This newsgroup is getting more and more useless because of the
"Peaceman". In the days where the most prominent "software war" rages
between the mega company we all love to hate and a smaller contender
who is just at bad at writing efficient software, claiming "lisp is
slow" is in this context not only false, but completely ridiculous,
when you watch what monsters people are able to create with those
supposedly good tools.
I'm working as a systems developer for an internet provider. Although
I used to use Lisp for research-oriented programming, I find myself
hacking a lot of perl code these days. I _do_ have CL-HTTP running
on an in-house web-server, though.
So here's my wishlist for making Lisp more useful for my usage:
- I'd love a Lisp replacement for inetd, where services are implemented
as lisp functions with stream arguments, and you don't have to
worry much about low-level socket handling. Just like CL-HTTP frees
you from the digusting habit of calling the perl compiler all the
time, the Lisp inetd would do the same for general internet services!
(the lisp inetd would probably be a little different, it could for
instance be configurable to either fork off a process (which, hopefully,
could share most of its memory with the mother process?), or, for
small services which should run as root anyway, it could maybe run
them "inline", as threads, to reduce overhead)
- In general, I'd like Lisp to have powerful tools for handling sockets.
- I'd like a powerful and easy-to-use regular expression package.
- I'd like all this to be standarized and be part of the Next Generation
Common Lisp Standard (which should also include multithreading along
the lines of e.g. Allegro CL or MCL).
- I think I even want a Lisp OS again, 9 years after the last
time I used an Interlisp-D machine for anything useful, probably
it should sit on top of some kind of unix kernel.
Because of all the noise in this group, I tend not to get the over-
view of ongoing useful work, so I hope that someone could summarize:
- Are there current significant projects for extending the Common
Lisp Standard?
- Do any of them work on sockets (or string manipulation tools)?
I apologise if these questions have just been answered, I should
probably start using a kill-file to make me able to read this
group again :-(
(in general, I don't like kill-files, partly because I think
using non-semantic filters is a dangerous thing to do
to your information. And comp.lang.lisp _used_ to be a
group where you could browse through the headers in a few
seconds - and actually most of the articles would be quite
interesting!)
--
regards,
Espen Vestre
In article <··············@gromit.online.no>, Espen Vestre <··@nextel.no> wrote:
> - I'd like a powerful and easy-to-use regular expression package.
It would be nice in the CL-HTTP context, too. It should be
a fast version.
> - I'd like all this to be standarized and be part of the Next Generation
> Common Lisp Standard (which should also include multithreading along
> the lines of e.g. Allegro CL or MCL).
Yep.
> - I think I even want a Lisp OS again, 9 years after the last
> time I used an Interlisp-D machine for anything useful, probably
> it should sit on top of some kind of unix kernel.
This is what some other people also want to do. There is a mailing
list for this purpose. It has slowed down, though.
> - Are there current significant projects for extending the Common
> Lisp Standard?
Not that I have heard of. Others?
-----------------------------------------------------------------------
We are thinking about our next Lisp User Group meeting here in Germany.
Yes, it will happen again!
Last time we met in Munich and most people were positively
surprised to see many other developers. The last topic was focused
on usage of Lisp for the Internet and related things (distributed
Lisp, Web site maintenance, CL-HTTP applications, new programming
styles for highly dynamic services, Statice, ...).
What would be a good topic for the next Lisp User Group meeting
early next year here in Germany?
I was thinking about:
1) New Common Lisp.
How would a new version of the Common Lisp
standard look like and what would be the micro steps to
move into that direction. Talks should discuss the
users perspective or provide ideas how to achieve
(to be defined) goals.
2) Common Lisp in an unfriendly world.
How to survive with Lisp in a non Lisp world. There
are two ways to survive: fit perfectly in the rest
of the computing world or build something better.
Currently PCs are dominating. Especially popular
are operating systems and standards provided by
Microsoft. Still a decent Lisp system for the PC
is not available. How would a perfect Lisp system
for the PC look like and what standards should
be supported? Users should get organized. Visibility
in the market should be improved.
3) New Lisp Platform
Can we work on a new Lisp-system much like Linux?
Major parts of the OS written in Lisp and the
applications also in Lisp? All running on standard
hardware? How to organize the Lisp OS movement.
Topic 1) may also be of interest here. The
problems of nowadays widely used architectures
(Windows, MacOS, Unix, ...) are clear: their
users and developers can't really take
advantage of the recent hardware achievements.
Modern software is bloated, unflexible, buggy
and only with massive amount of work evolvable.
Can a new Lisp platform with an open software
approach make a difference?
4) The Internet, revisited
Lisp applications for the Internet. Companies
and users are beginning to take advantage
of Lisp-based software. How can we make it
more popular? Extensive CL-HTTP tutorial.
Comments:
---------
Topic 1) may be very Common Lisp focused and would
talk about extensions that are Lisp-specific (threads,
TCP/IP in Lisp, FFI, ...)
Topic 2) may be to Microsoft specific.
Topic 3) sounds unrealistic?
Topic 4) extends this years theme. Again?
Opinions?
Greetings,
Rainer Joswig
--
http://www.lavielle.com/~joswig/