[Gardeners] CL Gardeners: good news, bad news, how to win big

[Jeremy Smith]

On Wednesday 22 February 2006 20:51, Jeremy Smith wrote:
> On Wednesday 22 February 2006 20:19, Paolo Amoroso wrote:
> > I hope that my comments will stimulate the realization that the future
> > of Lisp is in the hands of its users.  There are probably no
> > fundamental technical reasons why Lisp is not more widespread: the
> > fragmentation of implementations, the lack of cross-platform GUI
> > toolkits, the difficulty of generating standalone executables, the
> > parentheses, the lack of standardized APIs, whatever.  It's just that
> > not enough users do something.
> >
> > Ideas and discussions are no longer enough.  Lisp needs labor, not
> > praise.
> >
> >
> > Paolo
>
> I don't know if this is on-topic, but it's related. I just noticed that
> neither Corman nor Lispworks (I haven't checked Allegro) on Windows come
> with ASDF.lisp.
>
> I'm sure it can't help Lisp propagation if none of the commercial
> distributions come with what is effectively a portable, freely
> redistributable library (only 40kb too) that is crucial to the majority of
> these CLD efforts.
>
> I'm not saying they won't put it in their distribution, just that they
> haven't, and there's no reason that I can see, for it. Maybe nobody really
> considered it before.

Hi,

I should have been more specific.

My point is that a priority is making it as easy as possible for people to get 
started with Lisp. If you can include Asdf and even load it by default when 
you start Lisp, as Lisp in a Box does, then people don't have to spend half 
an hour finding it, figuring out how to install it, and installing it.

There's an old tale about the Apple Macintosh. Steve Jobs wanted an engineer 
to increase the boot time by a second (or some figure). The engineer balked 
at the idea, but Steve explained that if a million people were booting it 
once a day, that was 365 million seconds a year wasted.

So if we can save the average Lisp user half an hour here, and half an hour 
there, they are more likely to be interested in it.

Linux falls into this trap, where it can take hours to figure out how to do 
the simplest things. Of course, Googling the problem can help, but that takes 
time too (especially these days, with some very poor results).

I think simplicity (in terms of only having the absolute minimum) is the key 
to mass popularity. Lisp (not Common Lisp) is actually a very simple language 
with a few basic concepts, so it's hard to argue that people shouldn't use it 
because it's "too complex".

My other unstated point was that interaction between the commercial Lisp 
distributions would be good for all of them. I know they're not aiming at 
hobbyists with Web 2.0 startups, they're aiming at super-scientific 
applications with serious scientists who have money to spend, but there's the 
rub: Is this Lisp Gardener's project aimed at serious Lispers, or hobbyist 
Lispers who want to run a website and don't want it to take hours to get 
ASDF, get AServe, install it, run it, cope with the possible heap limit?

Scientists pay now, but hobbyists may pay in future.

There is a big difference between these 2 markets, and figuring this out is, I 
think, important to the future of Lisp.

Peter Siebel's comment about how I should have to pay a Lisp vendor to ask 
them to install ASDF seems wrong to me, because the suggestion is not for my 
benefit, but theirs. I will just continue using Clisp in the meantime, and 
the more dependent I get on it, the less interested I am in porting all my 
code to, say, Lispworks, because it might be 20% faster.

Cheers,

Jeremy.