Re: Is Lisp a Blub?

On Jul 3, 10:38 am, Slobodan Blazeski <slobodan.blaze...@xxxxxxxxx>
Yes it shares a simigliar problem with lisp, lacks consumer software
since all the DESKTOP software is written for Windows only.

I think there are cultural similarities, as well. Throughout the 90s,
the Mac really was better than Windows. However, Windows was getting
significantly better, and the Mac was not.

I think the overall sense of the general public was that the Mac might
be a little bit better, but not better enough to overcome the cultural
issues of more applications for Windows, more friends who knew Windows
and could give them advice, more places to buy a Windows PC, bigger
variety of supported hardware, etc.

This left the supporters of the Macintosh in the position of defending
things that seemed more and more esoteric, even though they were real
benefits. For example, the Mac had excellent meta-data support built
into the file system. This made it easy to track what applications
could open which files, which application created a file, which icon
to show, application resources bundled with the application, and all
kinds of other useful things. Windows relied on people typing the
correct file extension, and if the user made a mistake, everything
fell apart.

In Mac OS X, Apple decided to favor file extensions over the meta data
in the resource fork for associating files with applications. Many
long time supporters of the Mac were upset at Apple for deliberately
choosing the technically inferior solution.

However, the reasons for Apple's decision were cultural, not
technical. Windows users expected to associate a file with an
application through a file extension. They did not know anything
about resource forks. So when a Mac user gave a file to a Windows
user, the Windows could not easily open it without the expected file
extension. By making the Mac more like Windows, Apple made the choice
of buying a Mac easier.

Now, I'm not saying Lisp should become more like other languages just
to fit in. But I do think that there is a perception that even if
Lisp is better in certain ways than other languages, those other
languages are improving and Lisp is not. The perception might even be
wrong. But even the perception itself is a liability for Lisp

In short, sometimes the cultural outweighs the technical.

If you think you have found any
piece of utopia on this Earth, you are probably either deceiving
yourself or need to raise your standards.

I didn't say common lisp is perfect, but is the best thing (language +
implementation + libraries + community + literature) I've found after
going through: c, c++, basic, pascal/delphi, c#, TSQL and some
python , java, php , ocaml & erlang.

I would agree here, too. Lisp is not perfect, just the best so far
(of the languages I know).

Now, to your credit, you do mention the concurrency thing. But do you
really think there is nothing else that could make Lisp better?

AllegroCache in lisp that I could afford ( 2000$ at most)

Agreed, and note that I would consider this (pricing) a cultural issue
as well. There was a recent, extended thread on making a commercial
Lisp open source. Price and open source offerings definitely affect
perceptions of Common Lisp.

Lispers are spoiled of choice, that's very bad thing . The good thing
is that any lisp, whose development didn't stoppeed in 1986, will do
for a newbie. Afterwards they'll be able to choose what suits them

Choice is a barrier to entry for a newbie. By definition as a newbie,
they do not have the information they need to make a good choice. So
the newbie can spend time evaluating the major Lisp distributions, or
just pick Python or Ruby and start hacking. There are multiple
distributions of these, but not as many and there is a default
distribution that new programmers will be steered towards.

Also, please note I clearly said this is a good and a bad thing. The
availability of Common Lisps with different tradeoffs is a good thing
for someone knowledgable about Lisp.

Ever heard of starter pack , or lisp in a box.

Yes! My preferred solution would be for Starter Pack and Lisp in a
Box packaged together with either SBCL or CLisp is a double clickable
package. The only barrier then is knowing Emacs, but I think Phil
Armitage will have that covered, based on conversation elsewhere in
this thread :).

This one thing would do more than anything else I can think of for
erasing barriers to entry for new Lisp programmers.

You use :
a: Free commercial lisp like Allegro express, LW personal
b: Lisp in a box comes with everything you need.

I tried the Lisp in a Box with Allegro express. I was using it for a
homework project and immediately blew through the Allegro memory limit
because the assignment required processing a lot of data. So I found
AquaMacs (for Mac OS X), which had SLIME pre-installed, and got it
working with SBCL. (This is now my favorite development environment,
by the way, better than any IDE I've ever used.) Not that big of a
deal, but adds just a little bit more of a barrier to entry.

I suppose this could be subsumed under item 2., though.