercial ones.
Thaddeus L Olczyk
wrote:
>> How can Commercial Lisp Vendors expect people to pay 2K or more
>> for there Lisp Environment when you can get a Lispm for around 1K.
>
>They are still in business, so obviously they don't need to expect it,
>they just get it.
>
For now. But will that last into the future? I ( directly ) know of
several businesses moving away from Lisp. I ( directly ) know of
no business moving towards Lisp. The high cost is one determining
factor. And it becomes a deadly loop. Once they can't get it from
some, they have to raise prices to get it from the rest. Once that
happens, they lose even more, so even more higher prices.
>> Allegro CL $3K with student discount. (The ed. that makes exe. files under
>> windows.) no wonder the price of Allegro CL is not on there web site.
>
>You get ACL for free for private use. Why do you want to create exe
>files? In the Lisp world, it's quite safe to assume that you want to be
>more than a hobbyist when you want to create exe files.
No you get ACL free for private *experimentation*, and it's only in
the screwed up perspective of the "Lisp world" that you would expect
to be safe assuming that a person who is a "hobbyist" ( and I include
proffesional programmers, who are not proffesional Lisp programmers )
doesn't want exes.
In fact I have a question for you. How many people are using Linux,
FreeBSD, or even MacOS 10 ( which I belive has a layer of "gnu"
software running it )? How many Windows users are using Cygwin?
Look at all those "hobbyists" who have found a need for exe's.
Frankly I was going to point out that a hobbyist might want to submit
an ICFP entry in the free ACL if he could ( and anyone who says they
wouldn't rather see an ICFP entry done in Lisp do well in ICFP is
lieing ). Or my inability to create a backend "whitelist" filter for
Popfile in the free version of ACL. Or distribute it on SourceForge.
One of the ways that a "hobbyist" can stop being a hobbyist is to
distribute an open source program demostrating his capabilities. etc.
But the whole wealth of opensource software that makes up Linux and
FreeBSD say it much better. The fact that most of that software is
written in C/C++ and Perl makes for powerfull recruiting tools for new
programmers, and there isn't a Lisp programmer in this group who
wouldn't die for a chance for Lisp to have a fraction of that exposure
( anyone who says otherwise is lieing ).
--------------------------------------------------
Thaddeus L. Olczyk, PhD
Think twice, code once.
Thaddeus L Olczyk
<david....@verizon.net> wrote:
>"Franz Kafka" <Symbolics _ XL1201 _ Sebek _ Budo _ Kafka @ hotmail . com> writes:
>
>> Pascal writes:
>> >Why do you want to create exe files?
>>
>> I want to create some freeware lisp apps in Windows so that people can see
>> how useful Lisp is--and maybe encourage more people to use Lisp to develop
>> applications.
>
>That's all well and good. But real Free ware comes with source.
>
>> If I have to pay for the tools I am using I can not afford to give my apps
>> away.
>>
>> If I don't give my apps away, then I can not shout surprize that cool app
>> you used was written in Lisp--and could not have be written as easily in
>> C++/Java -- hopefully this will convince more people to try Lisp.
>
>Again, providing the source would go a long way.
>
I think you've gotten off the point here. Many people who download the
Freeware preffer executables. Me I download opensource in both binary
and source, hoping that things go the easy way with the source.
If he only distributed the source, then it would exacerbate the
situation more, as the people using the source would have to buy ACL
to compile it.
Friedrich Dominicus
> On Mon, 05 May 2003 22:58:37 +0200, Pascal Costanza <cost...@web.de>
> wrote:
>
> >> How can Commercial Lisp Vendors expect people to pay 2K or more
> >> for there Lisp Environment when you can get a Lispm for around 1K.
> >
> >They are still in business, so obviously they don't need to expect it,
> >they just get it.
> >
> For now. But will that last into the future? I ( directly ) know of
> several businesses moving away from Lisp. I ( directly ) know of
> no business moving towards Lisp.
Ok, we're the counterexample and moving towards Lisp, so now you know
(at least) one moving into the "right" direction.
> The high cost is one determining
> factor. And it becomes a deadly loop. Once they can't get it from
> some, they have to raise prices to get it from the rest. Once that
> happens, they lose even more, so even more higher prices.
The prices for LispWorks are very compatible.
>
> Frankly I was going to point out that a hobbyist might want to submit
> an ICFP entry in the free ACL if he could ( and anyone who says they
> wouldn't rather see an ICFP entry done in Lisp do well in ICFP is
> lieing ). Or my inability to create a backend "whitelist" filter for
> Popfile in the free version of ACL. Or distribute it on SourceForge.
> One of the ways that a "hobbyist" can stop being a hobbyist is to
> distribute an open source program demostrating his capabilities. etc.
> But the whole wealth of opensource software that makes up Linux and
> FreeBSD say it much better.
Well on Linux and FreeBSD you can use the free Lisp and everbody can
install them. Don't tell me you need executable for that.
> The fact that most of that software is
> written in C/C++ and Perl makes for powerfull recruiting tools for new
> programmers, and there isn't a Lisp programmer in this group who
> wouldn't die for a chance for Lisp to have a fraction of that exposure
> ( anyone who says otherwise is lieing ).
Hobbyist spend billons of dollars for the Hobby. So why can't you do
the same with Lisp implementations?
Friedrich
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
"Friedrich Dominicus" <fr...@q-software-solutions.com> wrote in message
news:87issm3...@fbigm.here...
> Hobbyist spend billons of dollars for the Hobby. So why can't you do
> the same with Lisp implementations?
What hobby, Classic Cars? There Rare, and costly.
The Classic Car company won't make
money by increasing the userbase--
software companyies will because
the more popular a lang. is
the more companyies will use it.
Reading Novels? I don't think you
know what a library is.
Listening to Classical Music? There is
a new invention called a radio.
Drawing? It cost under $200 not over
$2K to get started.
Poetry? Most artists are in it for
the fame not the money; I know a
lot of them from my college.
Rec. Drugs? You get an instant
benefit--that Lisp or even C++ won't
give you. Lisp is also less dangerous
to do. :)
Other Computer Lang.? Most other
Lang.'s give free or low cost (under
$150) compilers that will, yes will,
produce exe code.
>
> Friedrich
>
Do you think bitter people, not just Fred but everyone against a low cost
Lisp solution for hobbyist, and also for people who don't get why a hobbyist
would want an exe file, trying to fight for Commercial Lisps and telling
people to pay thousands for a compiler is a great way to attract people to
want to learn Lisp :)
This thread was starting to get in the right direction, but I guess that
Lisp
programmers really don't want to attract new users. I thought I was
wrong but most of the people on this thread seem to be against a
hobbyist who would like executable code.
You need a exe file if you want to trick people into using Lisp. Write a
very good freeware prog. in Lisp that people find useful and more
people will be willing to learn Lisp because thay saw first hand
what Lisp could do. You can't do that with source code--people
already have to want to use Lisp with that approch.
If people already wanted to learn Lisp we wouldn't need to attract
them to our language, now would we?
No, we wouldn't.
Franz,
PS
Think of Linux. They give Linux away free. But, if you need support
for comercial apps. you'll have to pay highly for it. I just want the
same done with Lisps. Hobbyists will not need support, and
will increase our user base.
I know I want a larger userbase of Lisp programmers; I can't
say the same about everyone else.
PPS
I'll rant until people understand the a good low cost
solution to making Lisp exe files is a big step
in the right direction to making Lisp a popular
language.
It must be in ANSI standard Lisp because only
then will the source code be portable.
PPPS
Out of neticate I warned you that this was
a rant. BTW, I kind of like Erik N.'s posting style.
Joe Marshall
> I can make sure I only get Lisp courses so I don't have to learn an
> other language.
That's an interesting approach. While I agree that knowing a few more
languages is a good thing, it is often easier to approach a language
as `like lisp, but stupid syntax and no GC'.
Friedrich Dominicus
> "Friedrich Dominicus" <fr...@q-software-solutions.com> wrote in message
> news:87issm3...@fbigm.here...
>
> > Hobbyist spend billons of dollars for the Hobby. So why can't you do
> > the same with Lisp implementations?
> What hobby, Classic Cars?
Oh come one: skiing, surfing, sailing, booting, motorcycling,
collecting, buying jewlry, golfing, tennis, horse-riding and and
Regards
Friedrich
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
You can rent skis, motorcycles, sail boats 4 a reasonable price.
You don't have to pay for a horse to ride one -- you can
pay for tine,
I think you can rent surf boards too.
booting -- do you mean drinking until you have to dump the
liquor into a boot because you drank too much ;) or boating.
You can't rent a Commercial Lisp Compiler. Don't give me the
copying excuse. You can rent movies at Blockbuster--they cost
much more to make than a Lisp Compiler and can be copied
just as easily.
If a firm makes illegal copys of a commercial Lisp--the corporate
lawyers can get more then the cost of the software from the bad
company. (But, it would give more people who want to learn Lisp
easy access to a high grade Dev. Env.)
If you can rent software--please tell me where.
USER @fun Dr. Leary
So is LSD Manufacture--why don't you include that
too. I'd rather help people learn coding than
LSD manufacture--but I guess u differ.
Tim Leary ;)
Thaddeus L Olczyk
Sebek _ Budo _ Kafka @ hotmail . com> wrote:
>Why is Allegro so pricey?
All that has been said and done, in the end it comes down to one
thing. Franz has to make back what it cost to produce ACL, and
a significant profit. If they do not make enough of a profit to
make investing in ACL better then the stock or bond market
( over the long term about a 10% ROI ), or other ventures,
then investors wont invest. ( Worse Franz is probably considered
a high risk venture because looking at it's balance sheet, client base
and the direction things have been going over the last 5, 10 years,
the probabilty is high that it won't be around much longer. So I
suspect that investors want a bigger ROI, probably at least 15%. )
They don't have a large customer base sso they have to charge what
they charge.
Make no mistake. If they could charge less they would. It would
certainly make it easier to sell to businesses, consulting companies
and individual consultants/contractors with a lower price. That in
turn would lead to a larger customer base, reducing the need for a big
profit. The larger customer base would mean more programmers
that know CL, which would mean that companies would be more willing
to use CL ( ie a larger cuctomer base ). The problem is that they
really can't right now.
Paolo Amoroso
_ Budo _ Kafka @ hotmail . com> wrote:
> How can Commercial Lisp Vendors expect people to pay 2K or more
> for there Lisp Environment when you can get a Lispm for around 1K.
Ever run a business?
Paolo
--
Paolo Amoroso <amo...@mclink.it>
Paolo Amoroso
_ Budo _ Kafka @ hotmail . com> wrote:
> I want to create some freeware lisp apps in Windows so that people can see
> how useful Lisp is--and maybe encourage more people to use Lisp to develop
> applications.
If you are going to use you LispM as a development environment, how do you
plan to deploy the application on Windows and distribute it as freeware?
Paul Wallich
Thaddeus L Olczyk <olc...@interaccess.com> wrote:
> On Mon, 05 May 2003 18:45:41 GMT, "Franz Kafka" <Symbolics _ XL1201 _
> Sebek _ Budo _ Kafka @ hotmail . com> wrote:
>
> >Why is Allegro so pricey?
[...]
> Make no mistake. If they could charge less they would. It would
> certainly make it easier to sell to businesses, consulting companies
> and individual consultants/contractors with a lower price. That in
> turn would lead to a larger customer base, reducing the need for a big
> profit. The larger customer base would mean more programmers
> that know CL, which would mean that companies would be more willing
> to use CL ( ie a larger cuctomer base ). The problem is that they
> really can't right now.
That's not necessarily true, at least not for large values of "less". As
long a CL is not a mainstream language (and maybe even if it were) the
cost of producing and maintaining the software mayb be less important
than the cost of supporting it. Even if Allegro were utterly stable,
customers would need support for installation, configuring, twiddly
little aspects of use and so forth. If anything, a larger customer base
would mean higher-than-proportional increases in the cost of providing
that support. If the market were orders of magnitude bigger, you might
be able to get away with a microsoft- or intuit-style support system,
but there's an enormous gulf between here and there.
paul
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
"Paolo Amoroso" <amo...@mclink.it> wrote in message
news:mDe5Pp7GXSAfwe...@4ax.com...
You can develop CL application--but you have to use ANSI CL, CLOS, and maybe
CLIM--as long as you stay away from Lispm only functions your fine.
A friend of mine wrote Pascal's code in Genera 8.3 and Paul Garham's
with-gensyms macro. It ran fine.
BK
> > Make no mistake. If they could charge less they would. <snip>
>
> That's not necessarily true, at least not for large values of "less". As
> long a CL is not a mainstream language (and maybe even if it were) the
> cost of producing and maintaining the software mayb be less important
> than the cost of supporting it. Even if Allegro were utterly stable,
> customers would need support for installation, configuring, twiddly
> little aspects of use and so forth. If anything, a larger customer base
> would mean higher-than-proportional increases in the cost of providing
> that support. If the market were orders of magnitude bigger, you might
> be able to get away with a microsoft- or intuit-style support system,
> but there's an enormous gulf between here and there.
I wonder if that was behind Digitools decision to open source OpenMCL.
At first it would seem crazy that they did this, paricularly as they
are only on the Mac platform which is a smaller community already.
But then, it may well be that a slightly-out of step and less complete
open source version will work in their favour. It allows more folks to
try Lisp on the Mac withough having to pay for the trial thus
potentially creating a larger user base but Digitool don't have to pay
a dime to support it.
Clearly if someone is going to deliver a commercial product they will
probably think twice whether they want to base their product on an
open source compiler without support, unless they have a resident Lisp
hacker who knows the whole thing well enough to fix everything that
might pop up.
Gary Byers mentioned that there were just under 60 people on the
OpenMCL mailing lisp, that was last December. That doesn't seem to be
a lot. If this was shareware and everyone had paid $100 for it, then
the author would have received $6000, that will certainly not be
enough to pay for the work it takes to make something like OpenMCL or
even Roger Corman's discontinued PowerLisp.
And in the case of OpenMCL, the excuse that people won't trust it
cause its shareware/open source doesn't really count because the bulk
of the code came from Digitool and at least when it was split off was
largely identical to the commercial MCL, so the quality of the product
cannot possibly be in question.
rgds
bk
BK
> You can't rent a Commercial Lisp Compiler. Don't give me the
> copying excuse. You can rent movies at Blockbuster--they cost
> much more to make than a Lisp Compiler and can be copied
> just as easily.
I asked my mother in law if she would rent the lisp compiler at $2 a
day or the Blockbuster video at $2 a day. She said she'd rather rent
the video. I asked again if she'd rent the lisp compiler if she had no
choice between the compiler and the video and she said if she can't
have the video she'd not rent the compiler either.
I guess this would be so for the overwhelming majority of people who
rent videos. It doesn't seem to be a sensible analogy by whatever
standard.
> If you can rent software--please tell me where.
DEC used to have a software rental program. I don't know if HPQ still
do it, though.
rgds
bk
Bob Bechtel
agree with you. There's a solution -- go build the tool that you want
yourself. If your analysis is correct, the world will beat a path to
your door, and you will have inarguably demonstrated your chops in the
open/free software environment (whichever you consider appropriate).
Even before you're finished, others may be convinced of your plan and
join in to get you to the goal more quickly. And just think of the
satisfaction in seeing the naysayers fall by the wayside!
Arguing that someone else should do something you want is usually futile
unless you can demonstrate to their satisfaction (not yours) how they're
better off doing so -- and making assertions about what might happen is
not demonstration.
bob bechtel
Christopher Browne
> Paul Wallich <p...@panix.com> wrote ...
> At first it would seem crazy that they did this, paricularly as they
> are only on the Mac platform which is a smaller community already.
>
> But then, it may well be that a slightly-out of step and less
> complete open source version will work in their favour. It allows
> more folks to try Lisp on the Mac withough having to pay for the
> trial thus potentially creating a larger user base but Digitool
> don't have to pay a dime to support it.
What it buys them is a perceived reduction of risk.
If they go belly up, the code at the base is clearly not lost.
And there *might* be some contributions that come back.
--
(reverse (concatenate 'string "gro.gultn@" "enworbbc"))
http://cbbrowne.com/info/spreadsheets.html
Rules of the Evil Overlord #83. "If I'm eating dinner with the hero,
put poison in his goblet, then have to leave the table for any reason,
I will order new drinks for both of us instead of trying to decide
whether or not to switch with him." <http://www.eviloverlord.com/>
Steven M. Haflich
>>Why is Allegro so pricey?
>
> I have no special knowledge of this, but I suspect the reason is that
> they do not want to make their money on Lisp sales, but rather on Lisp
> support. Implicitly, by screening out people who think that this is
> "pricey" they are eliminating people who don't have money for support,
> and streamlining their commercial dialog to be established companies
> with adequate cash flow to both be able to buy support and appreciate
> its need. From the tone of the your message, it sounds to me like if
> this is their plan, it's working.
Kent, my old friend, while your analysis has a lot of merit, I find that
it is expressed much more sourly than necessary, and that it distorts
the market reality.
[But first, I need to state that while I am an employee of Franz, I am
here voicing my own opinion, not that of Franz. If you want Franz'
opinion, send email to in...@franz.com.]
I don't think it is fair to talk about Franz' strategy of sales vs.
support. Please think back to that initial X3J13 meeting where the
ad hoc committee on the purpose of the language consed up the words
"Industrial Lisp". Indeed, Franz bundles support into its pricing,
but the reason (it seems to me, but not necessarily to Franz Inc.
itself) is that the company's support mechanisms are willing and
prepared to do what it takes to make a customer's application
succeed. It isn't _always_ possible to achieve success, but _usually_
Franz seems able to make customer applications succeed. Some
customers never need this service, and some customers need it
overmuch, but in my humble non-Franz-Inc-sanctioned opinion, that is
what customers are paying for. Given that a single programmer costs
more than an order of magnitude more than typical license fees,
paying for license and support is a wise bet for a commercial
enterprise. I'd like to think that this would be objectively true
even if those support fees were not what pays for my groceries.
As an example, last month I was packetized without checksum and put
onto a plane to Japan to work one week of 16-hour days helping an
important customer with several difficult application problems.
This trip cost more than Franz earns from this customer (even
assuming my lost time is completely valueless) but the result is that
the customer's application was delivered and remains in production.
If I were a CL hobbyist living on scattered consulting, I probably
would not be willing or able to pay for this backup support. But if
I were a solvent company with contractual obligations of my own, it
would be silly not to pay for it. There is room in the CL ecosystem
for both kinds of animals.
I feel (but this is in no way an official Franz statement) that Franz
has opted for a particular region of the price/support curve, and if
no one occupied that region, the Common Lisp we all love would be
relegated to an obscure historical hobbyist language. The Lisp machine
hobbyist lives on another distant region of that curve, and it is good
that he is there maintaining that different region of the CL universe.
(The LM was a wonderful peak of programmability, and will remain so as
long as enthusiasts can find enough low-sulfur coal to shovel into
those machines' boilers to keep them running... :-)
Ng Pheng Siong
> You need a exe file if you want to trick people into using Lisp. Write a
> very good freeware prog. in Lisp that people find useful and more
> people will be willing to learn Lisp because thay saw first hand
> what Lisp could do.
Here's a deal: I have a commercial Lispworks license. You write your very
good freeware prog in portable CL and I'll build your Windows exe for you.
What very good freeware prog do you have in mind to write?
--
Ng Pheng Siong <ng...@netmemetic.com>
http://firewall.rulemaker.net -+- Manage Your Firewall Rulebase Changes
http://www.post1.com/home/ngps -+- Open Source Python Crypto & SSL
Fernando Mato Mira
> "Friedrich Dominicus" <fr...@q-software-solutions.com> wrote in message
> news:871xza3...@fbigm.here...
> > "Franz Kafka" <Symbolics _ XL1201 _ Sebek _ Budo _ Kafka @ hotmail . com>
> writes:
> >
> > > "Friedrich Dominicus" <fr...@q-software-solutions.com> wrote in message
> > > news:87issm3...@fbigm.here...
> > >
> > > > Hobbyist spend billons of dollars for the Hobby. So why can't you do
> > > > the same with Lisp implementations?
> > > What hobby, Classic Cars?
> >
> > Oh come one: skiing, surfing, sailing, booting, motorcycling,
> > collecting, buying jewlry, golfing, tennis, horse-riding and and
> >
> > Regards
> > Friedrich
>
> You can't rent a Commercial Lisp Compiler. Don't give me the
You can't rent a model railroad, either. You can join a club, though.
Found a Lisp club with a floating license server. Demand floating
licenses if your selected vendor does not provide this option.
Jacek Generowicz
> Gary Byers mentioned that there were just under 60 people on the
> OpenMCL mailing lisp,
Doesn't the exstence of gmane.org makes that number even more
meaningless than it would otherwise be?
Matthias Heiler
On Mon, 05 May 2003 18:45:41 GMT, "Franz Kafka" <Symbolics _ XL1201 _
Sebek _ Budo _ Kafka @ hotmail . com> wrote:
>
>Why is Allegro so pricey?
Because they can be. It think it's ok for a company to charge as much as
the market is willing to pay. (As long as that company doesn't have a
monopoly, that is.)
The more interesting question is: Why do people and companies agree to these
prices if they can get 'alternatives' for free? There must be something
which makes ACL et al. _much_ more valuable than just plain-old-standard CL
which you can get decently implemented for free.
I have never used a commercial CL implementation, but I _guess_ it's the
libraries shipping with these systems. Is that right?
Matthias
Paul F. Dietz
> The more interesting question is: Why do people and companies agree to these
> prices if they can get 'alternatives' for free? There must be something
> which makes ACL et al. _much_ more valuable than just plain-old-standard CL
> which you can get decently implemented for free.
The cost of ACL is a small fraction of the cost of the employee using ACL,
so the increase in productivity doesn't have to be very large to be worthwhile.
Paul
David Steuber
> If you can rent software--please tell me where.
I believe that is the direction Microsoft is heading with Windows.
And all their other products.
David Steuber
> >Again, providing the source would go a long way.
> >
> I think you've gotten off the point here. Many people who download the
> Freeware preffer executables. Me I download opensource in both binary
> and source, hoping that things go the easy way with the source.
> If he only distributed the source, then it would exacerbate the
> situation more, as the people using the source would have to buy ACL
> to compile it.
Perhaps I have. But you cut out the part where I mention the Free
Lisps that would allow you to distribute an executable file.
One thing I don't quite get is why CMUCL (just as an example) doesn't
build an ELF executable or shared lib instead of a FASL file. Perhaps
when I understand the implementation I will know.
In the Free software world, it is quite common to distribute just the
source. You need a C compiler (if it is written in C) to get it to
work. GCC just happens to be available on every Linux distro I am
aware of, so a source only distro is not a big problem. However, it
is also very common to have precompiled executables for different
architectures. If it can be done with C, why not Lisp?
If that is the point you are trying to make, then I think we are on
the same page. My point is that a Free Lisp can be used as a
bootstrap to producing a Lisp system that does the desired thing if
none of the existing free ones do.
David Steuber
> As an example, last month I was packetized without checksum and put
> onto a plane to Japan to work one week of 16-hour days helping an
> important customer with several difficult application problems.
> This trip cost more than Franz earns from this customer (even
> assuming my lost time is completely valueless) but the result is that
> the customer's application was delivered and remains in production.
Wow. Franz sounds like a really cool company. Although, I don't know
what sort of A-Team trick they would have to pull to get me on a plane
to Japan. (Especially since I don't have a current passport ;-)
It must be a lot of fun working for a company with that sort of
dedication and commitment. Thanks for posting that small insight. If
nothing else, it is most interesting.
Matthias
> The cost of ACL is a small fraction of the cost of the employee using ACL,
> so the increase in productivity doesn't have to be very large to be
> worthwhile.
I was under the impression that runtime fees were charged. But google told
me that's not the case at least for some commercial CLs.
Matthias
Espen Vestre
> I believe that is the direction Microsoft is heading with Windows.
>
> And all their other products.
You bet. Guess what they'll use Palladium for (you thought it was
security? hah!)
--
(espen)
Paolo Amoroso
<fr...@q-software-solutions.com> wrote:
> Thaddeus L Olczyk <olc...@interaccess.com> writes:
[...]
> > For now. But will that last into the future? I ( directly ) know of
> > several businesses moving away from Lisp. I ( directly ) know of
> > no business moving towards Lisp.
> Ok, we're the counterexample and moving towards Lisp, so now you know
You might consider adding an entry to:
http://alu.cliki.net/Industry%20Application
Kent M Pitman
> Kent M Pitman wrote:
>
> >>Why is Allegro so pricey?
> > I have no special knowledge of this, but I suspect the reason is that
> > they do not want to make their money on Lisp sales, but rather on Lisp
> > support. Implicitly, by screening out people who think that this is
> > "pricey" they are eliminating people who don't have money for support,
> > and streamlining their commercial dialog to be established companies
> > with adequate cash flow to both be able to buy support and appreciate
> > its need. From the tone of the your message, it sounds to me like if
> > this is their plan, it's working.
>
> Kent, my old friend, while your analysis has a lot of merit, I find that
> it is expressed much more sourly than necessary, and that it distorts
> the market reality.
I'm not sour about it at all. You must have misread my tone! I was
just trying to offer a simple, plausible explanation of why a company
might reasonably price their product higher than others it competes
with and still expect to do ok in the market. If people are really
offering the identical product, you can't do that. But when there are
differences between what each vendor wants, you can.
Kent M Pitman
> The more interesting question is: Why do people and companies agree
> to these prices if they can get 'alternatives' for free? There must
> be something which makes ACL et al. _much_ more valuable than just
> plain-old-standard CL which you can get decently implemented for
> free.
ACL is definitely a high quality product, but I wanted to respond to
your more general claim, which I believe to be false: Just because a
product succeeds as higher priced doesn't prove it's better. It
proves only that the world believes it's better.
I don't have firsthand knowledge, but I recall hearing a story that
that Chivas Regal was selling a so-so alcoholic beverage and that they
were faced with bankruptcy when someone came up with a clever
strategy: Why don't we just up the price and tell people it's the one
for special occasions? That is, they manufactured the illusion of
being better. As I heard it, this plan worked and saved the company.
I'm sure there are other, less bold, examples of the same concept.
Even "pet rocks" are proof that good marketing can compensate for a
lot of non-value in delivered products.
Duane Rettig
> Kent M Pitman wrote:
>
> >>Why is Allegro so pricey?
> > I have no special knowledge of this, but I suspect the reason is that
>
> > they do not want to make their money on Lisp sales, but rather on Lisp
> > support. Implicitly, by screening out people who think that this is
> > "pricey" they are eliminating people who don't have money for support,
> > and streamlining their commercial dialog to be established companies
> > with adequate cash flow to both be able to buy support and appreciate
> > its need. From the tone of the your message, it sounds to me like if
> > this is their plan, it's working.
>
> Kent, my old friend, while your analysis has a lot of merit, I find that
> it is expressed much more sourly than necessary, and that it distorts
> the market reality.
I didn't read overly sour tone into Kent's view, if any sourness at
all. I saw his post as just his own view of reality as he sees Franz.
> [But first, I need to state that while I am an employee of Franz, I am
> here voicing my own opinion, not that of Franz. If you want Franz'
> opinion, send email to in...@franz.com.]
Frankly, although Steve is invoking the standard disclaimer explicitly,
there is at least one other developer in Franz who feels pretty much
the same way, and from my talks with others internally I would venture
to say that the number is larger than two as well. Thanks, Steve.
[ customer-success-orientation discussion elided ]
> If I were a CL hobbyist living on scattered consulting, I probably
> would not be willing or able to pay for this backup support. But if
> I were a solvent company with contractual obligations of my own, it
> would be silly not to pay for it. There is room in the CL ecosystem
> for both kinds of animals.
I think that this paragraph is the one part I don't completely agree
with:
1. If I were a CL hobbyist, I wouldn't be living on consulting; if
there were really consulting money coming in it would not be necessary
to make up for the money I put into my hobby. In fact, I would consider
a hobbyist almost definitionally to be one who isn't trying to make a
living on that on which he loves to spend time and money.
But, in fact, the true hobbyist tends to spend _tremendous_ amounts of
money on his hobby; elsewhere on this newsgroup (perhaps on this thread?)
someone mentioned hobbys that are extremely expensive, and not necessarily
that poipular; one has to love the hobby in order to spend so much time
and money on what they love to do. As I recall, the conversation got
sidetracked toward rentals, as in videos (I think I would also place
video games into the same category). However, I wouldn't tend to classify
something that is easily rentable as a hobby - and of course it depends
on the circumstances - it is hard for me to think of renting movies as
anything other than a means to unwind (unless I were secretly harboring
a desire to become the next Siskel & Ebert / Ebert & Roeper...). When
I truly get into a hobby, the money I make on work has to be carefully
divided so I don't starve while feeding my hobby...
2. If I were _living_ on my scattered consulting, it wouldn't be my
hobby; I would want to make more money than I spent. This amounts to
a profit-and-loss center, which is a similar model to that which a
larger company would have. Contractual obligations would indeed be
involved. And as I move an activity from hobbyist to carreer, it ceases
to become something I can just pour my discretionary money and time into;
I must scale back so that expenses are reduced below income.
So what about support? Support seems at first glance to be an unnecessary
expense, and thus one which should be cut at any opportunity. And the
free/open-source industry has given inexperienced programmers the
impression that such support is completely unnecessary, since if the
programmer has a problem, he/she can just go into the source and fix
it. And if in your consulting contract you have all the time in the
world to deviate from the primary goal of getting _XYZ_ product out
for an indeterminate time in order to work on this side issue (because
you have the source code and can thus do so) then more power to you.
But more often than not, such deviations from the primary contract
tend to tarnish your reputation of getting good product out _on_ _time_,
and unless you've built that slop into your initial estimation, the
excuse that "I had to go in and fix some of the free software I was
using" doesn't necessarily cut it with the customer you are consulting
for. Just be sure in that situation that you don't sign any contracts
with penalty clauses for lateness...
Instead it makes sense to have pieces of the project you are working on
to be covered by support contract, so that if anything goes wrong with
what you are using, you can get a quick fix or workaround (or lesson as
to how to use correctly) from those who know it best, and you can then
continue on quickly with your main goal without using up much of the
built-in slop factor in your schedule.
Finally, the statement that there is room in the CL ecosystem for both
kinds of animals is indeed true, but I tend to believe that both kinds
of animals (the hobbyist and the consultant/company-employee) are really
the same animal with a slightly different emphasis and level of
experience.
(I didn't mention a third group of CL programmers who are neither
hobbyists nor consultants/company-employees - I would call these the
"dabblers". I wonder how many dabblers view themselves as hobbyists...)
--
Duane Rettig du...@franz.com Franz Inc. http://www.franz.com/
555 12th St., Suite 1450 http://www.555citycenter.com/
Oakland, Ca. 94607 Phone: (510) 452-2000; Fax: (510) 452-0182
Matthias
> Matthias Heiler writes:
>> The more interesting question is: Why do people and companies agree
>> to these prices if they can get 'alternatives' for free? There must
>> be something which makes ACL et al. _much_ more valuable than just
>> plain-old-standard CL which you can get decently implemented for
>> free.
>
> ACL is definitely a high quality product, but I wanted to respond to
> your more general claim, which I believe to be false: Just because a
> product succeeds as higher priced doesn't prove it's better. It
> proves only that the world believes it's better.
You are right, and there are nice examples in the software industry where
expensive-but-shabby products succeeded thanks to excellent marketing.
But my question was really a sincere one, maybe badly phrased: What are the
main technical reasons to prefer ACL over CMUCL or CLISP? Or differently:
If you could add one feature of ACL to your favourite free CL, what would
it be?
(This is essentially the question of the original poster with "Lisp
Machines" replaced by "some free Lisp".)
Thanks,
Matthias
Paolo Amoroso
> Gary Byers mentioned that there were just under 60 people on the
> OpenMCL mailing lisp, that was last December. That doesn't seem to be
> a lot. If this was shareware and everyone had paid $100 for it, then
Only a minority of users subscribe to mailing lists or online forums.
Marc Battyani
"Matthias" <sp...@yourself.pl> wrote
> But my question was really a sincere one, maybe badly phrased: What are the
> main technical reasons to prefer ACL over CMUCL or CLISP? Or differently:
> If you could add one feature of ACL to your favourite free CL, what would
> it be?
I can't speak for ACL but I have Lispworks which is another commercial Common
Lisp.
The main features that I don't find in Open Source implementations:
Portability between Windows/Linux/MacOS X/Unix etc.
A development environment (debugger, profiler, inspectors, etc.)
Multi-threading
The possibility to deliver executables or dynamic libraries.
Support
For now, you don't have an Open Source implementation that offers this.
Marc
Thaddeus L Olczyk
It should be said that there are implementations that offer this
individually, but none offer it simultaneaously. The fact that there
is no such thing is something that has been holding Lisp back.
--------------------------------------------------
Thaddeus L. Olczyk, PhD
Think twice, code once.
Larry Hunter
But my question was really a sincere one, maybe badly phrased: What
are the main technical reasons to prefer ACL over CMUCL or CLISP?
Steve and Duane have discussed it from the vendor's point of view, but
let me give you the customer slant.
My lab has been collaborating with IBM Life Sciences in doing
knowledge-based bioinformatics for some time now. We had run into some
problems with the 4G (really 2G) address space limit of 32 bit
machines for some of the molecular biology knowledge-bases we were
trying to build. Through a long process, IBM eventually granted us a
16 processor p690 (64 bit power4 architecture) with 64GB of RAM to
facilitate this research.
We needed a lisp compiler for this machine if we were going to be able
to put it to use. We got bids from all the vendors, and looked into
hiring programmers who were expert in the open source compilers. The
The Franz proposal was faster, cheaper, and more likely to succeed
than any of the other options. They delivered a 64 bit power4 / AIX
common lisp compiler in less than 6 weeks for about $50k.
Although that may seem pricey to a hobbiest, compare it to the process
of preparing the controlled environment room to house the machine --
that took six months and cost us more than $100k to do the renovations
(the slowest and most expensive part was humidification!).
Then recently while benchmarking some of our code, I discovered that
the p690 was no faster than a 2GHz pentium on some sequence analysis
tasks. Franz tech support (thanks, Duane!) sped up our code by a
factor of 3, and is still working on trying to understand the factors
that are responsible for the difference between the low-level chip
benchmarks and our observed performance and whether they have to do
with the lisp compiler.
I believe that we have the fastest, largest memory lisp machine ever
built in my lab, and we are doing some pretty interesting things with
it. Without Franz' help, we wouldn't have gotten there. The price is
definitely worth it for us.
Larry
--
Lawrence Hunter, Ph.D.
Director, Center for Computational Pharmacology
Associate Professor of Pharmacology, PMB & Computer Science
phone +1 303 315 1094 UCHSC, Campus Box C236
fax +1 303 315 1098 School of Medicine rm 2817b
cell +1 303 324 0355 4200 E. 9th Ave.
email: Larry....@uchsc.edu Denver, CO 80262
PGP key on public keyservers http://compbio.uchsc.edu/hunter
Gabe Garza
> Or differently: If you could add one feature of ACL to your
> favourite free CL, what would it be?
>
A company to commercially back it.
Commercial support *really matters*. If I deploy on a platform, both
myself and my employer require a company that can be held accountable
for it.
If I find a bug that's causing downtime, I want it fixed *right
now*--not when a volunteer gets bored on sunday afternoon and decides
to hack on the free implementation a bit.
If I need an enhancement, I can easily get it accomplished by paying
for it (easily as in: I send an email to f...@bar.com, and he takes it
from there. Not easy as in: I send an email to
foo-mai...@xyzzy.org and wait for estimates, further questions,
and "so-and-so may be able to do it" replies to roll in (and if a
free-implementation maintainer *did* do the work, he better have a
company set up to go through the corporate payment dance with us...)
If I leave, my employer needs a company to go to for references to
consultants, etc.
Gabe Garza
BK
> Only a minority of users subscribe to mailing lists or online forums.
Fair enough. But likewise, not all users pay their shareware fees
either. So, my assumption here is that - in respect of compilers -
there is likely to be a strong correlation between users who subscribe
to a mailing list and those who pay their shareware fees. I may be
wrong about that, though.
What do you reckon how many OpenMCL users there are? Are there any
reasonably trustworthy figures?
rgds
bk
Fernando Mato Mira
> If you could add one feature of ACL to your favourite free CL, what would
> it be?
The Java Tool Suite. And not in a free CL, in another commercial one
with better runtime licensing conditions.
Daniel Barlow
> If I find a bug that's causing downtime, I want it fixed *right
> now*--not when a volunteer gets bored on sunday afternoon and decides
> to hack on the free implementation a bit.
>
> If I need an enhancement, I can easily get it accomplished by paying
> for it (easily as in: I send an email to f...@bar.com, and he takes it
> from there. Not easy as in: I send an email to
> foo-mai...@xyzzy.org and wait for estimates, further questions,
That's straightforward enough. Send an email to in...@metacircles.com
and we'll do your SBCL bug fixes and enhancements. We can't
necessarily guarantee fixes *right now*, because some bugs take finite
time to fix and the work is not infinitely parallelisable (even
supposing an infinite number of SBCL hackers to call on) but can
certainly give them immediate priority.
> and "so-and-so may be able to do it" replies to roll in (and if a
> free-implementation maintainer *did* do the work, he better have a
> company set up to go through the corporate payment dance with us...)
Sure ... but if you're not registered for VAT or equivalent in an EC
member country, you might want to find out whether you can reclaim it.
-dan, expecting a flood of similar followups from Clozure, PMSF,
Scieneer, etc ...
--
http://www.cliki.net/ - Link farm for free CL-on-Unix resources
David Steuber
> Matthias <sp...@yourself.pl> writes:
>
> > Or differently: If you could add one feature of ACL to your
> > favourite free CL, what would it be?
> >
>
> A company to commercially back it.
>
> Commercial support *really matters*. If I deploy on a platform, both
> myself and my employer require a company that can be held accountable
> for it.
I hear this a lot. I personally have never seen software that comes
with a warranty but that doesn't mean it doesn't exist. The premise
behind Free software (I'm not going to get into the distinction
between Free and Open Source) is that *YOU* are the accountable
support mechanism. I know that is unacceptable in your case, but that
is the way it works.
> If I find a bug that's causing downtime, I want it fixed *right
> now*--not when a volunteer gets bored on sunday afternoon and decides
> to hack on the free implementation a bit.
From what I've been reading, Franz (just as an example) seems to be
providing extremely good support. On the other hand, I've dealt with
a major software tool vendor who just happens to develop the most
popular desktop OS. For a lot of money, they provide poor developer
support (it's not like the support staff aren't trying though). They
sure as heck won't fix a bug just for you. Maybe in the next
release.
> If I need an enhancement, I can easily get it accomplished by paying
> for it (easily as in: I send an email to f...@bar.com, and he takes it
> from there. Not easy as in: I send an email to
> foo-mai...@xyzzy.org and wait for estimates, further questions,
> and "so-and-so may be able to do it" replies to roll in (and if a
> free-implementation maintainer *did* do the work, he better have a
> company set up to go through the corporate payment dance with us...)
Refer to my paragraph above. When you have the source, and you are
free to modify it, you have more control over how things work out for
you. That freedom comes at the price for taking responsibility for
your product rather than passing the buck to your vendor. Even so,
although the business models seem to have trouble as they are, there
are outfits that specialize in custom support.
> If I leave, my employer needs a company to go to for references to
> consultants, etc.
I hope you document all your work because any company or individual
that provides those services also has the overhead of learning the
architecture of your codebase just as a new hire would.
I have seen the problems caused when a consulting firm creates custom
software (using commercial tools) and then leaves at the end of the
contract with no documentation or code comments in all their very
expensive work. The fact that they used Java vs Perl is of no comfort
whatsoever.
BTW, has anyone read Sun's license and warranty for Java? Everything
is disclaimed. Same goes for Microsoft, Borland, etc. It seems that
commercial Lisp vendors are doing what most software vendors refuse to
do if they Warrant their work against bugs, defects, and actually say
that their product is indeed fit for a particular purpose.
From my limited experience, the entire commercial support argument is
a pickled herring.
--
(describe 'describe)
David Steuber
> What do you reckon how many OpenMCL users there are? Are there any
> reasonably trustworthy figures?
I would not trust any numbers for users of _any_ software. This goes
doubly so for Free software. Just because a Linux box (cardboard, not
PC) has been sold by CompUSA doesn't mean that purchaser installed the
system or didn't take it to the skeet range after a week of
frustration.
I own two computers with Windows licenses. I do not use Windows.
Download counts and serveys are also meaningless.
--
(describe 'describe)
rcj.p...@physics.org
> ...
> 1. If I were a CL hobbyist, I wouldn't be living on consulting; if
> there were really consulting money coming in it would not be necessary
> to make up for the money I put into my hobby. In fact, I would consider
> a hobbyist almost definitionally to be one who isn't trying to make a
> living on that on which he loves to spend time and money.
>
> But, in fact, the true hobbyist tends to spend _tremendous_ amounts of
> money on his hobby; elsewhere on this newsgroup (perhaps on this thread?)
> someone mentioned hobbys that are extremely expensive, and not necessarily
> that poipular; one has to love the hobby in order to spend so much time
> and money on what they love to do. As I recall, the conversation got
> sidetracked toward rentals, as in videos (I think I would also place
> video games into the same category). However, I wouldn't tend to classify
> something that is easily rentable as a hobby - and of course it depends
> on the circumstances - it is hard for me to think of renting movies as
> anything other than a means to unwind (unless I were secretly harboring
> a desire to become the next Siskel & Ebert / Ebert & Roeper...). When
> I truly get into a hobby, the money I make on work has to be carefully
> divided so I don't starve while feeding my hobby...
>
I think although it is certainly true that hobbies are expensive, the
perception of cost is relative to the products available.
There used to be (I'm not sure if its still around) a computer
magazine in the UK called PC Plus and certainly in the early days it
was aimed squarely at the programming hobbyist. Each month it would
contain a CD giving away the previous edition of Borland Builder,
Delphi etc. along with programming tutorials. As I recall, they were
mostly full products for non-commercial usage.
Many of my friends would buy the magazine, play with the compilers and
write little utilities to use at home such as displaying thumbnails of
photographs etc. There is no way they are going to pay several
thousand dollars for a professional compiler product when they can
enjoy programing for next to nothing. Whereas they would consider
spending that amount of money on a new telescope etc.
Incidently, I don't think the ACL Trial edition is of much interest to
these people either. It is a restricted product, you can't give away
any utilities you write to your friends and the re-registering each
month is too annoying for occasional use. i.e. it isn't much fun.
I would have thought it would be in Franz's interest to attract as
many users as possible. It is fairly unlikely that working for a
company which is not already using Lisp that you will be sent on a
training course. Common Lisp requires a good deal of learning effort
to be effective, so a way to add new users and get them learning Lisp
is to attract just this hobbyist market.
> 2. If I were _living_ on my scattered consulting, it wouldn't be my
> hobby; I would want to make more money than I spent. This amounts to
> a profit-and-loss center, which is a similar model to that which a
> larger company would have. Contractual obligations would indeed be
> involved. And as I move an activity from hobbyist to carreer, it ceases
> to become something I can just pour my discretionary money and time into;
> I must scale back so that expenses are reduced below income.
>
Moving from hobbyist Lisp programmer to using it as a carreer is
extremely difficult. I've tried persuading several companies to use
ACL for projects. Most managers are not unreasonable and will allow
staff in certain cases (usually the less important projects) to choose
their tools, if only to keep them happy.
Again the cost is not really the problem, its the perception that
there are 'similar' products for far less money. It feels like you're
trying to persude them you really need a Porsche for your next company
car.
At my workplace we have subscriptions for Microsoft, which gives us
their compilers and Novell which comes with CodeWarrior licenses etc.
We do use other compilers, but they are a few hundred UK pounds which
doesn't make your manager blink.
The final body blow comes when you mention that Franz wants the sales
projections for the project to work out the royalties, and that
database support etc. which come as standard with other compilers are
optional extra's. (I don't know if its changed but IIRC the profiler
was an extra!)
IMHO Franz simply doesn't provide a route for Common Lisp programming
to move from hobby to carreer.
> So what about support? Support seems at first glance to be an unnecessary
> expense, and thus one which should be cut at any opportunity. And the
> free/open-source industry has given inexperienced programmers the
> impression that such support is completely unnecessary, since if the
> programmer has a problem, he/she can just go into the source and fix
> it. And if in your consulting contract you have all the time in the
> world to deviate from the primary goal of getting _XYZ_ product out
> for an indeterminate time in order to work on this side issue (because
> you have the source code and can thus do so) then more power to you.
> But more often than not, such deviations from the primary contract
> tend to tarnish your reputation of getting good product out _on_ _time_,
> and unless you've built that slop into your initial estimation, the
> excuse that "I had to go in and fix some of the free software I was
> using" doesn't necessarily cut it with the customer you are consulting
> for. Just be sure in that situation that you don't sign any contracts
> with penalty clauses for lateness...
>
> Instead it makes sense to have pieces of the project you are working on
> to be covered by support contract, so that if anything goes wrong with
> what you are using, you can get a quick fix or workaround (or lesson as
> to how to use correctly) from those who know it best, and you can then
> continue on quickly with your main goal without using up much of the
> built-in slop factor in your schedule.
>
I agree that, once established in a business, the cost of ACL is
insignificant and that purchasing support is absolutely essential.
I'd have thought support would be the area that Franz could make their
profits.
I wish Franz would market ACL to make it more attractive for smaller
projects and hobbyists, perhaps by only offering support on a per
incident basis.
Wouldn't it be great if everyone was using AllegroServe for running
their company's website? I don't think Franz should consider what
immediate profit they could gain from this, but know that they have
established themselves inside new companies that would otherwise never
touch Lisp. As we've agreed, once a product has become part of a
company it is natural to pay for support.
Tim Bradshaw
> That's straightforward enough. Send an email to in...@metacircles.com
> and we'll do your SBCL bug fixes and enhancements. We can't
> necessarily guarantee fixes *right now*, because some bugs take finite
> time to fix and the work is not infinitely parallelisable (even
> supposing an infinite number of SBCL hackers to call on) but can
> certainly give them immediate priority.
This misses the point. The issue for most significant commercial
deployments is not just that there is someone who they can pay to
provide support, but that the company is plausible. To be plausible,
it has to be big enough that it can, for instance, allocate resource
rapidly (so it needs to have a significant number of people), and it
needs to be big enough so its response to anything bad happening to a
contract is not to instantly go bankrupt leaving the client without
the financial penalty they were trying to extract, and without any
support to boot.
I am not saying *anything* against metacircles, but I bet you fail on
several of those counts (so does Cley). Even something as large as
Franz is probably questionable for a large client.
To take the opposite extreme - if I'm a customer with an important
system, then if I buy that system from Sun/IBM/... I *can* hold them
accountable - I can have a contract with financial teeth, and be
fairly sure that they will pay up.
I assume that's what Gabe meant by `held accountable' - if the
supporting organisation is not large enough then it's not accountable
because it is too vulnerable to perturbations of various kinds.
--tim
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
<rcj.p...@physics.org> wrote in message news:vfwkp1...@physics.org...
Prehaps Franz should try the RedHat approch: make the software lower priced
to attract the hobbyists, and charge a lot of $$$ for yearly support
contracts.
Or, maybe release older versions of Allegro CL at low costs to build up the
user base, and maybe even attact more companies to move up to the full
version.
I'm not trying to tell Franz what to do, I am just giving some advice about
how to increase their market size. I also think that Symbolics should
release Open Genera for use on Alpha emulators running on Linux for simular
reasons.
Or, maybe even go the OpenMCL route and release a scaled down version of
Allgro CL--just the ANSI Lisp system so that a whole new
bunch of students might pick up on Lisp.
If people using the free version need help from Franz--Franz could either
charge them a per incident fee, or contractual fee.
Once more people start using Lisp; I'd expect several to want to use Lisp at
work, and be able to demonstrate to management what Lisp can do, and why the
cost of Allegro CL is worth it.
Old Lispers die, or retire. New ones need to be attracted.
If you know of a place where I can get used older ANSI Lisp compilers
cheaply tell me; I don't need the lastest & greatest, just a system to learn
either ANSI or CLtL2 CL and generate executables.
If anyone can give a good argument about why making it hard for people who
want to learn Lisp to learn it will help Lisp usage out in the long run
please tell me--I don't think it will.
Kent M Pitman
> Prehaps Franz should try the RedHat approch: make the software lower priced
> to attract the hobbyists, and charge a lot of $$$ for yearly support
> contracts.
Franz has its own dollars on the line. They don't need you telling them
how to change their business model.
> I'm not trying to tell Franz what to do, I am just giving some advice about
> how to increase their market size.
This assumes that they aren't aware of what's going on in the world
and have made their choices out of ignorance. I think you radically
underestimate their market awareness, to an a degree I would be
insulted by if I were them. It sounds like you think they are new
entrants to the market who have ignored the way the market works, when
in fact they've been very economically successful continuously for
longer than any of the other Lisp vendors.
For any given change we on the outside suggest, it could be that it
would improve their position. I am not saying Franz has made the best
decisions (nor am I saying they haven't). I'm saying they are
competent individuals with their ear constantly to the ground and they
understand what their market is and what the risks are. It's their own
money, after all.
It's easy to risk someone else's money by telling them what would work
better, but if you really want to show what a great idea it is to run
a service business on free software, there are plenty of offerings
already out there--why not pick one up and run with it?
If the only way you or someone else can succeed in offering service on
Lisp is to use Allegro as the base, then maybe Allegro has value in it
worth charging for (which they are doing). If Allegro has no value worth
charging for, you'd think Franz would notice that no one was buying and
would make the choice to change business models without you telling them.
> I also think that Symbolics should release Open Genera for use on
> Alpha emulators running on Linux for simular reasons.
Of course you do. You didn't pay to obtain it. If you think it's worth
it to pay to obtain it and then to give it away, why not get some financial
backing, buy it, and then open source it yourself? Then the risk will be
on you. If you don't think you can do that, then maybe you understand why
someone else might not either.
> Or, maybe even go the OpenMCL route and release a scaled down version of
> Allgro CL--just the ANSI Lisp system so that a whole new
> bunch of students might pick up on Lisp.
Oh good. A whole bunch of people with no money. Yes, maybe then the revenues
would roll in a few years later. Oh, by the way, what revenues would the
staff be living on while they waited for this delayed effect?
As I understand it, MCL was able to do this because, to round numbers,
Digitool has no staff. Franz is not in that situation.
> If people using the free version need help from Franz--Franz could either
> charge them a per incident fee, or contractual fee.
And if it turns out they just say "Thanks for the fish." and don't get in
touch at all?
> Once more people start using Lisp; I'd expect several to want to use Lisp at
> work, and be able to demonstrate to management what Lisp can do, and why the
> cost of Allegro CL is worth it.
"several" is not a market.
And "once more people start using Lisp" is NOT an overnight process.
Lisp is slowly working its way up from the depths again at commercial
companies like Franz and Xanalys that were hit hard by AI Winter. On
limited resources, they have persisted against amazing odds to continue.
But they have the responsibility of continuing to pay their bills in that
transition, and we who do not have those bills to pay should not be so
cavalier about making sure that gets done.
Free software or not, the failure of a Lisp company will not look good
for Lisp. The REAL thing working against Lisp is old hatreds of it for
marginal reasons, and the energy behind that anti-PR engine is still alive
looking for events to fuel its prejudices so that extant service
organizations not using Lisp can feel comfortable that they won't be
faced with the prospect of learning something new and perhaps even being
destabilized by a new set of experts who already know it.
Daniel Barlow
> This misses the point. The issue for most significant commercial
> deployments is not just that there is someone who they can pay to
It misses your point. Gabe's point, on the other hand, seemed to be
in large part concerned with being dependent on people he has no
contract with, and who are only working on his problems as and when
they feel like it. _That_ problem can be overcome, and there are
plenty of consultancies around who can help to overcome it.
> provide support, but that the company is plausible. To be plausible,
> it has to be big enough that it can, for instance, allocate resource
> rapidly (so it needs to have a significant number of people), and it
> needs to be big enough so its response to anything bad happening to a
> contract is not to instantly go bankrupt leaving the client without
> the financial penalty they were trying to extract, and without any
> support to boot.
It's funny, but the first reply I was going to make to Gabe's post
opened with the words "It costs about £100 (call it $150) to start a
company in the UK. The presence of a company behind a piece of
software is not magic pixie dust". By instead posting a "it's not as
bad as you describe" article I think I've provoked you to follow up
with exactly the article I was initially going to write. Sorry. Next
time I'll do my own work.
Because everything you say there is true, but has basically not a lot
to do with whether the software is free or not. You can get Linux
support from IBM or from Oracle; you can get, say, Corman Lisp support
from Corman Technologies. One of these syatems is free, and it's not
the one for which the supporting organization is likely to be able to
fly a team of consultants out to your site tomorrow.
-dan
Tim Bradshaw
> From my limited experience, the entire commercial support argument is
> a pickled herring.
I take it you've never worked in an environment where there are
support contracts which get you someone on site within a couple of
hours, at any time of day or night, and on the phone instantly. These
environments (and contracts) exist, and they aren't that uncommon.
--tim
Christopher Browne
> Prehaps Franz should try the RedHat approch: make the software lower priced
> to attract the hobbyists, and charge a lot of $$$ for yearly support
> contracts.
That is an interesting model, but I cannot imagine that they are so
stupid that they have never considered its merits and demerits.
You seem inclined to ignore that there are demerits to it.
> Or, maybe release older versions of Allegro CL at low costs to build
> up the user base, and maybe even attact more companies to move up to
> the full version.
There is a _big_ disadvantage to this, namely that it would encourage
users to stay with older versions, and not pay for "service contracts"
or the like to keep up to date.
> I'm not trying to tell Franz what to do, I am just giving some
> advice about how to increase their market size. I also think that
> Symbolics should release Open Genera for use on Alpha emulators
> running on Linux for simular reasons.
The thing is, "market size" isn't necessarily a metric that is of any
particular merit.
Personally, I'd be more concerned with "profitability" as a metric.
And having products that are high-priced allows profitability at a
much lower level of sales.
It would be totally irrational for Franz to drop pricing by 90% on the
basis of the assumption that this would lead to an increase of so
little as a 1000% increase in sales.
In contrast, if Franz were to increase pricing by 50%, and only lose
25% of their customer base, this would likely be a net _win_ for
profitability. And if that got rid of three customers that had been
irritating to the support staff, that might be a double win.
> Or, maybe even go the OpenMCL route and release a scaled down
> version of Allgro CL--just the ANSI Lisp system so that a whole new
> bunch of students might pick up on Lisp.
That _is_ an idea with some merit, although it requires devoting some
developer effort that may not have a clear way of leading to sales
that would pay for the effort.
--
If this was helpful, <http://svcs.affero.net/rm.php?r=cbbrowne> rate me
http://www3.sympatico.ca/cbbrowne/lisp.html
The human race will decree from time to time: "There is something at
which it is absolutely forbidden to laugh."
-- Nietzche on Common Lisp
Tim Bradshaw
> It misses your point. Gabe's point, on the other hand, seemed to be
> in large part concerned with being dependent on people he has no
> contract with, and who are only working on his problems as and when
> they feel like it. _That_ problem can be overcome, and there are
> plenty of consultancies around who can help to overcome it.
Yes, what I was trying to say is that a contract is only worth as much
as the company it's with.
--tim
Thaddeus L Olczyk
>
>I think although it is certainly true that hobbies are expensive, the
>perception of cost is relative to the products available.
>
The first thing I would say is that talking about "amateur" Lisp
programming as a hobby, is just so much BS. By this standard
Paul Grahams much vaunted venture was done with a "hobbyist"
compiler. What programming I do that is amateur I do for one of
several reasons:
1) I need a utility.
2) A friend needs a utility.
3) I am do so to become a better programmer ( learn a new library ie
Corba, expat,... or a new language or a new technique ).
In each case the term hobbyist seems to trivialize the action.
( If a construction worker puts a new room in his home, or does a home
repair, is he a hobbyist? )
Looking at Robert Martin's first book, he learned OO in his spare
time, after becoming convinced that it was a good thing to know. As OO
became popular, he became ( regardless of whether you like what he
said ) a leading authority. Was he engaging in a hobby?
As for the cost of a hobby. I've bicycled for about 30 years. In that
time I haven't spent ,much more then half amount of money that two
licenses ( one for Windows, one for Linux ) cost. Much less the next
upgrade for when the OS's upgrade to the point where I have to buy the
upgrade.
I also collect pens. This can get very expensive. ( One example, my
favorite pen is:
http://www.worldlux.com/cgi-bin/showmodel.cgi?field0=S.T.%20Dupont&field1=Orpheo%20Vertigo&dept=PENS&collect=
)
I haven't in fifteen years spent the cost of two licenses.
I quote an observation from the book "Have Spacesuit Will Travel" by
Robert Heinlein. ( Haven't read that book in at least 20 years, God is
my mind a junk yard. ) The main character had a space suit he won in a
contest. He made a hobby of refurbishing it. At the end of Summer he
pointed out that he didn't spend much money on it. Mostly it involved
his time. In fact, since it took his time, he didn't spend money on
other things. That's the way most hobbies are.
There are of course some hobbies that cost a lot. Someone I knew
just poured money down the drain maintaining a Corvette ( never
mind that he didn't make it that fast, the dumbest part, it was from
the late 90's ). Some fountain pen users just go out and buy
$2000 pens all the time.
But those are "status symbol" hobbies. People spending the money
to show that they can. Does anyone think that people buying
ACL as a status symbol is a *good* thing? ( Of course it would
help as these people would probably not need any support. There
are stores out there, not pen stores, but stores that stock pens,
Usually stationary stores, that don't stock ink for the pens. )
Oh. There is a lesson to be learned from fountain pen companies.
In the early 90's the pen industry was flourishing. It wasn't
something that grew 50% every year, but it had a steadily growing
base.
Then Scheaffer stopped selling Targas ( went for $45-$60 ) and the
cheapest quality pens they sold went fro $200-$300. Other followed
suit. At the same time Parker released a couple of Limited Edition
pens ( the most memorable had parts made from the casings of
SS-20's and Pershing missiles, to commemorate the end of the cold war
), with high prices. Within a few years the basic entry level pen
( $50 to $150 ) virtually died.
Suddenly there were no low priced quality pens out there. Newbies
would have to make a major commitment, spend $300 for a pen. So there
were few newbies. Old timers started looking at the prices and went a
new way. Now they go to pen shows or antiquing. They even have a
term for cheap high quality pens. They call them "sumgai".
Result: Parker closed it's Janesville plant. Many pen stores are now
defunct. Colorado Pen's has been dissolved. Parker/Waterman were
sold to a company hoping to bring them back.
>There used to be (I'm not sure if its still around) a computer
>magazine in the UK called PC Plus and certainly in the early days it
>was aimed squarely at the programming hobbyist. Each month it would
>contain a CD giving away the previous edition of Borland Builder,
>Delphi etc. along with programming tutorials. As I recall, they were
>mostly full products for non-commercial usage.
They may not have come with all the tools/libraries provided for by
the commercial versions. They also come with a restriction that it
cannot be used for commercial purposes.
There is also another thing. Since these products are sold over the
shelf, you can go to computer sales where you can find these things
for reduced price ( I once got Delphi enterprise for $150 because a
company was liquidating. It usually costs $1500. ).
>
>Many of my friends would buy the magazine, play with the compilers and
>write little utilities to use at home such as displaying thumbnails of
>photographs etc. There is no way they are going to pay several
>thousand dollars for a professional compiler product when they can
>enjoy programing for next to nothing. Whereas they would consider
>spending that amount of money on a new telescope etc.
>
>Incidently, I don't think the ACL Trial edition is of much interest to
>these people either. It is a restricted product, you can't give away
>any utilities you write to your friends and the re-registering each
>month is too annoying for occasional use. i.e. it isn't much fun.
>
As I have pointed out in the past most professional programmers
learn a language by writing such utilities.
>I would have thought it would be in Franz's interest to attract as
>many users as possible. It is fairly unlikely that working for a
>company which is not already using Lisp that you will be sent on a
>training course. Common Lisp requires a good deal of learning effort
>to be effective, so a way to add new users and get them learning Lisp
>is to attract just this hobbyist market.
>
>> 2. If I were _living_ on my scattered consulting, it wouldn't be my
>> hobby; I would want to make more money than I spent. This amounts to
>> a profit-and-loss center, which is a similar model to that which a
>> larger company would have. Contractual obligations would indeed be
>> involved. And as I move an activity from hobbyist to carreer, it ceases
>> to become something I can just pour my discretionary money and time into;
>> I must scale back so that expenses are reduced below income.
>>
>
>Moving from hobbyist Lisp programmer to using it as a carreer is
>extremely difficult. I've tried persuading several companies to use
>ACL for projects. Most managers are not unreasonable and will allow
>staff in certain cases (usually the less important projects) to choose
>their tools, if only to keep them happy.
>
If I were hiring Lisp programmers I wouldn't hire anyone who has
learned Lisp just by going through Paul Graham's books. I would
insist on professional experience, or barring that a demonstration
that the person has programmed in nonprofesionally in Lisp for
several years, as well as many years experience in other languages.
This is not as much a problem if you are a Linux programmer, but
if you are a Windows programmer, it becomes more problematic.
The question is then how do you get that nonprofessional experience.
>Again the cost is not really the problem, its the perception that
>there are 'similar' products for far less money. It feels like you're
>trying to persude them you really need a Porsche for your next company
>car.
>
>At my workplace we have subscriptions for Microsoft, which gives us
>their compilers and Novell which comes with CodeWarrior licenses etc.
>We do use other compilers, but they are a few hundred UK pounds which
>doesn't make your manager blink.
>
>The final body blow comes when you mention that Franz wants the sales
>projections for the project to work out the royalties, and that
>database support etc. which come as standard with other compilers are
>optional extra's. (I don't know if its changed but IIRC the profiler
>was an extra!)
>
>IMHO Franz simply doesn't provide a route for Common Lisp programming
>to move from hobby to carreer.
>
Like the pen manufacturers, they are not hooking the newbies.
Think how easier it would be to sell ACL to some managers if it won
ICFP. But the people who enter ICFP are not going to pay for ACL
just to enter ( especially the people most likely to win ).
Frankly whatever we say doesn't matter much. Franz is going to do what
Franz is going to do. But I suspect that it would be better for Franz
if they did what CinCom ( VisualWorks ) does. They allow people to
download ST for noncommercial use. I've met five Smalltalk
programmers, never met a Lisp programmers. Of he five, three got
hooked by VW. The other two go around encouraging people to download
it and try it out. They even got a several pilot projects going at
companies. The only experience I've had with professional Lisp
projects in the area is companies moving away from it.
Rainer Joswig
"Marc Battyani" <Marc.B...@fractalconcept.com> wrote:
LispWorks is an very cool piece of software. I really like it.
It has a very complete Lisp implementation and development environment.
I'm pretty sure, lots of people will want the Mac OX X version
of LispWorks. :-)
OCID
"Tj" <tj_sc...@yahoo.com> wrote in message
news:ccc7084.03050...@posting.google.com...
> Here's the link I mentioned earlier.
>
http://www.amazon.com/exec/obidos/tg/detail/-/013026248X/002-1380390-0472843
>
> BTW, I notice you ignore Corman Lisp, which IIRC is priced low.
Some people already mentioned Corman Lisp so I did'nt. Its win32 only which
may be a problem
for some people because everyone does not use windoze, and I don't like the
fact that it compiles
before execution. Its only a personal preference but I am not a big fan of
the compile-debug-test
cycle. I prefer the write-test-write-test-write-test-compile-optimize cycle
more and Clisp works
fine for me on cygwin/2K, OS X, Linux and BSD.
OpenMCL I believe can interact with the Cocoa framework as well and thats
free too.
Take Care
>
> Tj
Christopher Browne
They generally involve either:
a) Server hardware, or
b) Database software,
and involve paying enough money per year that you are effectively
paying the salaries of all of those people that will be answering the
phone, 7x24, as well as paying for a few flights between here and
California.
If you /aren't/ paying >$100K/year for the support contract, then it
is unlikely that the "support" will be worth much of anything...
--
(concatenate 'string "cbbrowne" "@cbbrowne.com")
http://cbbrowne.com/info/lsf.html
No, I believe it's his real name... he's "unique".
Think Duh-fferent.
Henrik Motakef
> Some people already mentioned Corman Lisp so I did'nt. Its win32
> only which may be a problem for some people because everyone does
> not use windoze, and I don't like the fact that it compiles before
> execution. Its only a personal preference but I am not a big fan of
> the compile-debug-test cycle. I prefer the
> write-test-write-test-write-test-compile-optimize cycle more and
> Clisp works fine for me on cygwin/2K, OS X, Linux and BSD.
As far as I know (and I didn't use it much), Corman has a REPL just
like every other sane Lisp. It just implicitly compiles everything
behind the scenes even when you just type in interactivly, or load a
source file. The workflow is not impacted at all. SBCL does the same,
essentially.
Regards
Henrik
Erann Gat
<cbbr...@acm.org> wrote:
> In contrast, if Franz were to increase pricing by 50%, and only lose
> 25% of their customer base, this would likely be a net _win_ for
> profitability.
This ignores the long-term dynamics of economic decisions involving
infrastructure. People don't make decisions about infrastructure based
solely on a local cost-benfit analysis, they also look at the decisions
being made by others. Towards the end of the VHS-Betamax war many people
who would have personally preferred to buy a Betamax were buying VHS
merely because everyone else was doing it, and of course by so doing
became part of the everyone else who was doing it and so turned the
ultimate victory of VHS into a self-fulfilling prophecy.
Franz raising its prices may very well raise its short-term profits, but
very likely at the expense of long-term growth. (And vice versa.)
E.
Peter Seibel
> The final body blow comes when you mention that Franz wants the
> sales projections for the project to work out the royalties, and
> that database support etc. which come as standard with other
> compilers are optional extra's. (I don't know if its changed but
> IIRC the profiler was an extra!)
As someone who's had some dealings with various people at Franz,
including some of the sales guys, I suspect that they may be more
flexible than they seem "officially". I--obviously--can't speak for
Franz and I might be totally wrong about this but here are several
ways that you might be able to get access to more of their software
than you get with the Trial edition for no (immediate) cost. Contact
them and ask to talk to a sales rep. Then say:
- I really want to use ACL on a project here at company X but I
can't sell it to management unless I can build a prototype that
includes features X, Y, and Z that aren't available in the Trial
edition.
- I'm an entrepreneur and I'm about to start a software startup. I'm
really not sure what our basic business model is going to be let
alone what our projected revenues, etc. are. We're still trying to
decide whether we're going to be a product company or a service
company. But we are exploring technology choices. I want to show
my co-founders how easy it is to whip up something that does X, Y,
and Z in Common Lisp but I need some features that aren't included
in the trial edition.
- I'm a student. As part of my research, I'm building an app that
requires very large, in-memory data sets. The Trial edition
doesn't work for me because of the heap limit. I'm not sure if
anything is going to come of this research but if it works out I
may be able to turn it into a product.
From the conversations I've had with their sales guys, I suspect that
if you really fall into any of these categories or any other category
that lets you make the case "Here's why I can't pay you now, but
here's the circumstances where I might be able to" they will be
interested in talking to you. They understand that they need to find
new customers and that not everyone who might be willing and able to
pay for Lisp in the future is necessarily able to now.
Obviously the more immanent the possibility of you paying them, the
more excited they are likely to be. But it doesn't cost them much to
give you you a time-limited license to their Enterprise Edition, as
long as they have a relation with you that gives them some confidence
that you aren't going to take advantage of that license to somehow
build a business using their software without paying them.
The good news about the high price of ACL is that they presumably are
prepared for the the cost-per-sale to be non-trivial. That is, they
actually have sales people who's job is to spend time dealing with
prospective customers and figuring out, "What's it going to take to
get you into a new license today?" Whereas, if they were selling the
thing for $19.99 on the web, it'd be a take it or leave it offer
simply because there'd be no sales force to deal with.
Now, this doesn't help the pure hobbyist who doesn't have any real
expectation of ever paying Franz any money. Presumably the Trial
edition is for those folks. Though even there, since most hobbyist
programmers are also professional programmers, you may have a case you
can make:
- I'm a Java programmer at company Foo. I've been playing around
with the Trial edition of ACL and love it. I've been trying to
convince some of my fellow programmers that Lisp is the greatest
but am having a hard time. I have a Java application I wrote in my
spare time--not for the company. If I could rewrite it in Lisp and
show them how much nicer it is, I think that would convince some
of these guys, but I can't do it on the Trial edition because of
X, Y, and Z. If I could get a license for the Enterprise edition I
think I could do it easily and it would be a good demonstration of
features A, B, and C of the Enterprise Edition that I'd be happy
to contribute to opensource.franz.com. For that matter, if
rewriting this app in Lisp is as easy as I expect it to be, I may
have to go find a Lisp job. Can you help me out?
No guarantees that any of these approaches will work, but it can't
hurt to try. The important thing to keep in mind is that you need to
be offering Franz something in return for whatever you want from them.
While offering them a P.O. is the most direct form, they also
presumably understand other forms of potential value--from the
possibility of a future P.O. to the potential of getting another
enthusiastic Lisp evangelist.
-Peter
--
Peter Seibel pe...@javamonkey.com
The intellectual level needed for system design is in general
grossly underestimated. I am convinced more than ever that this
type of work is very difficult and that every effort to do it with
other than the best people is doomed to either failure or moderate
success at enormous expense. --Edsger Dijkstra
Christopher Browne
I agree that Things Are More Complicated than they seem.
But the point I was trying to make was not that this was necessarily a
perfect strategy, but rather that it is just as plausible for a price
/increase/ to be rational as it is for price /decreases/ to be
rational.
We keep hearing the broken record of:
"If they started giving the product away, lots of people would use
it, and they could then make lots of money."
(Well, perhaps without the "make lots of money part.")
The assumption being that it is an unalloyed benefit to gain market
share.
I am objecting, on the basis that:
a) It is certanly *not* unalloyed, and
b) I would suggest that it may not even be a benefit at all.
--
(concatenate 'string "cbbrowne" "@acm.org")
http://www.ntlug.org/~cbbrowne/x.html
You can lead a horse to water, but if you can get him to swim on his
back, you've got something.
Erann Gat
<cbbr...@acm.org> wrote:
OK, I'll buy that. :-)
Ultimately, the proof is in the bottom line. Franz is still around so
they must be doing something right.
E.
Daniel Barlow
> download ST for noncommercial use. I've met five Smalltalk
> programmers, never met a Lisp programmers. Of he five, three got
i don't know where you live, but you might want to check out
http://alu.cliki.net/Local and see if there's a local Lisp User Group
near where you are. If not, try posting here and ask.
> hooked by VW. The other two go around encouraging people to download
> it and try it out. They even got a several pilot projects going at
> companies. The only experience I've had with professional Lisp
> projects in the area is companies moving away from it.
David Steuber
> "Franz Kafka" <Symbolics _ XL1201 _ Sebek _ Budo _ Kafka @ hotmail . com> writes:
>
> > Prehaps Franz should try the RedHat approch: make the software lower priced
> > to attract the hobbyists, and charge a lot of $$$ for yearly support
> > contracts.
>
> Franz has its own dollars on the line. They don't need you telling them
> how to change their business model.
As well meaning as Franz's suggestions are, they are also not
necessary. As you say, Franz (why we need seperate namespaces) can
quite easily go with whatever business model they like. As they
should. On Debian, there is a package called cl-acl-compat. Here is
the description for it:
A thin compatibility-layer that emulates Allegro Common Lisp. It is used
to support a number of Allegro's open-source packages.
.
This package requires cmucl-graystream to work with CMUCL. But, that
package is not required to allow cl-acl-compat to be installed on
platforms that do not support CMUCL.
A hacker like me can learn Lisp with a free implementation (I hope).
If I were to find myself able to do cool stuff that I was sure I could
sell, then I probably would look to Franz for "industrial grade" Lisp
and support. I would not have to waste Franz's time or money (or my
own) to "play" with Lisp and do it with ACL. BTW, Debian also allows
you to install ACL.
I could also fork a free implementation for my own purposes. That may
prove to be more expensive for me than going with Franz. I haven't
done the cost/benefit analysis. However, it is certainly hubris on my
part to assume I can just whip together something with all the
features of ACL when I am just one person who is just starting out in
Lisp. I'm not making that assumption ;-)
--
(describe 'describe)
Tim Bradshaw
> If you /aren't/ paying >$100K/year for the support contract, then it
> is unlikely that the "support" will be worth much of anything...
We pay well under 1000 for HW support that works (I think in fact we
pay well under 1000 total for 2 machines, with another (older) one
being about $1000).
I dunno how much really serious contracts cost, but given the price of
the machines involved in ones I've dealt with, I'd guess under 10k
(and I've made calls on those contracts at 6PM, and discussed whether
or not we wanted someone there between midnight and 4AM, and they
weren't even the most serious kind of contract).
So: sorry, that's wrong.
--tim
Tj
> Some people already mentioned Corman Lisp so I did'nt.
I didn't reply to your message, but Google thought I did. (It was a
response to the original msg.) I thought Google was just showing it
wrong. Mea culpa.
Tj
Patrik Nordebo
<olc...@interaccess.com> wrote:
> Think how easier it would be to sell ACL to some managers if it won
> ICFP. But the people who enter ICFP are not going to pay for ACL
> just to enter ( especially the people most likely to win ).
I can't think of a single manager I've known who whould care one whit
whether ACL has won ICFP or not. Those who are technical enough to
care at all about ICFP would likely realise it has much more to do
with the programmers involved than the language, and those who aren't
technical have no idea what an ICFP is or why they would care.[1]
Where do you find these managers?
[1] Disclaimer: I once took part in a failed attempt to do the ICFP
contest problem in Lisp. I blame me, not Lisp, for the failure.
Ingvar Mattsson
> Martha Stewart called it a Good Thing whenTim Bradshaw <t...@cley.com>wrote:
> > * David Steuber wrote:
> >> From my limited experience, the entire commercial support argument is
> >> a pickled herring.
>
> > I take it you've never worked in an environment where there are
> > support contracts which get you someone on site within a couple of
> > hours, at any time of day or night, and on the phone instantly.
> > These environments (and contracts) exist, and they aren't that
> > uncommon.
>
> They generally involve either:
> a) Server hardware, or
> b) Database software,
I've been on the other end of similar contracts. I was at the time
working for a company in the document management software biz and I
had multiple instances of "customer called, you need to be at the
customer's site tomorrow and untwist the knickers they've convinced
the DM server to twist. Call us at 11 am if you think you need a
hotel. Be there at 9 am."
Not quite "throw me onto an airplane to another country", and not
quite "there will be someone on-site". I usually either talked to the
local sysadmin or checked if I thought emote untwisting was feasible
before "get on a train early in the morning" was an option.
No, not *all* customers had that level of service.
//Ingvar
--
When in douFNORD! This signature has been hi-jacked by Fnord Information
systems, to fnordprovide you with unfnordlimited information.
_ XL1201 _ Sebek _ Budo _ Kafka @hotmail.com Franz Kafka
"Ingvar Mattsson" <ing...@cathouse.bofh.se> wrote in message
news:87y9118...@gruk.tech.ensign.ftech.net...
>
> No, not *all* customers had that level of service.
>
Not every Lisp user needs that level of service.
A student learning Lisp would be happy with
e-mail support. When you are learning a lang.
you don't expect things to be up yesterday.
People who don't need commerical grade support
but want a commerical grade compiler should pay
less?