http://www.joelonsoftware.com/items/2006/09/01.html
The best part:
"Oh and I know Paul told you that he made his app in Lisp and then he
made millions of dollars because he made his app in Lisp, but honestly
only�two people ever believed him and, a complete rewrite later, they
won't make that mistake again."
Well, you've come to the wrong place if you want to trash Lisp.
And why post something this here at all?
If you don't like Lisp, then fine, don't use it.
Anonymous coward #673 wrote:
> http://www.joelonsoftware.com/items/2006/09/01.html
>
> The best part:
>
> "Oh and I know Paul told you that he made his app in Lisp and then he
> made millions of dollars because he made his app in Lisp, but honestly
> only two people ever believed him and, a complete rewrite later, they
> won't make that mistake again."
"dpapathanasiou" <···················@gmail.com> writes:
> Well, you've come to the wrong place if you want to trash Lisp.
Well, you've come to the wrong place if you want to top post :-).
--
This post uses 100% post consumer electrons and 100% virgin photons.
At 2.6 miles per minute, you don't really have time to get bored.
--- Pete Roehling on rec.motorcycles
I bump into a lot of veteran riders in my travels.
--- David Hough: Proficient Motorcycling
Anonymous coward #673 wrote:
> http://www.joelonsoftware.com/items/2006/09/01.html
>
> The best part:
>
> "Oh and I know Paul told you that he made his app in Lisp and then he
> made millions of dollars because he made his app in Lisp, but honestly
> only two people ever believed him and, a complete rewrite later, they
> won't make that mistake again."
I think Joel is a little bit out of inspiration. He has written much
better articles than that, without needing to thrash anybody or
anything.
Regards,
Jurgen
Anonymous coward #673 wrote:
> http://www.joelonsoftware.com/items/2006/09/01.html
>
> The best part:
>
> "Oh and I know Paul told you that he made his app in Lisp and then he
> made millions of dollars because he made his app in Lisp, but honestly
> only two people ever believed him and, a complete rewrite later, they
> won't make that mistake again."
Perhaps you should read his article "Can Your Programming Language Do
This?"
http://www.joelonsoftware.com/items/2006/08/01.html
-- S
On 1 Sep 2006 13:03:49 -0700, ·········@juno.com:
>Perhaps you should read his article "Can Your Programming Language Do
>This?"
>
>http://www.joelonsoftware.com/items/2006/08/01.html
[Note: this message doesn't contains anything about my opinion; I'm
merely trying to analyze Joel's POV.]
What Joel says thourough his articles is something like: "Lisp is very
elegant, and contains lots of ideas. Pick those ideas and put them
into another language to make working programs."
Note that in that article, he doesn't speak much about Lisp; he just
says that some functions are named from their Lisp counterpart.
What he describes can be done in several languages. If I'm not
mistaken, you can do that in PHP and Javascript, and it's not that
complicated to do (nearly) the same in C++.
On Fri, 01 Sep 2006 09:33:15 -0700, Anonymous coward #673 wrote:
> http://www.joelonsoftware.com/items/2006/09/01.html
>
> The best part:
>
> "Oh and I know Paul told you that he made his app in Lisp and then he
> made millions of dollars because he made his app in Lisp, but honestly
> only two people ever believed him and, a complete rewrite later, they
> won't make that mistake again."
Who is Joel, and why should anyone care (more about) what he thinks?
..and who are you?
--
Lars Rune Nøstdal
http://lars.nostdal.org/
In article <······························@gmail.com>,
Lars Rune Nøstdal <···········@gmail.com> wrote:
> On Fri, 01 Sep 2006 09:33:15 -0700, Anonymous coward #673 wrote:
>
> > http://www.joelonsoftware.com/items/2006/09/01.html
> >
> > The best part:
> >
> > "Oh and I know Paul told you that he made his app in Lisp and then he
> > made millions of dollars because he made his app in Lisp, but honestly
> > only two people ever believed him and, a complete rewrite later, they
> > won't make that mistake again."
>
> Who is Joel, and why should anyone care (more about) what he thinks?
http://en.wikipedia.org/wiki/Joel_Spolsky
> ..and who are you?
I am nobody.
On Fri, 01 Sep 2006 19:55:47 +0200, Anonymous coward #673
<·······@devnull.com> wrote:
>>
>> Who is Joel, and why should anyone care (more about) what he thinks?
>
> http://en.wikipedia.org/wiki/Joel_Spolsky
>
Created Excel BASIC ... lol
Incidentally he is partly right.
Yahoo did rewritee ViaWeb in C.
But they do almost all their work in C so it
is easier for them to maintain that way I expect.
This is probably more due do their ineptitude in
Lisp that a defficiency of CL per se.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
John Thingstad <··············@chello.no> wrote:
> On Fri, 01 Sep 2006 19:55:47 +0200, Anonymous coward #673
><·······@devnull.com> wrote:
>
>>>
>>> Who is Joel, and why should anyone care (more about) what he thinks?
>>
>> http://en.wikipedia.org/wiki/Joel_Spolsky
>>
>
> Created Excel BASIC ... lol
>
> Incidentally he is partly right.
> Yahoo did rewritee ViaWeb in C.
> But they do almost all their work in C so it
> is easier for them to maintain that way I expect.
> This is probably more due do their ineptitude in
> Lisp that a defficiency of CL per se.
And of course, in the end they had to prove that Greenspun was right *g*
--
"Opportunity is missed by most people because it is dressed in overalls and
looks like work." -- Thomas A. Edison
In article <······························@gmail.com>,
Lars Rune Nøstdal <···········@gmail.com> wrote:
> On Fri, 01 Sep 2006 09:33:15 -0700, Anonymous coward #673 wrote:
>
> > http://www.joelonsoftware.com/items/2006/09/01.html
> >
> > The best part:
> >
> > "Oh and I know Paul told you that he made his app in Lisp and then he
> > made millions of dollars because he made his app in Lisp, but honestly
> > only two people ever believed him and, a complete rewrite later, they
> > won't make that mistake again."
>
> Who is Joel, and why should anyone care (more about) what he thinks?
Joel is Joel Spolsky. Among other things, he writes a widely read tech
blog. I don't know why people care about what he thinks, but the fact
of the matter is that they do.
rg
On Fri, 01 Sep 2006 11:01:00 -0700, Ron Garret
<·········@flownet.com>:
>I don't know why people care about what he thinks
I don't know, either. What he thinks is not important; his articles
are often interesting because they force people to have a different
view on seemingly-evident things.
Lars Rune Nøstdal skrev:
> On Fri, 01 Sep 2006 09:33:15 -0700, Anonymous coward #673 wrote:
>
> > http://www.joelonsoftware.com/items/2006/09/01.html
> >
> > The best part:
> >
> > "Oh and I know Paul told you that he made his app in Lisp and then he
> > made millions of dollars because he made his app in Lisp, but honestly
> > only two people ever believed him and, a complete rewrite later, they
> > won't make that mistake again."
>
> Who is Joel, and why should anyone care (more about) what he thinks?
>
> ..and who are you?
Oh, and I (or you) forgot the most important thing; what do _you_
think? :)
--
Lars Rune Nøstdal
http://lars.nostdal.org/
From: Pascal Costanza
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <4lrbc6F38kjhU1@individual.net>
Anonymous coward #673 wrote:
> http://www.joelonsoftware.com/items/2006/09/01.html
>
> The best part:
>
> "Oh and I know Paul told you that he made his app in Lisp and then he
> made millions of dollars because he made his app in Lisp, but honestly
> only two people ever believed him and, a complete rewrite later, they
> won't make that mistake again."
You have cited the wrong part, which means that you have probably missed
the point. The right part - indeed, the only part where the author
actually claims knowledge - is this:
"What I do know for sure, though, is two things:
* People all over the world are constantly building web applications
using .NET, using Java, and using PHP all the time. None of them are
failing because of the choice of technology.
* All of these environments are large and complex and you really need at
least one architect with serious experience developing for the one you
choose, because otherwise you'll do things wrong and wind up with messy
code that needs to be restructured."
What he doesn't seem to know is that people are also building web
applications in Lisp, Scheme, Smalltalk - those are tho ones I know for
sure - and probably many others. Considering this, the only really
relevant part is that you need developers with serious experience in at
least one particular technology. End of story.
Indeed, if you are an inexperienced Lisp programmer, you will probably
fail, at least the first time around. That's true for any language
and/or combination of technologies, because _programming is hard_.
It's true that Joel Spolsky makes claims that go beyond this simple
observation, but they are unfounded.
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
On Fri, 01 Sep 2006 15:07:31 -0400, Ken Tilton <·········@gmail.com>:
>"Squeak and Lisp and OCaml...are totally, truly brilliant programming
>languages worthy of great praise."
>
>Some nail. Oh, sorry, some guy named Spolsky wrote that:
To summarize, "Lisp is a great language, that you shouldn't use to
make applications".
I don't know enough Lisp to know how true that is, but lots of people
seem to believe it -- just look at some recent threads in this
newsgroup...
Fabien LE LEZ wrote:
> On Fri, 01 Sep 2006 15:07:31 -0400, Ken Tilton <·········@gmail.com>:
>
> >"Squeak and Lisp and OCaml...are totally, truly brilliant programming
> >languages worthy of great praise."
> >
> >Some nail. Oh, sorry, some guy named Spolsky wrote that:
>
> To summarize, "Lisp is a great language, that you shouldn't use to
> make applications".
> I don't know enough Lisp to know how true that is, but lots of people
> seem to believe it -- just look at some recent threads in this
> newsgroup...
I just read the article and some of the comments to it, and some of
this thread.
Never is more true, than on the internet, that when all is said and
done, more is said than done.
I mean really, Joel is a good writer, he entertains and interests lots
of people and he has at least two useful applications that earn him a
living. So most people in this thread cannot touch that.
And regardless of his opinions on lisp, it's a great language for web
development for various reasons. If you want to write a great web app
in lisp you will be able to. If you want to write a great web app in
Ruby and Ruby on rails you can do that too. But stop talking and start
doing.
On 1 Sep 2006 15:14:38 -0700, "justinhj" <········@gmail.com>:
>And regardless of his opinions on lisp, it's a great language for web
>development for various reasons.
Which are?
>If you want to write a great web app in lisp you will be able to.
If I want to write a web app in Lisp, and then let someone else take
care of it, will I find a Lisp programmer easily?
>But stop talking and start doing.
I'm pretty lazy, so I prefer to think a little bit before trying.
And discussion is one way to find the best language/framework/method
for a program.
Fabien LE LEZ wrote:
> I'm pretty lazy, so I prefer to think a little bit before trying.
> And discussion is one way to find the best language/framework/method
> for a program.
In this case, discussion is a waste of time.
Mind you, discussion between open-minded people concentrating on the
technical (and social, like are there programmers who know it)
strengths and weaknesses probably wouldn't be. But usually, for values
of usual asymptotically nearing 100 per cent, discussion about best
language/framework/method devolves into religion.
On 1 Sep 2006 21:14:04 -0700, "Christopher Koppler"
<········@chello.at>:
>But usually, for values
>of usual asymptotically nearing 100 per cent, discussion about best
>language/framework/method devolves into religion.
I quite don't get what an asymptote is doing here, but I see your
point.
However, discussing with evangelists of one language can be
instructive too: you can ask them the positive points of that
language, and what is missing from the list is the negative points.
I'll have to make quite a big website, and I'm trying to sort out the
different possibilities. On comp.lang.lisp I certainly won't get an
answer to the question "Which language shall I choose"; but the thread
"Website in LISP?" gave me quite an insight on the advantages and
drawbacks of Lisp in that domain.
Fabien LE LEZ <········@gramster.com> wrote:
> On 1 Sep 2006 15:14:38 -0700, "justinhj" <········@gmail.com>:
>>If you want to write a great web app in lisp you will be able to.
>
> If I want to write a web app in Lisp, and then let someone else take
> care of it, will I find a Lisp programmer easily?
I can't speak for Lisp, but Common Lisp is really easy to learn.
Compare this with Haskell. :-)
Any _real_ programmer should be able to pick up the language.
Just try to find a programmer instead of a single-language
code-monkey.
--
Web (en): http://www.no-spoon.de/ -*- Web (de): http://www.frell.de/
Stefan Scholl wrote:
> Fabien LE LEZ <········@gramster.com> wrote:
>
>>On 1 Sep 2006 15:14:38 -0700, "justinhj" <········@gmail.com>:
>>
>>>If you want to write a great web app in lisp you will be able to.
>>
>>If I want to write a web app in Lisp, and then let someone else take
>>care of it, will I find a Lisp programmer easily?
>
>
> I can't speak for Lisp, but Common Lisp is really easy to learn.
> Compare this with Haskell. :-)
>
> Any _real_ programmer should be able to pick up the language.
>
> Just try to find a programmer instead of a single-language
> code-monkey.
>
>
"Code-monkeys" are encouraged by business. Business would
rather have brain-damaged warm-bodies to fill up the
resources needed for a team than smart developers.
--
goose
Have I offended you? Send flames to ····@localhost
real email: lelanthran at gmail dot com
website : www.lelanthran.com
Fabien LE LEZ wrote:
> On 1 Sep 2006 15:14:38 -0700, "justinhj" <········@gmail.com>:
> >And regardless of his opinions on lisp, it's a great language for web
> >development for various reasons.
>
> Which are?
lisp maps very well to html. using a library like lml2 you can see that
very clearly. lisp code using macros looks almost identical to xml and
html, with one difference, it is code. You can easily intermingle lisp
evaluations with html style markup. You can also compile the web pages
so the are not interpreted at runtime.
Since you can implement continuations quite easily in lisp, you can use
a continuation style for maintaining state. What the user sees then, is
a live lisp program, running on his browser.
There are free web servers like araneida and allegro serve, or you can
use mod_lisp to run lisp programs serving web content on apache1 or 2.
You can also hook directly to mysql using cl-sql.
Unlike the web framework ruby on rails, the production system can be
fully debuggable. You can pause the server, recompile functions, change
data and restart it.
> >If you want to write a great web app in lisp you will be able to.
>
> If I want to write a web app in Lisp, and then let someone else take
> care of it, will I find a Lisp programmer easily?
>
> >But stop talking and start doing.
>
> I'm pretty lazy, so I prefer to think a little bit before trying.
> And discussion is one way to find the best language/framework/method
> for a program.
Thinking is good, I wouldn't dispute that. Doing is better. Build
mini-systems in various ways... using python and the libraries you can
get for that, then again with ruby on rails, maybe even .net. That's a
great way to learn and it should take no more than a few days to put
together mini test systems in each language so you get a feel for what
each has to offer.
Justin
On 3 Sep 2006 18:11:49 -0700, "justinhj" <········@gmail.com>:
>Thinking is good, I wouldn't dispute that. Doing is better.
But doing what?
If I try to make a web app in Lisp right now, I'll find it very
difficult, and the result won't be good. Not because of Lisp, but
because I don't know much of that language.
OTOH, if I try the same in C++, I'll find it far easier -- and get the
impression that C++ is better than Lisp for web apps.
On Mon, 04 Sep 2006 03:23:11 +0200, Fabien LE LEZ <········@gramster.com>
wrote:
> On 3 Sep 2006 18:11:49 -0700, "justinhj" <········@gmail.com>:
>
>> Thinking is good, I wouldn't dispute that. Doing is better.
>
> But doing what?
> If I try to make a web app in Lisp right now, I'll find it very
> difficult, and the result won't be good. Not because of Lisp, but
> because I don't know much of that language.
>
> OTOH, if I try the same in C++, I'll find it far easier -- and get the
> impression that C++ is better than Lisp for web apps.
>
Well I started with PHP which has a lot of good books for beginners.
When you have a good working knowledge of web programming then you can
go over to applying it to Lisp.
Indeed most of the documentation is either a user guide or a reference.
There aren't any good and comprehensive tutorials of starting web
programming that I am aware of.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
On Mon, 04 Sep 2006 12:21:55 +0200, "John Thingstad"
<··············@chello.no>:
>When you have a good working knowledge of web programming then you can
>go over to applying it to Lisp.
Yeah, but I still have to actually learn the language.
And since learning Lisp means spending several weeks being
unproductive, I have to convince my boss it's a good investment.
Fabien LE LEZ <········@gramster.com> writes:
> I don't know enough Lisp to know how true that is, but lots of people
> seem to believe it -- just look at some recent threads in this
> newsgroup...
Much of Emacs is written in its own primitive dialect of Lisp. It is
a popular enough editor that the vi vs emacs wars are still going with
no end in sight.
SBCL is written in Common Lisp. So is OpenMCL.
AutoCAD contains a fair bit of its own dialect of Lisp.
There's actually lots of stuff. Even some web apps.
--
This post uses 100% post consumer electrons and 100% virgin photons.
At 2.6 miles per minute, you don't really have time to get bored.
--- Pete Roehling on rec.motorcycles
I bump into a lot of veteran riders in my travels.
--- David Hough: Proficient Motorcycling
In article <··············@individual.net>,
Pascal Costanza <··@p-cos.net> wrote:
> What he doesn't seem to know is that people are also building web
> applications in Lisp, Scheme, Smalltalk
Of course he knows that. What many in the Lisp community do not seem to
understand is that the mere fact that people use Lisp is not in and of
itself a good thing. The problem in the real world is not that no one
uses it -- the problem is that no one [1] is making any money using it.
Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
failures (1) is greater than the number of high-profile Lisp successes
(0). Having people using Lisp and consistently losing money is much,
much worse than not having people using it at all.
rg
[1] To a reasonable approximation.
Ron Garret wrote:
> Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> failures (1) is greater than the number of high-profile Lisp successes
> (0). Having people using Lisp and consistently losing money is much,
> much worse than not having people using it at all.
I have very little interest in whether Lisp is suitable for commercial
applications (since I don't write them and don't want to write them),
but I don't see how one high-profile failure in eight years is a can be
a sign of anything *consistent*.
Cheers, Matt
In article <······················@h48g2000cwc.googlegroups.com>,
·········@gmail.com wrote:
> Ron Garret wrote:
>
> > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > failures (1) is greater than the number of high-profile Lisp successes
> > (0). Having people using Lisp and consistently losing money is much,
> > much worse than not having people using it at all.
>
> I have very little interest in whether Lisp is suitable for commercial
> applications (since I don't write them and don't want to write them),
That is precisely the problem.
> but I don't see how one high-profile failure in eight years is a can be
> a sign of anything *consistent*.
The one failure is not nearly as significant as the zero successes.
rg
Ron Garret wrote:
> In article <······················@h48g2000cwc.googlegroups.com>,
> ·········@gmail.com wrote:
> > Ron Garret wrote:
> > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > > failures (1) is greater than the number of high-profile Lisp successes
> > > (0). Having people using Lisp and consistently losing money is much,
> > > much worse than not having people using it at all.
> > I have very little interest in whether Lisp is suitable for commercial
> > applications (since I don't write them and don't want to write them),
> That is precisely the problem.
I don't see how. People do lots of things other than commerce, and
still need programming languages to do them. I'm quasi-convinced that
CL is well-suited to doing many things that I do, but those things have
nothing to do with writing lucrative web applications. I know
Mathematica is well-suited to doing other stuff I do, and it's hardly a
problem that no one writes web apps in *that*.
Cheers,
Matt
In article <························@i3g2000cwc.googlegroups.com>,
·········@gmail.com wrote:
> Ron Garret wrote:
>
> > In article <······················@h48g2000cwc.googlegroups.com>,
> > ·········@gmail.com wrote:
>
> > > Ron Garret wrote:
>
> > > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > > > failures (1) is greater than the number of high-profile Lisp successes
> > > > (0). Having people using Lisp and consistently losing money is much,
> > > > much worse than not having people using it at all.
>
> > > I have very little interest in whether Lisp is suitable for commercial
> > > applications (since I don't write them and don't want to write them),
>
> > That is precisely the problem.
>
> I don't see how. People do lots of things other than commerce,
Yes. Those are called "hobbies." That Lisp may be suitable for hobby
projects is not in dispute.
> still need programming languages to do them. I'm quasi-convinced that
> CL is well-suited to doing many things that I do, but those things have
> nothing to do with writing lucrative web applications. I know
> Mathematica is well-suited to doing other stuff I do, and it's hardly a
> problem that no one writes web apps in *that*.
The reason it's not a problem is that people use Mathematica to support
other profitable enterprises. The reason Joel focuses on Web apps (I
think) is just because the Web seems to be where most of the lucrative
opportunities are to be found nowadays.
rg
Ron Garret wrote:
> In article <························@i3g2000cwc.googlegroups.com>,
> ·········@gmail.com wrote:
[...]
> > I don't see how. People do lots of things other than commerce,
> Yes. Those are called "hobbies."
Sometimes. Sometimes those are called other things. In my case, I
started Lisp as a "hobby"[1], but have begun experimenting with it as a
research tool at work.
[...]
> > I know Mathematica is well-suited to doing other stuff I do, and it's hardly a
> > problem that no one writes web apps in *that*.
> The reason it's not a problem is that people use Mathematica to support
> other profitable enterprises.
I strongly suspect that the majority of the enterprises people support
with Mathematica are of the non-profit variety.
Cheers, Matt
[1] Hell, that's how I got started with Mathematica back in my college
days....
In article <·······················@m73g2000cwd.googlegroups.com>,
·········@gmail.com wrote:
> Ron Garret wrote:
> > In article <························@i3g2000cwc.googlegroups.com>,
> > ·········@gmail.com wrote:
> [...]
> > > I don't see how. People do lots of things other than commerce,
>
> > Yes. Those are called "hobbies."
>
> Sometimes. Sometimes those are called other things. In my case, I
> started Lisp as a "hobby"[1], but have begun experimenting with it as a
> research tool at work.
> [...]
If you get paid for your research work then you are engaged in commerce
whether you like it or not.
> > > I know Mathematica is well-suited to doing other stuff I do, and it's
> > > hardly a
> > > problem that no one writes web apps in *that*.
>
> > The reason it's not a problem is that people use Mathematica to support
> > other profitable enterprises.
>
> I strongly suspect that the majority of the enterprises people support
> with Mathematica are of the non-profit variety.
Non-profit is not the same as non-commercial. What distinguishes a
non-profit organization from a for-profit one is not that one makes a
profit and the other doesn't. The distinction is that a for-profit
returns the profit to its shareholders and a non-profit doesn't. (They
have to roll their profits back into operations.) There are other
distinctions as well. For-profits pay taxes, non-profits don't.
Non-profits are more constrained in the kinds of business they may
engage in. But people can (and do) make a living from non-profit
organizations. Some even get rich (although when that happens it's
generally frowned upon).
rg
Le Sat, 02 Sep 2006 09:27:02 -0700, Ron Garret a écrit :
> If you get paid for your research work then you are engaged in
> commerce whether you like it or not.
Then you use the word commerce in an artificially broad way, which seems
not to me to be the accepted way.
Semantically,
Nowhere man
--
···········@levallois.eu.org
OpenPGP 0xD9D50D8A
In article <······························@levallois.eu.org>,
Pierre THIERRY <···········@levallois.eu.org> wrote:
> Le Sat, 02 Sep 2006 09:27:02 -0700, Ron Garret a écrit :
> > If you get paid for your research work then you are engaged in
> > commerce whether you like it or not.
>
> Then you use the word commerce in an artificially broad way, which seems
> not to me to be the accepted way.
Commerce means the buying and selling of goods and services. If you are
paid to do research then you are selling a service.
rg
From: Pascal Costanza
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <4lt97jF3i1btU2@individual.net>
Ron Garret wrote:
> The reason Joel focuses on Web apps (I
> think) is just because the Web seems to be where most of the lucrative
> opportunities are to be found nowadays.
I don't think so. It's probably because that's where the hype is.
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
Le Fri, 01 Sep 2006 20:32:59 -0700, Ron Garret a écrit :
>> I don't see how. People do lots of things other than commerce,
> Yes. Those are called "hobbies."
Or research, prototyping, in-house development or scripting (a.k.a.
automatization).
Alternatively,
Nowhere man
--
···········@levallois.eu.org
OpenPGP 0xD9D50D8A
On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
wrote:
> Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> failures (1) is greater than the number of high-profile Lisp successes
> (0). Having people using Lisp and consistently losing money is much,
> much worse than not having people using it at all.
Interesting. Care to quantify that.
(Links, examples..)
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
John Thingstad wrote:
> On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
> wrote:
>
> > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > failures (1) is greater than the number of high-profile Lisp successes
> > (0). Having people using Lisp and consistently losing money is much,
> > much worse than not having people using it at all.
>
> Interesting. Care to quantify that.
Word on the street is: Reddit tried with Lisp and "failed" (for some
definition of failed). Popular (ie. known but the large JAVA, .NET
masses) software written in Lisp == 0.
I believe that's what Ron meant.
Daniel
In article <························@i3g2000cwc.googlegroups.com>,
"Daniel Santa Cruz" <·······@gmail.com> wrote:
> John Thingstad wrote:
> > On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
> > wrote:
> >
> > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > > failures (1) is greater than the number of high-profile Lisp successes
> > > (0). Having people using Lisp and consistently losing money is much,
> > > much worse than not having people using it at all.
> >
> > Interesting. Care to quantify that.
What exactly do you want me to quantify? Do you not see the numbers
"8", "1" and "0"?
> Word on the street is: Reddit tried with Lisp and "failed" (for some
> definition of failed). Popular (ie. known but the large JAVA, .NET
> masses) software written in Lisp == 0.
>
> I believe that's what Ron meant.
>
> Daniel
Exactly.
rg
On Fri, 01 Sep 2006 22:29:48 +0200, Ron Garret <·········@flownet.com>
wrote:
> In article <························@i3g2000cwc.googlegroups.com>,
> "Daniel Santa Cruz" <·······@gmail.com> wrote:
>
>> John Thingstad wrote:
>> > On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
>> > wrote:
>> >
>> > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
>> > > failures (1) is greater than the number of high-profile Lisp
>> successes
>> > > (0). Having people using Lisp and consistently losing money is
>> much,
>> > > much worse than not having people using it at all.
>> >
>> > Interesting. Care to quantify that.
>
> What exactly do you want me to quantify? Do you not see the numbers
> "8", "1" and "0"?
How many companies have succeeded vs. how many have failed basically.
Also how does the choice of Lisp as a language contribute to the failure.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
In article <·················@pandora.upc.no>,
"John Thingstad" <··············@chello.no> wrote:
> On Fri, 01 Sep 2006 22:29:48 +0200, Ron Garret <·········@flownet.com>
> wrote:
>
> > In article <························@i3g2000cwc.googlegroups.com>,
> > "Daniel Santa Cruz" <·······@gmail.com> wrote:
> >
> >> John Thingstad wrote:
> >> > On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
> >> > wrote:
> >> >
> >> > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> >> > > failures (1) is greater than the number of high-profile Lisp
> >> successes
> >> > > (0). Having people using Lisp and consistently losing money is
> >> much,
> >> > > much worse than not having people using it at all.
> >> >
> >> > Interesting. Care to quantify that.
> >
> > What exactly do you want me to quantify? Do you not see the numbers
> > "8", "1" and "0"?
>
> How many companies have succeeded
Zero.
> vs. how many have failed
At least one high-profile failure, probably more that we don't know
about. People tend not to advertise their failures.
> Also how does the choice of Lisp as a language contribute to the failure.
In the case of Reddit, they implemented it in Lisp and had so many
problems that they re-implemented it in Python. Since then they seem to
be on a trajectory to success. That doesn't prove anything of course,
but do you see how even a not completely unreasonable person might reach
the conclusion that Python was a better choice than Lisp?
rg
Ron Garret wrote:
> In article <·················@pandora.upc.no>,
> "John Thingstad" <··············@chello.no> wrote:
>
> > On Fri, 01 Sep 2006 22:29:48 +0200, Ron Garret <·········@flownet.com>
> > wrote:
> >
> > > In article <························@i3g2000cwc.googlegroups.com>,
> > > "Daniel Santa Cruz" <·······@gmail.com> wrote:
> > >
> > >> John Thingstad wrote:
> > >> > On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
> > >> > wrote:
> > >> >
> > >> > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > >> > > failures (1) is greater than the number of high-profile Lisp
> > >> successes
> > >> > > (0). Having people using Lisp and consistently losing money is
> > >> much,
> > >> > > much worse than not having people using it at all.
> > >> >
> > >> > Interesting. Care to quantify that.
> > >
> > > What exactly do you want me to quantify? Do you not see the numbers
> > > "8", "1" and "0"?
> >
> > How many companies have succeeded
>
> Zero.
>
> > vs. how many have failed
>
> At least one high-profile failure, probably more that we don't know
> about. People tend not to advertise their failures.
>
> > Also how does the choice of Lisp as a language contribute to the failure.
>
> In the case of Reddit, they implemented it in Lisp and had so many
> problems that they re-implemented it in Python. Since then they seem to
> be on a trajectory to success. That doesn't prove anything of course,
> but do you see how even a not completely unreasonable person might reach
> the conclusion that Python was a better choice than Lisp?
>
> rg
I was surprised that they went with FreeBSD/sbcl at the time. There
were known problems with the combination.
Reddit never really struck me as an application that really needed a
language like Lisp. Maybe I haven't seen all of the features that
Reddit has to offer, but it seems like something that could be written
really quickly in PHP. I really like Reddit, BTW.
On 2006-09-01 21:34:52 -0500, ···········@gmail.com"
<··········@gmail.com> said:
> I was surprised that they went with FreeBSD/sbcl at the time. There
> were known problems with the combination.
And they spent time doing completely unnecessary things like writing
"low-level socket and threading code" [1] for such a laughably simple
app. Gotta wonder just how low is the Y-Combinator barrier to entry.
> Reddit never really struck me as an application that really needed a
> language like Lisp. Maybe I haven't seen all of the features that
> Reddit has to offer, but it seems like something that could be written
> really quickly in PHP. I really like Reddit, BTW.
Twenty minutes... in Lisp:
http://homepage.mac.com/svc/LispMovies/index.html#2
(As demonstrated by a competent programmer, I might add.)
-Mike
[1] http://redditblog.blogspot.com/2005/12/on-lisp.html
--
Michael J. Forster <····@sharedlogic.ca>
On Sat, 02 Sep 2006 00:44:12 +0200, Ron Garret <·········@flownet.com>
wrote:
>
> In the case of Reddit, they implemented it in Lisp and had so many
> problems that they re-implemented it in Python. Since then they seem to
> be on a trajectory to success. That doesn't prove anything of course,
> but do you see how even a not completely unreasonable person might reach
> the conclusion that Python was a better choice than Lisp?
>
> rg
In general it is a bad idea to write a large app without having at least
one
programmer with extensive experience in the language.
Say he has been using it for a couple of years and written 20 smaller
programs
and 30 000 lines of code. That's a pretty minimal figure and the numbers
are
somewhat arbitrary, but you get the picture.
If you have been writing applications in Java for 10 years and just learned
Lisp you are probably better off sticking with Java.
I haven't studied the 'redit incident' in detail but it was my
impression that the lack of success was due to inexperience and
that they went back to Python because it was more familiar.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
John Thingstad wrote:
> In general it is a bad idea to write a large app without having at least
> one
> programmer with extensive experience in the language.
> Say he has been using it for a couple of years and written 20 smaller
> programs
> and 30 000 lines of code. That's a pretty minimal figure and the numbers
> are
> somewhat arbitrary, but you get the picture.
> If you have been writing applications in Java for 10 years and just learned
> Lisp you are probably better off sticking with Java.
> I haven't studied the 'redit incident' in detail but it was my
> impression that the lack of success was due to inexperience and
> that they went back to Python because it was more familiar.
I rarely post here, so forgive any breaches of etiquette... ;)
I think this is an important point. I regularly read Joel's postings
and usually find them interesting, but it seems like a point is being
overlooked here by several of the replies. I've seen more than a few
articles/papers/postings/etc about web applications being developed in
"non-traditional" languages, and I've also read Paul Graham's writings
on ViaWeb.
I can't say there have been zero successes with Lisp (Naughty Dog,
Orbitz, the original ViaWeb are examples of successes), but the problem
with picking a language like Lisp, OCaml or Smalltalk is simple: you
might be the guru who implements the original system, but the company
might be in a heap of trouble if you leave and they try to find someone
as good as you are.
People pick Java or .NET not because they are better or worse than Lisp
or Smalltalk, but because they are familiar with it and can find
resources for the project. Unfortunately, it seems to be easier to
find Java or .NET developers for most projects. I've had similar
problems with my company... I'd love to develop some wonderful web
applications using something other than Java, but the response I
usually get is "that's great, but that means you are the only person
that can do anything with this code". I also had a very good coworker
who loved developing applications in Erlang, OCaml and Haskell, but he
had a hell of a time convincing people to let him use these languages,
and only succeeded in getting OCaml used on one small system. Of
course, he eventually left the company and everyone's hoping this
application never breaks. :)
Just my two cents...
Mark
> I also had a very good coworker
> who loved developing applications in Erlang, OCaml and Haskell, but he
> had a hell of a time convincing people to let him use these languages,
> and only succeeded in getting OCaml used on one small system. Of
> course, he eventually left the company and everyone's hoping this
> application never breaks. :)
>
> Just my two cents...
>
> Mark
A pet peeve of mine : programmers who aren't curious about other
languages.
Regards,
Jurgen
jurgen_defurne wrote:
> > I also had a very good coworker
> > who loved developing applications in Erlang, OCaml and Haskell, but he
> > had a hell of a time convincing people to let him use these languages,
> > and only succeeded in getting OCaml used on one small system. Of
> > course, he eventually left the company and everyone's hoping this
> > application never breaks. :)
> >
> > Just my two cents...
> >
> > Mark
>
> A pet peeve of mine : programmers who aren't curious about other
> languages.
>
> Regards,
>
> Jurgen
Agreed... I'm a language nut. I love learning/playing
with/experimenting with languages.
Mark
On 1 Sep 2006 18:29:49 -0700, "Mark" <··········@mac.com>:
>Unfortunately, it seems to be easier to
>find Java or .NET developers for most projects.
... so Java (or .Net) is more used than the others languages, so
people tend to learn Java, so there are more Java developers, etc.
It seems that if you want your pet language to be used, you need a
hell of a lot of money to advertise it.
Fabien LE LEZ wrote:
> On 1 Sep 2006 18:29:49 -0700, "Mark" <··········@mac.com>:
>
> >Unfortunately, it seems to be easier to
> >find Java or .NET developers for most projects.
>
> ... so Java (or .Net) is more used than the others languages, so
> people tend to learn Java, so there are more Java developers, etc.
>
> It seems that if you want your pet language to be used, you need a
> hell of a lot of money to advertise it.
...or you need to be one hell of an evangelist. :)
It does seem like some kind of vicious cycle... a language/platform
gets used more, so more people study it, it gets used in companies,
becomes popular, so more people study it.
I, personally, try to evangelize languages like Lisp (which I use/learn
in my spare time) even though I am a Java programmer by trade...
Mark
On 1 Sep 2006 21:07:27 -0700, "Mark" <··········@mac.com>:
>> It seems that if you want your pet language to be used, you need a
>> hell of a lot of money to advertise it.
>
>...or you need to be one hell of an evangelist. :)
I believe that most of those who decide which language to choose,
don't read Usenet or other places you might write on.
To reach them, you need glossy paper and conferences in sunny places.
Fabien LE LEZ <········@gramster.com> wrote:
> It seems that if you want your pet language to be used, you need a
> hell of a lot of money to advertise it.
One full page in the New York Times. That would be fun. :-)
Mark wrote:
<snipped>
> I'd love to develop some wonderful web
> applications using something other than Java, but the response I
> usually get is "that's great, but that means you are the only person
> that can do anything with this code". I also had a very good coworker
> who loved developing applications in Erlang, OCaml and Haskell, but he
> had a hell of a time convincing people to let him use these languages,
> and only succeeded in getting OCaml used on one small system. Of
> course, he eventually left the company and everyone's hoping this
> application never breaks. :)
>
I'd love to work in a company like that; give me the
erlang, lisp, ocaml, haskell or whatever minority
language project you have - I'm willing to
learn it on company time[1] even if the others aren't.
[1] I have a theory about the superiority of languages
like lisp, erlang, etc over languages like Java, C#, etc.
1. The latter languages are the ones demanded by our employers.
2. The former languages are forbidden by our employers (at
work, anyway).
3. A programmer who has decided to learn the former languages
will have invested a significant portion of his time
into learning it without expecting any financial gain.
4. A programmer who has /not/ learnt the former languages
doesn't really care about learning for the sake of
learning.
5. Therefore the people who know the former languages
are self-selected as the harder-workers (or smarter
through extra learning, etc) and will usually have
better problem solving skills than the rest and will
produce better solutions.
6. They (former languages people) implement their
better solutions in the former languages.
7. The non-learning people (latter languages) will
tend to produce only satisfactory or mediocre solutions.
This seems, to everyone else, that all the better solutions
are only possible with the former languages.
Alternatively, I'm just full of it today :-)
--
goose
Have I offended you? Send flames to ····@localhost
real email: lelanthran at gmail dot com
website : www.lelanthran.com
Mark <··········@mac.com> wrote:
> I can't say there have been zero successes with Lisp (Naughty Dog,
> Orbitz, the original ViaWeb are examples of successes), but the problem
> with picking a language like Lisp, OCaml or Smalltalk is simple: you
> might be the guru who implements the original system, but the company
> might be in a heap of trouble if you leave and they try to find someone
> as good as you are.
The company starts being in trouble when they let single persons
develop a project, instead of a team.
Stefan Scholl <······@no-spoon.de> writes:
> Mark <··········@mac.com> wrote:
>> I can't say there have been zero successes with Lisp (Naughty Dog,
>> Orbitz, the original ViaWeb are examples of successes), but the problem
>> with picking a language like Lisp, OCaml or Smalltalk is simple: you
>> might be the guru who implements the original system, but the company
>> might be in a heap of trouble if you leave and they try to find someone
>> as good as you are.
>
> The company starts being in trouble when they let single persons
> develop a project, instead of a team.
Yes. Sometimes they even take the prosperous risk to send their whole
developer team to some convention in the same plane!
Actually, companies start being in trouble when the let a single team
develop a project. They should really make everybody, from the
janitor to the CEO work on the project, and they should even involve
customers and passerbies, just in case a meteorite falls on the
company building.
Right now, I'm going to ask my baker to come over do some coding on my
project, least I get struck down.
--
__Pascal Bourguignon__ http://www.informatimago.com/
NEW GRAND UNIFIED THEORY DISCLAIMER: The manufacturer may
technically be entitled to claim that this product is
ten-dimensional. However, the consumer is reminded that this
confers no legal rights above and beyond those applicable to
three-dimensional objects, since the seven new dimensions are
"rolled up" into such a small "area" that they cannot be
detected.
Pascal Bourguignon <···@informatimago.com> wrote:
> Stefan Scholl <······@no-spoon.de> writes:
>> Mark <··········@mac.com> wrote:
>>> I can't say there have been zero successes with Lisp (Naughty Dog,
>>> Orbitz, the original ViaWeb are examples of successes), but the problem
>>> with picking a language like Lisp, OCaml or Smalltalk is simple: you
>>> might be the guru who implements the original system, but the company
>>> might be in a heap of trouble if you leave and they try to find someone
>>> as good as you are.
>>
>> The company starts being in trouble when they let single persons
>> develop a project, instead of a team.
>
> Yes. Sometimes they even take the prosperous risk to send their whole
> developer team to some convention in the same plane!
I was aiming for code quality, not redundancy of coders.
Stefan Scholl <······@no-spoon.de> writes:
> Pascal Bourguignon <···@informatimago.com> wrote:
>> Stefan Scholl <······@no-spoon.de> writes:
>>> Mark <··········@mac.com> wrote:
>>>> I can't say there have been zero successes with Lisp (Naughty Dog,
>>>> Orbitz, the original ViaWeb are examples of successes), but the problem
>>>> with picking a language like Lisp, OCaml or Smalltalk is simple: you
>>>> might be the guru who implements the original system, but the company
>>>> might be in a heap of trouble if you leave and they try to find someone
>>>> as good as you are.
>>>
>>> The company starts being in trouble when they let single persons
>>> develop a project, instead of a team.
>>
>> Yes. Sometimes they even take the prosperous risk to send their whole
>> developer team to some convention in the same plane!
>
> I was aiming for code quality, not redundancy of coders.
I was aiming for developer's quality.
Perhaps it would be better to have one good developer than a bad team.
Of course, if you can gather a team of good developers, better,
but not everybody's Google.
But who am I, what do I know? Managers seems to know that they're
better off with teams of average people than with less, but better
people.
I'm only wondering of the average Java/C/C++-monkey people they use
wouldn't be, most of them, better if they used Lisp. I bet yes, since
apparently, they're already better for using Java rather than C.
What we need is an opinion cascade to make PHBs want their teams to
use Lisp instead of Java. And it might be easier than we think, I've
met a number of technical managers who were told undercover to know
some Lisp, even in Cobol shops! Let's confess! How many of us DON'T
use Lisp in our job?
--
__Pascal Bourguignon__ http://www.informatimago.com/
ADVISORY: There is an extremely small but nonzero chance that,
through a process known as "tunneling," this product may
spontaneously disappear from its present location and reappear at
any random place in the universe, including your neighbor's
domicile. The manufacturer will not be responsible for any damages
or inconveniences that may result.
From: Pascal Costanza
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <4lt8jsF3i1btU1@individual.net>
Mark wrote:
> People pick Java or .NET not because they are better or worse than Lisp
> or Smalltalk, but because they are familiar with it and can find
> resources for the project. Unfortunately, it seems to be easier to
> find Java or .NET developers for most projects. I've had similar
> problems with my company... I'd love to develop some wonderful web
> applications using something other than Java, but the response I
> usually get is "that's great, but that means you are the only person
> that can do anything with this code". I also had a very good coworker
> who loved developing applications in Erlang, OCaml and Haskell, but he
> had a hell of a time convincing people to let him use these languages,
> and only succeeded in getting OCaml used on one small system. Of
> course, he eventually left the company and everyone's hoping this
> application never breaks. :)
>
> Just my two cents...
Yes, mainstream programming languages turn programmers into
plug-compatible units.
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
In article <·················@pandora.upc.no>,
"John Thingstad" <··············@chello.no> wrote:
> I haven't studied the 'redit incident' in detail but it was my
> impression that the lack of success was due to inexperience and
> that they went back to Python because it was more familiar.
Nope. The original blog entries seem to have vanished, but I found
salient excerpts on lemonodor:
"On my Mac, my choices of threaded Lisp implementations was limited to
OpenMCL, and in FreeBSD it's CMUCL. Because of the low-level socket and
threading code we had to write, reddit would not run on my Mac, and I
was always tethered to our FreeBSD development server. Not being able to
program offline is a pain.
If Lisp is so great, why did we stop using it? One of the biggest issues
was the lack of widely used and tested libraries. Sure, there is a CL
library for basically any task, but there is rarely more than one, and
often the libraries are not widely used or well documented."
rg
On Sat, 02 Sep 2006 05:40:25 +0200, Ron Garret <·········@flownet.com>
wrote:
> In article <·················@pandora.upc.no>,
> "John Thingstad" <··············@chello.no> wrote:
>
>> I haven't studied the 'redit incident' in detail but it was my
>> impression that the lack of success was due to inexperience and
>> that they went back to Python because it was more familiar.
>
> Nope. The original blog entries seem to have vanished, but I found
> salient excerpts on lemonodor:
>
> "On my Mac, my choices of threaded Lisp implementations was limited to
> OpenMCL, and in FreeBSD it's CMUCL. Because of the low-level socket and
> threading code we had to write, reddit would not run on my Mac, and I
> was always tethered to our FreeBSD development server. Not being able to
> program offline is a pain.
>
> If Lisp is so great, why did we stop using it? One of the biggest issues
> was the lack of widely used and tested libraries. Sure, there is a CL
> library for basically any task, but there is rarely more than one, and
> often the libraries are not widely used or well documented."
>
> rg
Erm.. But why didn't he know what libraries were available?
BSD/OS X and SBCL is asking for trouble if you want native
thread support. Inexperience is also appearance in the
fact that the requirements and the ability to meet them wasn't
known from the start. Seems to me most of the problems stem from
being stuck on one set of hardware and the choosing solutions that didn't
have
enough support on that hardware.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
On 2006-09-02 01:48:15 -0500, "John Thingstad" <··············@chello.no> said:
> On Sat, 02 Sep 2006 05:40:25 +0200, Ron Garret <·········@flownet.com> wrote:
>
>> In article <·················@pandora.upc.no>,
>> "John Thingstad" <··············@chello.no> wrote:
>>
>>> I haven't studied the 'redit incident' in detail but it was my
>>> impression that the lack of success was due to inexperience and
>>> that they went back to Python because it was more familiar.
>>
>> Nope. The original blog entries seem to have vanished, but I found
>> salient excerpts on lemonodor:
>>
>> "On my Mac, my choices of threaded Lisp implementations was limited to
>> OpenMCL, and in FreeBSD it's CMUCL. Because of the low-level socket and
>> threading code we had to write, reddit would not run on my Mac, and I
>> was always tethered to our FreeBSD development server. Not being able to
>> program offline is a pain.
>>
>> If Lisp is so great, why did we stop using it? One of the biggest issues
>> was the lack of widely used and tested libraries. Sure, there is a CL
>> library for basically any task, but there is rarely more than one, and
>> often the libraries are not widely used or well documented."
>>
>> rg
>
> Erm.. But why didn't he know what libraries were available?
And shouldn't the Y-Combinator mentors have noted this? Perhaps they
did, but, if so, why didn't they demand better of the kids before letting them
continue? Really, if this is all you gotta muster to get PG to hand over the
$$$, well, I have this great idea for a Lisp Machine ;-)
> BSD/OS X and SBCL is asking for trouble if you want native
> thread support. Inexperience is also appearance in the
> fact that the requirements and the ability to meet them wasn't
> known from the start. Seems to me most of the problems stem from
> being stuck on one set of hardware and the choosing solutions that didn't have
> enough support on that hardware.
On the other hand, the combination of Portable AllegroServe and either
CMUCL on FreeBSD or LispWorks on OS X has proven to be a fast and
stable base for our work. And I should note that my partner writes in
LispWorks on Windows, just to underscore the absence of multi-platform
headaches. Neither do we find ourselves having to write "low-level socket
and threading code..." for apps that are undoubtedly more complex than
reddit.
-Mike
--
Michael J. Forster <····@sharedlogic.ca>
"John Thingstad" <··············@chello.no> writes:
>
> Inexperience is also appearance in the fact that the requirements and
> the ability to meet them wasn't known from the start. Seems to me most
> of the problems stem from being stuck on one set of hardware and the
> choosing solutions that didn't have enough support on that hardware.
And then they went with a solution which _is_ supported on the hardware:
Python. Which is unfortunate.
--
Robert Uhl <http://public.xdi.org/=ruhl>
progress (n): the process through which Usenet has evolved from smart
people in front of dumb terminals to dumb people in front of smart
terminals --obs at burnout.demon.co.uk
"Daniel Santa Cruz" <·······@gmail.com> writes:
>
> Word on the street is: Reddit tried with Lisp and "failed" (for some
> definition of failed).
It boiled down to two things: no common Common Lisp (heh) between
FreeBSD and OS X which supported what they needed (sockets being one of
the requirements, as I recall); and stability problems with threads.
Interestingly, the former is exactly what folks have been complaining
about for awhile now--and it's exactly what a lot of other folks claim
is no big deal.
--
Robert Uhl <http://public.xdi.org/=ruhl>
Architects cannot learn to design grand cathedrals if they are taught
all their drawing courses using only an Etch-a-Sketch because the
company struck a deal with the university... --Gene Spafford
John Thingstad wrote:
> On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
> wrote:
>
> > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > failures (1) is greater than the number of high-profile Lisp successes
> > (0). Having people using Lisp and consistently losing money is much,
> > much worse than not having people using it at all.
>
> Interesting. Care to quantify that.
> (Links, examples..)
I consider Naughty Dog to be a failure. This example says that yes,
companies will try out Lisp, even invest significant resources and ship
products with it. But eventually, they'll think better of it and
abandon Lisp. Paul Graham's example also shows the world that Lisp has
no staying power. It may get a guru programmer to a certain point,
even a very profitable point, but the commercial examples do not
demonstrate repeatability.
What the world needs is a killer app that has to get done in Lisp
because one simply isn't gonna achieve it with a pile of code monkeys.
The competitive advantage has to be sustainable, not just a one hit
wonder.
Cheers,
Brandon Van Every
Mallor wrote:
> John Thingstad wrote:
> > On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
> > wrote:
> >
> > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > > failures (1) is greater than the number of high-profile Lisp successes
> > > (0). Having people using Lisp and consistently losing money is much,
> > > much worse than not having people using it at all.
> >
> > Interesting. Care to quantify that.
> > (Links, examples..)
>
> I consider Naughty Dog to be a failure.
Please? What are you smoking?
The guys made millions. They created fantastic games which were rated
extremely high by critics and gamers.They sold millions of copies of
these games. Plus they sold the company for even more money to Sony.
You can't have much better success than that. The first games were
developed with Lisp systems and the later games even were based on a
Lisp system running on the Playstation. Which means they sold millions
of these Lisp-based games.
Btw., if you actually could program some Lisp, you might even get a job
at the new company of them. But all you seem to do is to post some poor
comments on Usenet. If you want to spend your time with some Lisp
system, get Jak & Dexter. It has been created by a Lisp system (Allegro
CL) and runs on a Lisp system (GOAL runtime).
> This example says that yes,
> companies will try out Lisp, even invest significant resources and ship
> products with it. But eventually, they'll think better of it and
> abandon Lisp.
Some not. Some are running Lisp systems for more than twenty years.
> Paul Graham's example also shows the world that Lisp has
> no staying power.
Bla, it enabled some guys to enter a market very quickly.
Moving from prototypes to early production systems is one of the main
strengths of Lisp.
This has been demonstrated thousands of times.
> It may get a guru programmer to a certain point,
> even a very profitable point, but the commercial examples do not
> demonstrate repeatability.
What commercial examples do you know? Three, or four? In which Lisp
products were you involved? Or is it all just hearsay?
> What the world needs is a killer app that has to get done in Lisp
Bla. What the world may need is better software that doesn't get
cracked by schoolkids.
> because one simply isn't gonna achieve it with a pile of code monkeys.
> The competitive advantage has to be sustainable, not just a one hit
> wonder.
>
>
> Cheers,
> Brandon Van Every
In article <·······················@p79g2000cwp.googlegroups.com>,
·······@corporate-world.lisp.de" <······@corporate-world.lisp.de>
wrote:
> Mallor wrote:
> > John Thingstad wrote:
> > > On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
> > > wrote:
> > >
> > > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > > > failures (1) is greater than the number of high-profile Lisp successes
> > > > (0). Having people using Lisp and consistently losing money is much,
> > > > much worse than not having people using it at all.
> > >
> > > Interesting. Care to quantify that.
> > > (Links, examples..)
> >
> > I consider Naughty Dog to be a failure.
>
> Please? What are you smoking?
>
> The guys made millions.
I don't wish to argue that Naughty Dog was a failure, but under no
circumstances can you consider it a *high profile* success. It was
acquired by Sony in the depths of the dotcom crash for an undisclosed
sum. That sure makes it seem like it was a fire sale. Furthermore,
once the acquisition was complete Lisp was promptly abandoned.
But as always I'm happy to be proven wrong on this. Do you have any
data to show how much money Naughty Dog made for its investors?
rg
Ron Garret wrote:
> In article <·······················@p79g2000cwp.googlegroups.com>,
> ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de>
> wrote:
>
> > Mallor wrote:
> > > John Thingstad wrote:
> > > > On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
> > > > wrote:
> > > >
> > > > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > > > > failures (1) is greater than the number of high-profile Lisp successes
> > > > > (0). Having people using Lisp and consistently losing money is much,
> > > > > much worse than not having people using it at all.
> > > >
> > > > Interesting. Care to quantify that.
> > > > (Links, examples..)
> > >
> > > I consider Naughty Dog to be a failure.
> >
> > Please? What are you smoking?
> >
> > The guys made millions.
>
> I don't wish to argue that Naughty Dog was a failure, but under no
> circumstances can you consider it a *high profile* success.
It is a HUGE high profile success. You don't have an idea what you
are talking about. Sony made Crash Bandicoot their (unofficial)
platform masco for the playstation. The games were just amazing.
Just read the critics. Should I look for some reviews for you?
Otherwise Google is your friend.
IGN rates Jak II and Jak III as 'incredible'. Jak 3 is rated to have
the
3rd best graphics of all games published for the Playstation 2.
Animation and sound were also technical perfect.
> It was
> acquired by Sony in the depths of the dotcom crash for an undisclosed
> sum.
Ron, we are talking about computer games companies. Not dot.coms.
Naughty Dog was not a dot.com. It was/is a company creating games for
the Sony Playstation 1 and then Playstation 2.
Sony bought them because the games were so successful, and not
because they were cheap or something else. They pushed Crash Bandicoot
very big.
> That sure makes it seem like it was a fire sale.
Do you have ANY evidence to back that up? Not just a GUESS?
> Furthermore,
> once the acquisition was complete Lisp was promptly abandoned.
Promptly? It was? Who said that? Where did you read that? Naughty Dog
released for example Jak2 and Jak3 as Lisp-based games for Sony for the
PlayStation 2.
A few years after they have been bought by Sony they decided not to use
Lisp for the
next-gen console games (not to write a new Lisp-based engine for the
Playstation 3 that means essentially). The founders also left Sony.
>
> But as always I'm happy to be proven wrong on this. Do you have any
> data to show how much money Naughty Dog made for its investors?
Dunno. I guess the founders can easily finance the gasoline for their
Ferraris. But lets look at the sales of the Crash Series:
http://www.naughtydog.com/crash/crash/timeline.htm
Release somewhen in 1996.
1998, 'Crash 1 & 2 become the number one and number two best selling
PlayStation titles of all-time in North America.'
1998, 'As Crash 3 tops one million units in the U.S., WORLDWIDE SALES
OF THE TRILOGY SURPASS 10 MILLION!'
So in not even two years they sold ten million games. In 2002 sales
reached more than
22 million copies of Crash Bandicoot 1/2/3.
Just a guess, but I'd think games sold on Naughty Dogs (Crash
bandicoot, Jak & Dexter, ...) concepts should by now have reached the
50 million or more mark. Multiply that with the cost of a typical game,
add the merchandize stuff and you might get an idea how huge the
success is. Like sales of easily more than (really a guess) two hundred
million US$. Would that be a 'success' for you? Huge even?
IMHO, their (Lisp-based)games were total fun and beauty. Well, I had
fun playing Crash Bandicoot 1 when it came out. ;-)
But then I had also fun at that time playing Super Mario 64 from
Nintendo. Which was developed with the help of another Lisp-based
system: N-World. ;-)
Ron, when will you get over your Lisp depression?
······@corporate-world.lisp.de wrote:
> Just a guess, but I'd think games sold on Naughty Dogs (Crash
> bandicoot, Jak & Dexter, ...) concepts should by now have reached the
> 50 million or more mark. Multiply that with the cost of a typical game,
> add the merchandize stuff and you might get an idea how huge the
> success is. Like sales of easily more than (really a guess) two hundred
> million US$. Would that be a 'success' for you? Huge even?
It's late. Let's calculate.
Let's assume 50 million games were sold for each $40.
Fifty million is not unreasonable, since Crash 1/2/3 sold alone of 22
million until 2002. Add the other games, add that Crash has been
released on some other platforms, too. Just guessing.
* (* 50000000 40)
2,000,000,000
Which would be around two billion $ in sales. Just think about it. Is
that huge? Just in sales?
Huge enough for Ron?
In article <························@b28g2000cwb.googlegroups.com>,
·······@corporate-world.lisp.de" <······@corporate-world.lisp.de>
wrote:
> ······@corporate-world.lisp.de wrote:
> > Just a guess, but I'd think games sold on Naughty Dogs (Crash
> > bandicoot, Jak & Dexter, ...) concepts should by now have reached the
> > 50 million or more mark. Multiply that with the cost of a typical game,
> > add the merchandize stuff and you might get an idea how huge the
> > success is. Like sales of easily more than (really a guess) two hundred
> > million US$. Would that be a 'success' for you? Huge even?
>
> It's late. Let's calculate.
>
> Let's assume 50 million games were sold for each $40.
> Fifty million is not unreasonable, since Crash 1/2/3 sold alone of 22
> million until 2002. Add the other games, add that Crash has been
> released on some other platforms, too. Just guessing.
>
> * (* 50000000 40)
>
> 2,000,000,000
>
> Which would be around two billion $ in sales. Just think about it. Is
> that huge? Just in sales?
> Huge enough for Ron?
Not bad.
How much of that was profit?
rg
If somebody wants to know more facts about GOAL, and it sucess, and why
when they were bought by Sony they switched to C++ - get all the
messages from Scott Schumacher...
http://lists.midnightryder.com/pipermail/sweng-gamedev-midnightryder.com/2005-August/003798.html
and the thread itself
http://lists.midnightryder.com/pipermail/sweng-gamedev-midnightryder.com/2005-August/thread.html#3789
······@corporate-world.lisp.de wrote:
> Ron Garret wrote:
> > In article <·······················@p79g2000cwp.googlegroups.com>,
> > ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de>
> > wrote:
> >
> > > Mallor wrote:
> > > > John Thingstad wrote:
> > > > > On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret <·········@flownet.com>
> > > > > wrote:
> > > > >
> > > > > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > > > > > failures (1) is greater than the number of high-profile Lisp successes
> > > > > > (0). Having people using Lisp and consistently losing money is much,
> > > > > > much worse than not having people using it at all.
> > > > >
> > > > > Interesting. Care to quantify that.
> > > > > (Links, examples..)
> > > >
> > > > I consider Naughty Dog to be a failure.
> > >
> > > Please? What are you smoking?
> > >
> > > The guys made millions.
> >
> > I don't wish to argue that Naughty Dog was a failure, but under no
> > circumstances can you consider it a *high profile* success.
>
> It is a HUGE high profile success. You don't have an idea what you
> are talking about. Sony made Crash Bandicoot their (unofficial)
> platform masco for the playstation. The games were just amazing.
> Just read the critics. Should I look for some reviews for you?
> Otherwise Google is your friend.
>
> IGN rates Jak II and Jak III as 'incredible'. Jak 3 is rated to have
> the
> 3rd best graphics of all games published for the Playstation 2.
> Animation and sound were also technical perfect.
>
> > It was
> > acquired by Sony in the depths of the dotcom crash for an undisclosed
> > sum.
>
> Ron, we are talking about computer games companies. Not dot.coms.
> Naughty Dog was not a dot.com. It was/is a company creating games for
> the Sony Playstation 1 and then Playstation 2.
>
> Sony bought them because the games were so successful, and not
> because they were cheap or something else. They pushed Crash Bandicoot
> very big.
>
> > That sure makes it seem like it was a fire sale.
>
> Do you have ANY evidence to back that up? Not just a GUESS?
>
> > Furthermore,
> > once the acquisition was complete Lisp was promptly abandoned.
>
> Promptly? It was? Who said that? Where did you read that? Naughty Dog
> released for example Jak2 and Jak3 as Lisp-based games for Sony for the
> PlayStation 2.
> A few years after they have been bought by Sony they decided not to use
> Lisp for the
> next-gen console games (not to write a new Lisp-based engine for the
> Playstation 3 that means essentially). The founders also left Sony.
>
> >
> > But as always I'm happy to be proven wrong on this. Do you have any
> > data to show how much money Naughty Dog made for its investors?
>
> Dunno. I guess the founders can easily finance the gasoline for their
> Ferraris. But lets look at the sales of the Crash Series:
>
> http://www.naughtydog.com/crash/crash/timeline.htm
>
> Release somewhen in 1996.
>
> 1998, 'Crash 1 & 2 become the number one and number two best selling
> PlayStation titles of all-time in North America.'
>
> 1998, 'As Crash 3 tops one million units in the U.S., WORLDWIDE SALES
> OF THE TRILOGY SURPASS 10 MILLION!'
>
> So in not even two years they sold ten million games. In 2002 sales
> reached more than
> 22 million copies of Crash Bandicoot 1/2/3.
>
> Just a guess, but I'd think games sold on Naughty Dogs (Crash
> bandicoot, Jak & Dexter, ...) concepts should by now have reached the
> 50 million or more mark. Multiply that with the cost of a typical game,
> add the merchandize stuff and you might get an idea how huge the
> success is. Like sales of easily more than (really a guess) two hundred
> million US$. Would that be a 'success' for you? Huge even?
>
> IMHO, their (Lisp-based)games were total fun and beauty. Well, I had
> fun playing Crash Bandicoot 1 when it came out. ;-)
>
> But then I had also fun at that time playing Super Mario 64 from
> Nintendo. Which was developed with the help of another Lisp-based
> system: N-World. ;-)
>
> Ron, when will you get over your Lisp depression?
In article <·······················@74g2000cwt.googlegroups.com>,
·······@corporate-world.lisp.de" <······@corporate-world.lisp.de>
wrote:
> Ron Garret wrote:
> > In article <·······················@p79g2000cwp.googlegroups.com>,
> > ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de>
> > wrote:
> >
> > > Mallor wrote:
> > > > John Thingstad wrote:
> > > > > On Fri, 01 Sep 2006 21:15:21 +0200, Ron Garret
> > > > > <·········@flownet.com>
> > > > > wrote:
> > > > >
> > > > > > Since Paul sold ViaWeb 8 years ago, the number of high-profile Lisp
> > > > > > failures (1) is greater than the number of high-profile Lisp
> > > > > > successes
> > > > > > (0). Having people using Lisp and consistently losing money is
> > > > > > much,
> > > > > > much worse than not having people using it at all.
> > > > >
> > > > > Interesting. Care to quantify that.
> > > > > (Links, examples..)
> > > >
> > > > I consider Naughty Dog to be a failure.
> > >
> > > Please? What are you smoking?
> > >
> > > The guys made millions.
> >
> > I don't wish to argue that Naughty Dog was a failure, but under no
> > circumstances can you consider it a *high profile* success.
>
> It is a HUGE high profile success. You don't have an idea what you
> are talking about. Sony made Crash Bandicoot their (unofficial)
> platform masco for the playstation. The games were just amazing.
> Just read the critics. Should I look for some reviews for you?
I do not dispute that the games were good and popular -- but did they
turn a profit?
> Sony bought them because the games were so successful, and not
> because they were cheap or something else.
How do you know? Were you involved in the acquisition?
> They pushed Crash Bandicoot very big.
That proves nothing. Just because Sony picked up a valuable asset
doesn't necessarily mean it wasn't a fire sale.
> > That sure makes it seem like it was a fire sale.
>
> Do you have ANY evidence to back that up? Not just a GUESS?
I don't need evidence to back up this particular point. A lack of
evidence suffices. If there is no evidence that something was a high
profile success then ipso facto it was not. It may well have been a
success, but it was not high profile.
(Note that by "high profile success" I mean "high profile financial
success that is attributable at least in part to the use of Lisp." I do
not dispute that ND's games were wildly popular.)
> > Furthermore,
> > once the acquisition was complete Lisp was promptly abandoned.
>
> Promptly? It was? Who said that? Where did you read that?
http://bc.tech.coop/blog/060118.html
But I probably should not have used the word "promptly." It did take a
couple of years.
Of course, it took a couple of years for Yahoo to abandon Lisp too
because it took them that long to find someone who could rewrite Viaweb
it in C++.
> > But as always I'm happy to be proven wrong on this. Do you have any
> > data to show how much money Naughty Dog made for its investors?
>
> Dunno. I guess the founders can easily finance the gasoline for their
> Ferraris. But lets look at the sales of the Crash Series:
>
> http://www.naughtydog.com/crash/crash/timeline.htm
>
> Release somewhen in 1996.
>
> 1998, 'Crash 1 & 2 become the number one and number two best selling
> PlayStation titles of all-time in North America.'
>
> 1998, 'As Crash 3 tops one million units in the U.S., WORLDWIDE SALES
> OF THE TRILOGY SURPASS 10 MILLION!'
>
> So in not even two years they sold ten million games. In 2002 sales
> reached more than
> 22 million copies of Crash Bandicoot 1/2/3.
>
> Just a guess, but I'd think games sold on Naughty Dogs (Crash
> bandicoot, Jak & Dexter, ...) concepts should by now have reached the
> 50 million or more mark. Multiply that with the cost of a typical game,
> add the merchandize stuff and you might get an idea how huge the
> success is. Like sales of easily more than (really a guess) two hundred
> million US$. Would that be a 'success' for you? Huge even?
OK, I'm convinced. Naughty Dog was (probably -- I still don't really
know for sure) a financial success.
That's one.
> Ron, when will you get over your Lisp depression?
I dunno. When Digitool is profitable again maybe?
rg
Ron Garret wrote:
...
> > Sony bought them because the games were so successful, and not
> > because they were cheap or something else.
>
> How do you know? Were you involved in the acquisition?
At that time Nintendo was THE game company and with Super Mario
they had a huge success and a mascot. The Nintendo 64 was
their new console and it had the 3d adaption called Super Matio 64
which was a HUGE hit. It was revolutionary.
Sony was entering the market with the Playstation as a direct
competition.
But they lacked a character like Mario and they lacked a game like
Super Mario 64 to show off the 3d capabilities in a so-called 'platform
game'.
I guess the best they saw was Crash Bandicoot. It was a character
and a very competetive game (though not revolutionary like Super Mario
64).
Sony was desperately looking for something to put against the Nintendo
64's
platform game hit Super Mario 64.
It was (and still is) a highly competitive business and Naughty Dog
kindof had what Sony needed. I really don't think this would be a bad
situation for negotating a contract. ;-) Just the opposite.
...
> > Ron, when will you get over your Lisp depression?
>
> I dunno. When Digitool is profitable again maybe?
Okay, that will be when they offer a competitive product and
get an idea who to run a company to sell that. Not sure,
but that definitely would help. Since both is unlikely,
you really should look for another way to get over your
Lisp depression, IMHO.
> I do not dispute that the games were good and popular -- but did they
> turn a profit?
Hi, sorry to jump in here, but as a serious gamer, Nuaghty Dog were
well reknowned for their games amongst both gamers and critics alike.
Other companies have had huge critical acclaim and not turned a profit
(Looking Glass Software as a poster child), but Naughty Dog were never
in that boat - I've read about them for years in the UK magazine EDGE
(Jak & Dexter is awesome, but I'm not sure if the final one (#3) was a
GOAL game or if it used C++
> How do you know? Were you involved in the acquisition?
>
> > They pushed Crash Bandicoot very big.
>
> That proves nothing. Just because Sony picked up a valuable asset
> doesn't necessarily mean it wasn't a fire sale.
At the time that Crash sort of became the Sony mascot, there's no way
that Naughty Dog sold out cheap on that. Sega had Sonic, Nintendo had
(Mario|Luigi|Zelda|Turok) Sony had nothing and were desperate, they
needed a mascot to 'compete', and Naughty Dog was the company in the
position of power to dictate terms to Sony. - This was back in the
playstation days, when Sony were the new boys in town. By time the PS2
was available, Sony had become the 800lb gorilla and yes they would
have been able to dictate terms to Naughty Dog.
> I don't need evidence to back up this particular point. A lack of
> evidence suffices. If there is no evidence that something was a high
> profile success then ipso facto it was not. It may well have been a
> success, but it was not high profile.
>
> (Note that by "high profile success" I mean "high profile financial
> success that is attributable at least in part to the use of Lisp." I do
> not dispute that ND's games were wildly popular.)
Within the gaming community the company had a lot of respect (similar
to Rare before Nintendo bought them, or Bungie before Microsoft got
them) - not sure what they are doing now - but I had heard of GOAL
mentioned in the gaming press (not the gaming 'tabloids', but serious
in interviews it was mentioned as Naughty Dog's 'special' system that
allowed them to build these games - however there wasn't an explicit
'we made a ton of cash because we used Lisp'
>
> > > Furthermore,
> > > once the acquisition was complete Lisp was promptly abandoned.
> >
> > Promptly? It was? Who said that? Where did you read that?
>
Yeah as I mentioned, I think the latest Jak & Dexter game #3, is a C++
game (no way to tell really, but it did come out after the announcement
that Sony were moving all internal teams over to a common dev platform,
so that suggests that by now all games made by Sony studios will not
use Lisp|Smalltalk|Ocaml.
From: Rainer Joswig
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <C1231AE6.4F6E4%joswig@lisp.de>
Am 05.09.2006 10:38 Uhr schrieb ·········@gmail.com" unter
<········@gmail.com> in
························@h48g2000cwc.googlegroups.com:
>>>> Furthermore,
>>>> once the acquisition was complete Lisp was promptly abandoned.
>>>
>>> Promptly? It was? Who said that? Where did you read that?
>>
>
> Yeah as I mentioned, I think the latest Jak & Dexter game #3, is a C++
> game
Here ND shows a snippet of GOAL code from Jak3:
http://lists.midnightryder.com/pipermail/sweng-gamedev-midnightryder.com/200
5-August/003804.html
> (no way to tell really, but it did come out after the announcement
> that Sony were moving all internal teams over to a common dev platform,
> so that suggests that by now all games made by Sony studios will not
> use Lisp|Smalltalk|Ocaml.
>
On 1 Sep 2006 17:44:28 -0700, ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> wrote:
>
> Ron Garret wrote:
> >
> > I don't wish to argue that Naughty Dog was a failure, but under no
> > circumstances can you consider it a *high profile* success.
>
> It is a HUGE high profile success. You don't have an idea what you
> are talking about. Sony made Crash Bandicoot their (unofficial)
> platform masco for the playstation. The games were just amazing.
> Just read the critics. Should I look for some reviews for you?
> Otherwise Google is your friend.
Let me see if I can help with a translation. "High profile" in
in this context means people who put valuations on potential
investments will have been able, for a few months at least, to
read articles in Fortune or Inc or Technology Review about companies
that use Common Lisp, where the aim of said articles is to position
those companies as "the next big thing" and the founders' use of
Common Lisp as evidence of what great "out of the box" thinking
those brilliant fellows are capable of.
Then certain well-known and prolific "industry thought leaders"
would have to start banging the drum about Common Lisp, writing
articles and blog entries and giving seminars, leading to a new
generation of consultants springing up around Common Lisp.
And so finally big name consultancies like Accenture would
eventually feel compelled to start recommending Common Lisp to their
clients. And big name analyst groups like Gartner would start
writing reports about how companies who did not use Common Lisp were
at a competitive disadvantage, facing an uphill climb, etc., etc.
That would be "high profile." Whether it's "good" is another
question.
--
Jack Unrue
Jack Unrue wrote:
> On 1 Sep 2006 17:44:28 -0700, ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> wrote:
> >
> > Ron Garret wrote:
> > >
> > > I don't wish to argue that Naughty Dog was a failure, but under no
> > > circumstances can you consider it a *high profile* success.
> >
> > It is a HUGE high profile success. You don't have an idea what you
> > are talking about. Sony made Crash Bandicoot their (unofficial)
> > platform masco for the playstation. The games were just amazing.
> > Just read the critics. Should I look for some reviews for you?
> > Otherwise Google is your friend.
>
> Let me see if I can help with a translation.
You can bend the meaning in any direction you want.
Ron's current topic is that using Lisp makes something automatically a
failure.
Reread the thread. Naughty Dog used Lisp, so Naughty Dog can't be a
successful
company. I gave a few indications that Naughty Dog might have been a
very successful
company. But it seems to be for Ron and you (?) unsual that this could
be the case.
Now you add the restriction that 'high profile' means that the product
generated
media attention for the tools that had been used. Sorry, I can't follow
you.
Is it really that hard to accept that Naughty Dog created with Lisp
(and other tools)
some games that literally sold many millions and were huge success with
critics
and gamers? It was atleast that interesting that some game magazines
wrote background stories on Naughty Dog.
Ron just lives through his (Common) Lisp depression openly on
comp.lang.lisp.
I'm not sure if that is the right therapy or if that create any useful
lines of Lisp code.
Probably not.
> "High profile" in
> in this context means people who put valuations on potential
> investments will have been able, for a few months at least, to
> read articles in Fortune or Inc or Technology Review about companies
> that use Common Lisp, where the aim of said articles is to position
> those companies as "the next big thing" and the founders' use of
> Common Lisp as evidence of what great "out of the box" thinking
> those brilliant fellows are capable of.
>
> Then certain well-known and prolific "industry thought leaders"
> would have to start banging the drum about Common Lisp, writing
> articles and blog entries and giving seminars, leading to a new
> generation of consultants springing up around Common Lisp.
>
> And so finally big name consultancies like Accenture would
> eventually feel compelled to start recommending Common Lisp to their
> clients. And big name analyst groups like Gartner would start
> writing reports about how companies who did not use Common Lisp were
> at a competitive disadvantage, facing an uphill climb, etc., etc.
>
> That would be "high profile." Whether it's "good" is another
> question.
>
> --
> Jack Unrue
In article <·······················@74g2000cwt.googlegroups.com>,
·······@corporate-world.lisp.de" <······@corporate-world.lisp.de>
wrote:
> Jack Unrue wrote:
> > On 1 Sep 2006 17:44:28 -0700, ·······@corporate-world.lisp.de"
> > <······@corporate-world.lisp.de> wrote:
> > >
> > > Ron Garret wrote:
> > > >
> > > > I don't wish to argue that Naughty Dog was a failure, but under no
> > > > circumstances can you consider it a *high profile* success.
> > >
> > > It is a HUGE high profile success. You don't have an idea what you
> > > are talking about. Sony made Crash Bandicoot their (unofficial)
> > > platform masco for the playstation. The games were just amazing.
> > > Just read the critics. Should I look for some reviews for you?
> > > Otherwise Google is your friend.
> >
> > Let me see if I can help with a translation.
>
> You can bend the meaning in any direction you want.
>
> Ron's current topic is that using Lisp makes something automatically a
> failure.
No, that is not what I said.
> Reread the thread. Naughty Dog used Lisp, so Naughty Dog can't be a
> successful company.
No. What I said was that ND is not a "high profile success." What I
meant by that is that it is hard to tell how much money ND actually
made. (Contrast that situation with Viaweb where the attention-getting
number of $49M is abundantly published.)
I have actually since changed my position and accepted that ND was a
financial success.
> Now you add the restriction that 'high profile' means that the product
> generated
> media attention for the tools that had been used. Sorry, I can't follow
> you.
No. I don't want media attention. I just want something on the record
that shows that their investors got a good return on their investment,
and that they think that this was *because* they used Lisp rather than
despite it. Again I cite Viaweb as my poster child for the high-profile
success, where there is no doubt whatsoever where the principles stand.
> Is it really that hard to accept that Naughty Dog created with Lisp
> (and other tools)
> some games that literally sold many millions and were huge success with
> critics
> and gamers? It was atleast that interesting that some game magazines
> wrote background stories on Naughty Dog.
I have never disputed that. Webvan got a lot of attention in its day
too. But investors focus more on the endgame.
> Ron just lives through his (Common) Lisp depression openly on
> comp.lang.lisp.
> I'm not sure if that is the right therapy or if that create any useful
> lines of Lisp code.
> Probably not.
Or maybe I'm here because I'm the consummate optimist who still believes
that it is possible to make money using Lisp despite overwhelming
evidence to the contrary.
rg
On 1 Sep 2006 18:45:13 -0700, ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> wrote:
>
> Jack Unrue wrote:
> > On 1 Sep 2006 17:44:28 -0700, ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> wrote:
> > >
> > > Ron Garret wrote:
> > > >
> > > > I don't wish to argue that Naughty Dog was a failure, but under no
> > > > circumstances can you consider it a *high profile* success.
> > >
> > > It is a HUGE high profile success. You don't have an idea what you
> > > are talking about. Sony made Crash Bandicoot their (unofficial)
> > > platform masco for the playstation. The games were just amazing.
> > > Just read the critics. Should I look for some reviews for you?
> > > Otherwise Google is your friend.
> >
> > Let me see if I can help with a translation.
>
> You can bend the meaning in any direction you want.
>
> Ron's current topic is that using Lisp makes something automatically
> a failure. Reread the thread. Naughty Dog used Lisp, so Naughty Dog
> can't be a successful company. I gave a few indications that Naughty Dog
> might have been a very successful company. But it seems to be for Ron
> and you (?) unsual that this could be the case.
> Now you add the restriction that 'high profile' means that the product
> generated media attention for the tools that had been used. Sorry, I
> can't follow you.
I failed in my translation attempt, obviously, because you're completely
missing the gist, which I had hoped would become clear when I questioned
whether that whole process was good.
> Is it really that hard to accept that Naughty Dog created with Lisp
> (and other tools) some games that literally sold many millions and
> were huge success with critics and gamers? It was atleast that
> interesting that some game magazines wrote background stories on
> Naughty Dog.
It's not hard for me to accept, quite the opposite. I'm with
you, I would overjoyed to be as successful Naughty Dog.
--
Jack Unrue
I wrote:
> On 1 Sep 2006 18:45:13 -0700, ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> wrote:
> >
> > Jack Unrue wrote:
> > > On 1 Sep 2006 17:44:28 -0700, ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> wrote:
> > > >
> > > > Ron Garret wrote:
> > > > >
> > > > > I don't wish to argue that Naughty Dog was a failure, but under no
> > > > > circumstances can you consider it a *high profile* success.
> > > >
> > > > It is a HUGE high profile success. You don't have an idea what you
> > > > are talking about. Sony made Crash Bandicoot their (unofficial)
> > > > platform masco for the playstation. The games were just amazing.
> > > > Just read the critics. Should I look for some reviews for you?
> > > > Otherwise Google is your friend.
> > >
> > > Let me see if I can help with a translation.
> >
> > You can bend the meaning in any direction you want.
> >
> > Ron's current topic is that using Lisp makes something automatically
> > a failure. Reread the thread. Naughty Dog used Lisp, so Naughty Dog
> > can't be a successful company. I gave a few indications that Naughty Dog
> > might have been a very successful company. But it seems to be for Ron
> > and you (?) unsual that this could be the case.
> > Now you add the restriction that 'high profile' means that the product
> > generated media attention for the tools that had been used. Sorry, I
> > can't follow you.
>
> I failed in my translation attempt, obviously, because you're completely
> missing the gist, which I had hoped would become clear when I questioned
> whether that whole process was good.
IOW, I am confident Ron used the term "high profile" to mean something
other than what I believe you (based on your response) interpreted it
to mean. I should not have said "translation" as that might have conveyed
an impression that I didn't intend. I was trying to be clever.
--
Jack Unrue
In article <··································@4ax.com>,
Jack Unrue <·······@example.tld> wrote:
> On 1 Sep 2006 17:44:28 -0700, ·······@corporate-world.lisp.de"
> <······@corporate-world.lisp.de> wrote:
> >
> > Ron Garret wrote:
> > >
> > > I don't wish to argue that Naughty Dog was a failure, but under no
> > > circumstances can you consider it a *high profile* success.
> >
> > It is a HUGE high profile success. You don't have an idea what you
> > are talking about. Sony made Crash Bandicoot their (unofficial)
> > platform masco for the playstation. The games were just amazing.
> > Just read the critics. Should I look for some reviews for you?
> > Otherwise Google is your friend.
>
> Let me see if I can help with a translation. "High profile" in
> in this context means people who put valuations on potential
> investments will have been able, for a few months at least, to
> read articles in Fortune or Inc or Technology Review about companies
> that use Common Lisp, where the aim of said articles is to position
> those companies as "the next big thing" and the founders' use of
> Common Lisp as evidence of what great "out of the box" thinking
> those brilliant fellows are capable of.
It doesn't have to be Fortune. It just has to be on the record
somewhere that the principles got rich and attribute their success at
least in part to the use of Lisp.
> Then certain well-known and prolific "industry thought leaders"
> would have to start banging the drum about Common Lisp, writing
> articles and blog entries and giving seminars, leading to a new
> generation of consultants springing up around Common Lisp.
>
> And so finally big name consultancies like Accenture would
> eventually feel compelled to start recommending Common Lisp to their
> clients. And big name analyst groups like Gartner would start
> writing reports about how companies who did not use Common Lisp were
> at a competitive disadvantage, facing an uphill climb, etc., etc.
>
> That would be "high profile." Whether it's "good" is another
> question.
In between the scenario you describe and the present state of affairs
lies a very broad range of possibilities, at least some of which are
"good" according to certain not altogether unreasonable quality metrics.
rg
Ron Garret wrote:
>
> It doesn't have to be Fortune. It just has to be on the record
> somewhere that the principles got rich and attribute their success at
> least in part to the use of Lisp.
The articles I've read about Naughty Dog's success didn't beat the drum
of Lisp. They beat the drum of aggressively negotiating your IP
ownership, your title crediting, and not taking "no" for an answer from
publishers.
Cheers,
Brandon Van Every
From: Thomas Lindgren
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <87r6yby686.fsf@dev.null>
·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> writes:
> Just a guess, but I'd think games sold on Naughty Dogs (Crash
> bandicoot, Jak & Dexter, ...) concepts should by now have reached the
> 50 million or more mark. Multiply that with the cost of a typical game,
> add the merchandize stuff and you might get an idea how huge the
> success is. Like sales of easily more than (really a guess) two hundred
> million US$. Would that be a 'success' for you? Huge even?
Here are some revenue data from a summary of the 100 greatest console
(PS2 gen) games. It's unclear to me whether all the games in a series
are counted for revenue or just the title mentioned, but:
Ratchet & Clank: 1.1m units, $31m revenue (#49 of 100)
Jak & Daxter: 1.7m units, $49m revenue (#19)
http://www.next-gen.biz/index.php?option=com_content&task=view&id=3537&Itemid=34
Ratchet & Clank was written by Insomniac Games, but I believe they
sourced the engine from Naughty Dog. The list also includes a Crash
Bandicoot game (#22, $45m), but not made by Naughty Dog. I haven't heard
whether that one used the ND engine or not, so I excluded it.
Seems quite respectable to me.
Best,
Thomas
--
Thomas Lindgren
······@corporate-world.lisp.de wrote:
> Mallor wrote:
> >
> > I consider Naughty Dog to be a failure.
>
> Please? What are you smoking?
>
> The guys made millions. They created fantastic games which were rated
> extremely high by critics and gamers.They sold millions of copies of
> these games. Plus they sold the company for even more money to Sony.
> You can't have much better success than that. The first games were
> developed with Lisp systems and the later games even were based on a
> Lisp system running on the Playstation. Which means they sold millions
> of these Lisp-based games.
And last I heard, Sony ditched the Lisp stuff. It's not a failure to
cash in. It's a failure to demonstrate to the game industry that Lisp
has value or staying power. Nobody's beating down the door to try to
repeat what they did with Lisp. They wrote a postmortem with both
"GOAL Rules!" in What Went Right, and "GOAL Sucks!" in What Went Wrong.
Really it's a treatise on the two-edged sword of writing your own
scripting language.
> Btw., if you actually could program some Lisp, you might even get a job
> at the new company of them. But all you seem to do is to post some poor
> comments on Usenet.
Gee I guess I'm just too busy making Chicken Scheme work on Windows.
Pardon me for being such a slouch.
Cheers,
Brandon Van Every
Mallor wrote:
> ······@corporate-world.lisp.de wrote:
> > Mallor wrote:
> > >
> > > I consider Naughty Dog to be a failure.
> >
> > Please? What are you smoking?
> >
> > The guys made millions. They created fantastic games which were rated
> > extremely high by critics and gamers.They sold millions of copies of
> > these games. Plus they sold the company for even more money to Sony.
> > You can't have much better success than that. The first games were
> > developed with Lisp systems and the later games even were based on a
> > Lisp system running on the Playstation. Which means they sold millions
> > of these Lisp-based games.
>
> And last I heard, Sony ditched the Lisp stuff. It's not a failure to
> cash in. It's a failure to demonstrate to the game industry that Lisp
> has value or staying power.
But nobody cares. Sony can use whatever they want to develop
for the Playstation 3. Sony did not buy Naughty Dog for the
Lisp technology. They bought the games, the characters and so on.
Anyway, I guess they are busy to have a half decent programming
platform
for the cell processor based Playstation at all.
> Nobody's beating down the door to try to
> repeat what they did with Lisp. They wrote a postmortem with both
> "GOAL Rules!" in What Went Right, and "GOAL Sucks!" in What Went Wrong.
Yeah, and then Naughty Dog created the next game (Jak 2) in Lisp.
And after that Jak 3 also got developed also in Lisp.
> Really it's a treatise on the two-edged sword of writing your own
> scripting language.
They were not writing a scripting language. They were developing a
full programming system for platform games for the Playstation 2 in
Lisp
(compiler, runtime, libraries).
>
> > Btw., if you actually could program some Lisp, you might even get a job
> > at the new company of them. But all you seem to do is to post some poor
> > comments on Usenet.
>
> Gee I guess I'm just too busy making Chicken Scheme work on Windows.
> Pardon me for being such a slouch.
Wow, I never thought you would write one line of Lisp code. Well, show
us
some screenshots when your application is ready. Everybody
here likes screenshots.
>
>
> Cheers,
> Brandon Van Every
······@corporate-world.lisp.de wrote:
> Mallor wrote:
> >
> > And last I heard, Sony ditched the Lisp stuff. It's not a failure to
> > cash in. It's a failure to demonstrate to the game industry that Lisp
> > has value or staying power.
>
> But nobody cares. Sony can use whatever they want to develop
> for the Playstation 3. Sony did not buy Naughty Dog for the
> Lisp technology.
Bingo.
> > Really it's a treatise on the two-edged sword of writing your own
> > scripting language.
>
> They were not writing a scripting language. They were developing a
> full programming system for platform games for the Playstation 2 in
> Lisp (compiler, runtime, libraries).
It was not good enough to use for the core 3D rendering, so I don't
know what definition of "full programming system" you're using here.
> > > Btw., if you actually could program some Lisp, you might even get a job
> > > at the new company of them. But all you seem to do is to post some poor
> > > comments on Usenet.
> >
> > Gee I guess I'm just too busy making Chicken Scheme work on Windows.
> > Pardon me for being such a slouch.
>
> Wow, I never thought you would write one line of Lisp code. Well, show
> us
> some screenshots when your application is ready. Everybody
> here likes screenshots.
Why don't you go look at the fucking Chicken repository meanwhile. Oh
nevermind. You'll wait for the official tarball that unifies the old
and new build systems. Oh nevermind, you don't actually care about my
actual level of productivity.
Cheers,
Brandon Van Every
On 2006-09-02 12:02:06 +0100, ·······@corporate-world.lisp.de"
<······@corporate-world.lisp.de> said:
> But nobody cares. Sony can use whatever they want to develop
> for the Playstation 3. Sony did not buy Naughty Dog for the
> Lisp technology. They bought the games, the characters and so on.
> Anyway, I guess they are busy to have a half decent programming
> platform
> for the cell processor based Playstation at all.
That's an important point. The PS2 was a big success (with ND games
using Lisp). We don't yet know that the PS3 will be, and all the
intimations I've seen are that it may not be the dominant platform -
too expensive, too late, too hard to program etc etc.
Additionally I think there's the issue that ND was sold to Sony who
eventually stopped using Lisp. Well, so what? When you sell your
company you do it because you want to cash in your investment in time
and hard work and typically go on and do something else. You pretty
much know that whoever buys it is going to screw it around at some
point, but you need to stop worrying about that and go on to the next
thing / retire or whatever you want to do. The tool they used helped
make them a lot of money: perhaps not billions, but more than anyone
posting to this thread has ever made.
--tim
Tim Bradshaw wrote:
> The tool they used helped
> make them a lot of money: perhaps not billions, but more than anyone
> posting to this thread has ever made.
But there are people using C++ and even Java that make a lot of money.
This hasn't got anything to do with the language of choice.
Not for using Lisp you are going to become rich. This is only a Graham
mith.
"Javier" <·······@gmail.com> writes:
> Tim Bradshaw wrote:
>> The tool they used helped
>> make them a lot of money: perhaps not billions, but more than anyone
>> posting to this thread has ever made.
>
> But there are people using C++ and even Java that make a lot of money.
> This hasn't got anything to do with the language of choice.
> Not for using Lisp you are going to become rich. This is only a Graham
> mith.
Exactly!
So while doing what it takes to become rich, why not have some fun
using Lisp instead of a crappy programming language?
And don't imagine you'll get rich with 20 minutes of programming, be
it lisp or ruby.
--
__Pascal Bourguignon__ http://www.informatimago.com/
I need a new toy.
Tail of black dog keeps good time.
Pounce! Good dog! Good dog!
Pascal Bourguignon wrote:
> So while doing what it takes to become rich, why not have some fun
> using Lisp instead of a crappy programming language?
Sure!
So we are here. (Although most of us are not going to get rich, sure we
are having some fun!)
On Fri, 01 Sep 2006 21:57:36 +0200, Mallor <···········@gmail.com> wrote:
>
> What the world needs is a killer app that has to get done in Lisp
> because one simply isn't gonna achieve it with a pile of code monkeys.
> The competitive advantage has to be sustainable, not just a one hit
> wonder.
>
I think you overstate Lisp's importance in these failures.
The truth is that 80% of all small businesses fail in the first
5 years. Usually due to poor business management.
Since mostly small businesses use Lisp you would expect
the failure rate to be high.
What I am missing is the link.
How is it that choosing Lisp leads to a failure to
develop a successful product?
What would be surprising is if most people who chose to program in
Lisp succeeded just because they chose Lisp..
In a sense autocad is a killer app.
They use Lisp as a application language.
Emacs of course.
Chystler's CIM software is written in Lisp I think though I am having some
problem finding it on the net.
I guess the real problem is that Lisp isn't typically used to write
end user applications. So you are not likely to find much desktop
software written in lisp. Prime Trader is one exception.
Lisp also has a faily good hold in Bio-Informatics.
Again I don't think you will see a killer app coming from this.
Why is it so important to you whether people typically choose
C++ or Java for desktop apps? The reason for this is fairly simple.
The first language I used extensively on my PC was Pascal, the C then C++.
With this background it is fairly easy to pick up Java or C#.
Lisp was not a issue at the time I learned this because it was to
big. (Lisp is no bigger than Java but Java is easier to move to if you
already know C/C++)
This is a course no a issue any more. The point is the
large amount of programmers that already know these languages
and the fact that you need to support and maintain software already written
in these languages.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
From: Tim Bradshaw
Subject: About six months (was Re: Joel hammers the final nail into Lisp's coffin)
Date:
Message-ID: <eda6ep$msu$1$830fa79d@news.demon.co.uk>
On 2006-09-01 19:10:12 +0100, Pascal Costanza <··@p-cos.net> said:
> You have cited the wrong part, which means that you have probably
> missed the point. The right part - indeed, the only part where the
> author actually claims knowledge - is this:
>
> "What I do know for sure, though, is two things:
>
> * People all over the world are constantly building web applications
> using .NET, using Java, and using PHP all the time. None of them are
> failing because of the choice of technology.
>
> * All of these environments are large and complex and you really need
> at least one architect with serious experience developing for the one
> you choose, because otherwise you'll do things wrong and wind up with
> messy code that needs to be restructured."
And what he missed completely is that anyone can build the thousandth
web application, which is much the same as 992 of the rest, using the
tools developed to write the rest, basing it on the books written by
people who wrote the rest and probably hiring some of the people who
wrote the rest. They can even get paid reasonably well to do it.
Writing (almost) the first is a different story. There are no tools,
there are no books, there is no received wisdom No one knows how it
should be done. But if you've spotted something people want you stand
to get rich. But you don't have long, because you're likely not that
far ahead of the herd: may be you have about six months before they
catch you.
Lisp can buy you about six months (no more): look for the places where
six months makes a difference.
--tim
From: Friedrich Dominicus
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <87irk2wb8l.fsf@flarge.here>
Pascal Costanza <··@p-cos.net> writes:
> What he doesn't seem to know is that people are also building web
> applications in Lisp, Scheme, Smalltalk - those are tho ones I know
> for sure - and probably many others. Considering this, the only really
> relevant part is that you need developers with serious experience in
> at least one particular technology. End of story.
Good point.
>
> Indeed, if you are an inexperienced Lisp programmer, you will probably
> fail, at least the first time around. That's true for any language
> and/or combination of technologies, because _programming is hard_.
That's true, but he certainly has a few points.
- libraries are important and those for the there mentioned languages
are surely large and more usefule then most of the Common Lisp
libraries.
>
> It's true that Joel Spolsky makes claims that go beyond this simple
> observation, but they are unfounded.
I think he forget a simply thing. If you go the C#, Java, PHP way you
can not expect to really gain something against anyone in the long
run. Well yes if you a are super brilliant Java programmer you'll be
better then "Joe Java Average" programmer. But Java solutions will be
all very similiar, you will build Object hierachies and that's
it. You'll think you'll be fine with the OO way, however there are
quite a lot of shortcomings in doing it all the OO way.
Of course you can use some language compiling to Java which offers
HOFS and the like but I doubt those really get a share, so at least
you do have at least on degree of freedom less then with a language
supporting functional programming also.
Now it gets interesting, how does one have to judge Macros? There are
just a handful of languages offering something like that, so how much
do macros buy you?
I'd argue serious Lisp programmers can write macros and they know how
to use them and this gets them another degree of freedom and at least
the possibilty for another round of abstractions. But it's for sure
that Macros are harder to understand then some "library" class.
Now we get in very muddy waters and that's what Joel has missed
really. There is not hint in his posting that he knows about either Smalltalks, Ocaml or
Lisp so he judges about something he does not know and that's really
the thing he really missed. Now let's see his language list he
mentioned C# (really a choice on a system besides Windows), Java, PHP
and Python. But the all are very similiar. Someone knowing Java will
be learning C# in nearly no time, and doing the step from Java to
Python is not really a challenge.
However the way to Lisp, Ocaml and Smalltalk will be quite a different
one, obviously he has not even tried. So his posting in the end is FUD
(at least about the language he has not clue about).....
Another point, which was not discussed, the wrote about a in-house
developed language, and why it was necesarry, IMHO this arguments are
also more then weak, if they know PHP why shouldn't it be possible to
use that for example?
And another "weak" points. Efficiency: Choosing Java or C# is not
really that much better then any decent scripting language and
discounting Perl which is in "use" for ages just shows certain dislike
for "anything" beeing a bit different.
So in the end he shows quite a bit of ignorance, half baked opinions
and prejudices. However he's not alone with that..., but this opinions
are quite common and such are the decisions made.....
You do not get fired while choosing "IBM, HP, Microsoft, Intel ....."
Regards
Friedrich
--
Please remove just-for-news- to reply via e-mail.
From: Pascal Costanza
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <4m5rhsF4mmv9U1@individual.net>
Friedrich Dominicus wrote:
> Pascal Costanza <··@p-cos.net> writes:
>
>> Indeed, if you are an inexperienced Lisp programmer, you will probably
>> fail, at least the first time around. That's true for any language
>> and/or combination of technologies, because _programming is hard_.
> That's true, but he certainly has a few points.
> - libraries are important and those for the there mentioned languages
> are surely large and more usefule then most of the Common Lisp
> libraries.
I don't think so. Maybe you have to look a little harder, and maybe you
have to pay for it, but all the important stuff exists. And this is
actually not that different in other languages, it only appears to be
different. The prepackaged libraries for a given language implementation
are not necessarily the best for the task at hand, so you are taking the
risk that you use a solution that's not optimal. If you don't want to
take that risk, you basically also have to spend the time on doing your
research. That's why I also typically suggest to pick a particular CL
implementation and use its libraries to their fullest advantage.
And always remember: time is also money.
> Now it gets interesting, how does one have to judge Macros? There are
> just a handful of languages offering something like that, so how much
> do macros buy you?
Macros are pretty useful for building domain-specific abstractions, but
I think that other features are more important. I'd vote on the
interactive, bottom-up development style that Lisp enables, and the fact
that Lisp doesn't impose any structure on your programs, but that you
can structure them in any way you like.
> I'd argue serious Lisp programmers can write macros and they know how
> to use them and this gets them another degree of freedom and at least
> the possibilty for another round of abstractions. But it's for sure
> that Macros are harder to understand then some "library" class.
I disagree. Well-implemented macros make code easier to understand, not
harder. And that's the case for any construct for building abstractions:
if you don't use them appropriately, they make the code harder to
understand. Note that it's always the programmer who controls whether
some code is hard to understand or not.
There's an important difference to single-paradigm languages here: when
you have only one way to build abstractions, you cannot make any
mistakes when choosing the right one. ;) But the code will still be hard
to understand when it's not the appropriate one...
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
From: Friedrich Dominicus
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <87zmddv5xu.fsf@flarge.here>
Pascal Costanza <··@p-cos.net> writes:
> Friedrich Dominicus wrote:
>> Pascal Costanza <··@p-cos.net> writes:
>>
>>> Indeed, if you are an inexperienced Lisp programmer, you will probably
>>> fail, at least the first time around. That's true for any language
>>> and/or combination of technologies, because _programming is hard_.
>> That's true, but he certainly has a few points.
>> - libraries are important and those for the there mentioned languages
>> are surely large and more usefule then most of the Common Lisp
>> libraries.
>
> I don't think so. Maybe you have to look a little harder, and maybe
> you have to pay for it, but all the important stuff exists.
Well I paid for quite few implementations, however they do not have
the stuff I was interested in and many of the free ones can just be
used by "Lisp hackers", if you are are one fine for you, then you
certainly can get along with them.
> And this
> is actually not that different in other languages, it only appears to
> be different.
Well my experience have shown me something different.
>The prepackaged libraries for a given language
> implementation are not necessarily the best for the task at hand, so
> you are taking the risk that you use a solution that's not optimal.
This is an argument I do not like. And it's probably the way I like to
"use" things. I'm not very keen on implementing all that stuff by
myself and I'm quite happy to use pre-existing software and I'd argue
if it's in widespread use it's certainly better then an ad-hoc
implementation from myself.
>If
> you don't want to take that risk, you basically also have to spend the
> time on doing your research.
I spend certainly a few months on learning the stuff I wanted to use,
however it has not lead very far. It always was a fight to get things
running as I liked them.
> That's why I also typically suggest to
> pick a particular CL implementation and use its libraries to their
> fullest advantage.
Oh, I did so and was in the end quite happy with a commercial
offer. This software has worked, it has needed much more "hand-work"
then I liked but it was understandable and one could apply it. So yes
in the end I found something, however in other languages I found much
ore advanced stuff, at least as easy to use while I know the language
and it was easy to apply it to myself. So I could get something the
way I liked it up and running in a week which has took me at least
three times the time doing it with the commercial Lisp offering and
more than 20 times as long as the other solution I tried to apply.
Fine if your experience is a better one, however that's mine and
that's very convinding not to follow the CL route.
>
> And always remember: time is also money.
Great, I paid for the solution and needed more time to implement what
I wanted to do than with a free alternative. So yes I "burned" money
and my time on using Common Lisp in this area....
>
> Macros are pretty useful for building domain-specific abstractions,
> but I think that other features are more important. I'd vote on the
> interactive, bottom-up development style that Lisp enables, and the
> fact that Lisp doesn't impose any structure on your programs, but that
> you can structure them in any way you like.
Well this argument still holds for Smalltalk, Ocaml, Erlang and even
Ruby.
>
>
> I disagree. Well-implemented macros make code easier to understand,
> not harder.
Not always, try to follow the flow of control in UCW and then follow
the same stuff in Webactions...
> And that's the case for any construct for building
> abstractions: if you don't use them appropriately, they make the code
> harder to understand. Note that it's always the programmer who
> controls whether some code is hard to understand or not.
Well just one point a Macro is not the code you really work on, you
have to "step" into it to see the code behind.
>
> There's an important difference to single-paradigm languages here:
> when you have only one way to build abstractions, you cannot make any
> mistakes when choosing the right one. ;) But the code will still be
> hard to understand when it's not the appropriate one...
That's true to some degree, but you do not get so many suprises, you
know you are in a certain swing and it can't change suddenly that's
surely not true for Common Lisp solutions...
Regards
Friedrich
--
Please remove just-for-news- to reply via e-mail.
From: Pascal Costanza
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <4m8nesF54lh5U1@individual.net>
Friedrich Dominicus wrote:
> Pascal Costanza <··@p-cos.net> writes:
>
>> Friedrich Dominicus wrote:
>>> Pascal Costanza <··@p-cos.net> writes:
>>>
>>>> Indeed, if you are an inexperienced Lisp programmer, you will probably
>>>> fail, at least the first time around. That's true for any language
>>>> and/or combination of technologies, because _programming is hard_.
>>> That's true, but he certainly has a few points.
>>> - libraries are important and those for the there mentioned languages
>>> are surely large and more usefule then most of the Common Lisp
>>> libraries.
>> I don't think so. Maybe you have to look a little harder, and maybe
>> you have to pay for it, but all the important stuff exists.
> Well I paid for quite few implementations, however they do not have
> the stuff I was interested in and many of the free ones can just be
> used by "Lisp hackers", if you are are one fine for you, then you
> certainly can get along with them.
Can you be more specific about what didn't work?
>> The prepackaged libraries for a given language
>> implementation are not necessarily the best for the task at hand, so
>> you are taking the risk that you use a solution that's not optimal.
> This is an argument I do not like. And it's probably the way I like to
> "use" things. I'm not very keen on implementing all that stuff by
> myself and I'm quite happy to use pre-existing software and I'd argue
> if it's in widespread use it's certainly better then an ad-hoc
> implementation from myself.
Hm, I am trying to guess what you are trying to get at here. Are you
missing a working component model for Lisp?
>> Macros are pretty useful for building domain-specific abstractions,
>> but I think that other features are more important. I'd vote on the
>> interactive, bottom-up development style that Lisp enables, and the
>> fact that Lisp doesn't impose any structure on your programs, but that
>> you can structure them in any way you like.
> Well this argument still holds for Smalltalk, Ocaml, Erlang and even
> Ruby.
Yes, but I didn't say that macros are not important. ;)
>> And that's the case for any construct for building
>> abstractions: if you don't use them appropriately, they make the code
>> harder to understand. Note that it's always the programmer who
>> controls whether some code is hard to understand or not.
> Well just one point a Macro is not the code you really work on, you
> have to "step" into it to see the code behind.
That's also the same for other constructs.
>> There's an important difference to single-paradigm languages here:
>> when you have only one way to build abstractions, you cannot make any
>> mistakes when choosing the right one. ;) But the code will still be
>> hard to understand when it's not the appropriate one...
> That's true to some degree, but you do not get so many suprises, you
> know you are in a certain swing and it can't change suddenly that's
> surely not true for Common Lisp solutions...
Yes, I think that's a valid advantage of single-paradigm languages, and
that's the central trade-off between singe- and multi-paradigm languages.
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
From: Friedrich Dominicus
Subject: Re: Joel hammers the final nail into Lisp's coffin
Date:
Message-ID: <873bb444d8.fsf@flarge.here>
Pascal Costanza <··@p-cos.net> writes:
> Friedrich Dominicus wrote:
>> Pascal Costanza <··@p-cos.net> writes:
>>
>>> Friedrich Dominicus wrote:
>>>> Pascal Costanza <··@p-cos.net> writes:
>>>>
>>>>> Indeed, if you are an inexperienced Lisp programmer, you will probably
>>>>> fail, at least the first time around. That's true for any language
>>>>> and/or combination of technologies, because _programming is hard_.
>>>> That's true, but he certainly has a few points.
>>>> - libraries are important and those for the there mentioned languages
>>>> are surely large and more usefule then most of the Common Lisp
>>>> libraries.
>>> I don't think so. Maybe you have to look a little harder, and maybe
>>> you have to pay for it, but all the important stuff exists.
>> Well I paid for quite few implementations, however they do not have
>> the stuff I was interested in and many of the free ones can just be
>> used by "Lisp hackers", if you are are one fine for you, then you
>> certainly can get along with them.
>
> Can you be more specific about what didn't work?
cl-xml (applying)
ucw (applying, what has to be defined when and where?, debug traces
longer than just a few pages, finding any problem in it)
Regards
Friedrich
--
Please remove just-for-news- to reply via e-mail.
Friedrich Dominicus <···················@q-software-solutions.de> writes:
> I'd argue serious Lisp programmers can write macros and they know how
> to use them and this gets them another degree of freedom and at least
> the possibilty for another round of abstractions. But it's for sure
> that Macros are harder to understand then some "library" class.
It's possible that some macro be harder to understand. But good
macros, even if they're harder to understand themselves, should render
the code using them easier to understand.
Somebody might not understand the quicksort algorithm, but can still
understand and use easily the SORT function.
--
__Pascal Bourguignon__ http://www.informatimago.com/
THIS IS A 100% MATTER PRODUCT: In the unlikely event that this
merchandise should contact antimatter in any form, a catastrophic
explosion will result.
On 2006-09-01 11:33:15 -0500, Anonymous coward #673 <·······@devnull.com> said:
> http://www.joelonsoftware.com/items/2006/09/01.html
>
> The best part:
>
> "Oh and I know Paul told you that he made his app in Lisp and then he
> made millions of dollars because he made his app in Lisp, but honestly
> only�two people ever believed him and, a complete rewrite later, they
> won't make that mistake again."
... since it underscores Spolsky's willingness to sacrifice accuracy
for sake of wit. Of course, he isn't the only Boy Band Hacker to
talk such nonsense. Fortunately, there *are* authentic, skilled,
and experienced programmers like John Foderaro to tell it like
it is:
http://www.aaronsw.com/weblog/rewritingreddit#c29
-Mike
--
Michael J. Forster <····@sharedlogic.ca>
Michael J. Forster <····@sharedlogic.ca> wrote:
> talk such nonsense. Fortunately, there *are* authentic, skilled,
> and experienced programmers like John Foderaro to tell it like
> it is:
>
> http://www.aaronsw.com/weblog/rewritingreddit#c29
It was not the choice of implementation.
When you want to run a high traffic site you should develop on
the same systems as the production servers. There are so many
differences between operating systems (and even distributions of
Linux!) that this is way to risky.
Hey, you can't even use .NET on different versions of Windows
without risk!
On 2006-09-02 02:49:32 -0500, Stefan Scholl <······@no-spoon.de> said:
> Michael J. Forster <····@sharedlogic.ca> wrote:
>> talk such nonsense. Fortunately, there *are* authentic, skilled,
>> and experienced programmers like John Foderaro to tell it like
>> it is:
>>
>> http://www.aaronsw.com/weblog/rewritingreddit#c29
>
> It was not the choice of implementation.
>
> When you want to run a high traffic site you should develop on
> the same systems as the production servers. There are so many
> differences between operating systems (and even distributions of
> Linux!) that this is way to risky.
>
> Hey, you can't even use .NET on different versions of Windows
> without risk!
I would agree that developing and deploying on the same hardware
and OS would have been wiser.
However, I think a bigger issue was that they were writing, as they put it,
"low-level socket and threading code..." Why were they doing this for
such a simple app? It would be like a PHP programmer trying to write
reddit and taking a detour to write his own SQL DBMS because he
disliked MySQL. A solid business plan would have all but dictated that
they choose an implementation and libs that would see them
writing no more detour code than we saw Sven Van Caekenberghe
write in his video.
Moreover, you know as well as I do that computing today is a cross-
platform endeavour. Even if Java failed on the write-once-run-everywhere
promise, as programmers we are not off the hook. We have to select
the right tools and understand the platform differences intimately. Often,
that means developing and deploying on different platforms.
My point is that a better implementation choice would have all but
eliminated hardware and OS issues. I'm quite certain that was John
Foderaro's point as well.
Regards,
Mike
--
Michael J. Forster <····@sharedlogic.ca>
Le Sat, 02 Sep 2006 07:20:05 -0500, Michael J. Forster a écrit :
> However, I think a bigger issue was that they were writing, as they
> put it, "low-level socket and threading code..." Why were they doing
> this for such a simple app?
Hell, since I begun reading about this reddit scandal, I was wondering
why none asked that question! ;-)
I really don't see where they had to hack such low-level things. It
seems to me they took the wrong route with Lisp, and took another while
switching to Python.
Did they explain why they had to mess with threads and sockets
themselves?
Curiously,
Nowhere man
--
···········@levallois.eu.org
OpenPGP 0xD9D50D8A
On 2006-09-04 11:00:24 -0500, Pierre THIERRY
<···········@levallois.eu.org> said:
> Le Sat, 02 Sep 2006 07:20:05 -0500, Michael J. Forster a �crit�:
>> However, I think a bigger issue was that they were writing, as they
>> put it, "low-level socket and threading code..." Why were they doing
>> this for such a simple app?
>
> Hell, since I begun reading about this reddit scandal, I was wondering
> why none asked that question! ;-)
Me too :-). The thing seemed to spawn thread-after-lengthy-thread
about Lisp library deficiencies, platform compatibility, and etc.; all of
which completely missed the real problem.
> I really don't see where they had to hack such low-level things. It
> seems to me they took the wrong route with Lisp, and took another while
> switching to Python.
>
> Did they explain why they had to mess with threads and sockets
> themselves?
>
> Curiously,
> Nowhere man
I haven't seen or heard of one. However, since I would be dubious
of any explanation for such an obvious mistake, I haven't gone looking
either ;-)
I suspect, from reading their announcement, that a predilection for
Open Source tools might have been a factor. Again, however, there are
Open Source implementations and libs (paserve) that would have
obviated the low-level hacking. Or, perhaps, it was a matter of keeping
startup costs down. That, at least, I can respect.
-Mike
--
Michael J. Forster <····@sharedlogic.ca>
Michael J. Forster <····@sharedlogic.ca> wrote:
> Open Source tools might have been a factor. Again, however, there are
> Open Source implementations and libs (paserve) that would have
> obviated the low-level hacking. Or, perhaps, it was a matter of keeping
> startup costs down. That, at least, I can respect.
Last time I tested portable aserve with CMUCL I had to send in a
patch. (Due to someone making some Debian-only changes ... :-/)
Maybe they tried paserve and hadn't the nerves to continue after
the first errors.
--
Web (en): http://www.no-spoon.de/ -*- Web (de): http://www.frell.de/
On 2006-09-06 00:45:17 -0500, Stefan Scholl <······@no-spoon.de> said:
> Michael J. Forster <····@sharedlogic.ca> wrote:
>> Open Source tools might have been a factor. Again, however, there are
>> Open Source implementations and libs (paserve) that would have
>> obviated the low-level hacking. Or, perhaps, it was a matter of keeping
>> startup costs down. That, at least, I can respect.
>
> Last time I tested portable aserve with CMUCL I had to send in a
> patch. (Due to someone making some Debian-only changes ... :-/)
>
> Maybe they tried paserve and hadn't the nerves to continue after
> the first errors.
Possibly. However, I don't think Linux was in the mix:
"... but I develop on a Mac, and reddit.com is a FreeBSD box." [1]
and we haven't had a problem with CMUCL and paserve on FreeBSD,
though I should note that we went for the latest CVS straight away rather
than the tarball.
-Mike
[1] http://redditblog.blogspot.com/2005/12/on-lisp.html
--
Michael J. Forster <····@sharedlogic.ca>
On Fri, 01 Sep 2006 09:33:15 -0700, Anonymous coward #673
<·······@devnull.com>:
>http://www.joelonsoftware.com/items/2006/09/01.html
One thing is very nice about this blog: since he can say interesting
things and complete bullshit in the same article, you always have to
think a lot after reading. Maybe that's the point.
I won't comment about Lisp, but when he says that Ruby is to slow
while PHP is acceptable, well... as I say, everything in his articles
doesn't always make sense...
Le Fri, 01 Sep 2006 20:50:53 +0200, Fabien LE LEZ a écrit :
> I won't comment about Lisp, but when he says that Ruby is to slow
> while PHP is acceptable, well... as I say, everything in his articles
> doesn't always make sense...
At least he's consistent with some benchmarks. Go check shootout:
http://shootout.alioth.debian.org/
On many tests, Ruby is dramatically slower than anything. Wether the
benchmakr itself makes sense is another story, though.
Factually,
Nowhere man
--
···········@levallois.eu.org
OpenPGP 0xD9D50D8A
On Sat, 02 Sep 2006 00:40:32 +0200, Stefan Scholl <······@no-spoon.de>
wrote:
> Damn! Lisp is dead (again)!
>
No deader than usual.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
Anonymous coward #673 wrote:
> http://www.joelonsoftware.com/items/2006/09/01.html
>
> The best part:
>
> "Oh and I know Paul told you that he made his app in Lisp and then he
> made millions of dollars because he made his app in Lisp, but honestly
> only two people ever believed him and, a complete rewrite later, they
> won't make that mistake again."
I wouldn't really worry about what Joel says. He is sometimes
correct and sometimes misinformed.
--
goose
Have I offended you? Send flames to ····@localhost
real email: lelanthran at gmail dot com
website : www.lelanthran.com
In article <············@ctb-nnrp2.saix.net>,
goose <··············@webmail.co.za> wrote:
> Anonymous coward #673 wrote:
> > http://www.joelonsoftware.com/items/2006/09/01.html
> >
> > The best part:
> >
> > "Oh and I know Paul told you that he made his app in Lisp and then he
> > made millions of dollars because he made his app in Lisp, but honestly
> > only two people ever believed him and, a complete rewrite later, they
> > won't make that mistake again."
>
> I wouldn't really worry about what Joel says. He is sometimes
> correct and sometimes misinformed.
But his blog is read by very many people.
Anonymous coward #673 wrote:
> In article <············@ctb-nnrp2.saix.net>,
> goose <··············@webmail.co.za> wrote:
>
>
>>Anonymous coward #673 wrote:
>>
>>>http://www.joelonsoftware.com/items/2006/09/01.html
>>>
>>>The best part:
>>>
>>>"Oh and I know Paul told you that he made his app in Lisp and then he
>>>made millions of dollars because he made his app in Lisp, but honestly
>>>only two people ever believed him and, a complete rewrite later, they
>>>won't make that mistake again."
>>
>>I wouldn't really worry about what Joel says. He is sometimes
>>correct and sometimes misinformed.
>
>
> But his blog is read by very many people.
I'm really too old to be worried about popularity
contests (and if anyone here was, they wouldn't
be interested in Lisp in the first place).
--
goose
Have I offended you? Send flames to ····@localhost
real email: lelanthran at gmail dot com
website : www.lelanthran.com
goose <··············@webmail.co.za> wrote:
> I'm really too old to be worried about popularity
> contests (and if anyone here was, they wouldn't
> be interested in Lisp in the first place).
Are you sure? Lisp is cool. People who use Lisp are called
wizards by lesser lifeforms.
Anonymous coward #673 wrote:
> http://www.joelonsoftware.com/items/2006/09/01.html
>
> The best part:
>
> "Oh and I know Paul told you that he made his app in Lisp and then he
> made millions of dollars because he made his app in Lisp, but honestly
> only two people ever believed him and, a complete rewrite later, they
> won't make that mistake again."
I think he is right up to a point.
The reason because Reddit abandoned Lisp is not because the language is
not powerfull, nor because the didn't like it (they said so), but
because the lack of support from the Lisp comunity.
They could spend some money buying a comercial version, but I think
this in not the point. Most webmasters in the world are not going to
pay for a language when there are free, well supported free
implementations of other ones.
What happened for Reddit is that when they were going to mantain their
code, they found themselves with poor or no support at all from the
comunity. They even had to write threads and other things that there
are supposed to be already there for any implementation.
Another important issue about the comunity, not only its lazyness, is
how they are wasting their time instead of working. You can find a lot
of blogs, webpages, forums, articles about how impresively good is
Lisp, but vey little support for what really matters: libraries and
implementations. Take for example c.l.l and compare to c.l.python or
c.l.java, or even c.l.perl. In c.l.l about 50% of posts are dedicated
to criticique other languages, and to say how good is Lisp. In those
other newsgroups there are 5% or less, and people is actually working.
You can go for Python, Perl, and Java, and have good support, well
finished libraries, free and complete implementations, and a lot of
people which are polite and believe in the filosophy of mutual help.
We are in the point of being a better comunity. We've got the best
language, but really need to do a lot of work in other areas. This is
very difficult.
Paul Graham has done right his job. He is probably not good doing
"core" work by himself, but very capable of promoting Lisp. But his
work of promoting will be totally wasted if others, which are good at
writing code, don't do what they have to do. And I know there are a lot
of people working very hard to actually write code, but infortunaly it
is not enoght, because there are others that waste to much time in
front of GNUS and very little in front of SLIME.
Just my opinion, and I admit that in some way I also have part of
fault, even if I am a newbie in Lisp.
Javier <·······@gmail.com> wrote:
> The reason because Reddit abandoned Lisp is not because the language is
> not powerfull, nor because the didn't like it (they said so), but
> because the lack of support from the Lisp comunity.
Hmm??
http://common-lisp.net/pipermail/tbnl-devel/2005-August/000364.html
ff.
Joel has good points from time to time:
http://www.joelonsoftware.com/articles/HighNotes.html
But he is hardly much more than just another pundit.
--
This post uses 100% post consumer electrons and 100% virgin photons.
At 2.6 miles per minute, you don't really have time to get bored.
--- Pete Roehling on rec.motorcycles
I bump into a lot of veteran riders in my travels.
--- David Hough: Proficient Motorcycling