Hi All,

I'm someone who has been attempting to get into lisp for about a year
now. Unfortunately, my job has until recently been way too demanding
for me to make a lot of progress, but I intend to keep at it as I do
enjoy what little time I do get to work at things. 

Until recently, I was manager of a data centre at the local
University. However, after 2.5 years, I came to the conclusion I'd
rather earn less and have a more technical focus. My background is as
a programmer and developer (C, C++, Tcl/TK, perl, Java and various
database 4GLs/SQL). Most of my work has been on various *nixes and
Linux. I started working in IT in 1986. Like others on the list, I'd
class myself as entering the getting to be an old fart catagory, born
in '61. I'm also a blind programmer, having lost my sight in 1987.

I live in Armidale NSW Australia (half way between Sydney and
Brisbane). Small town of about 20k people. Very few programming jobs
and pretty much no chance of getting a lisp programming job apart from
possibly scoring a contract with someone in one of the larger cities
or overseas. I like where I live, just wish there were more
programming jobs available. Still, I have permanent work which is
relatively well paid, so I should be thankful for that. 

I'm currently in the process of changing positions with my employer,
but the full definition of my responsabilities is still being
formulated, so I'm still unsure exactly how much time I will have to
contribute to things. However, I'm very keen to help out where I
can. For at least the next 2 months, I'm covering for many of the
managers who are taking annual leave, so things will be quite
unpredictable for a time. 

So far, most of my lisp work has been pretty minor in nature and
mainly oriented to 'experiments' aimed to improve my understanding. I
have added compatibility for additional lisp implementations to a
debian lisp package called cl-ssip, which is a lisp interface to the
speech-dispatcher package. This was quite an interesting little simple
project as the original package only works with SBCL, but the original
author didn't include any compile pragmas to warn of missing
compatability with other lisps - it took me sometime to work out why I
couldn't get things to work with CMUCL until I realised there were
functions that contained no real code for any version of lisp other
than SBCL (i.e. (defun blah () +SBCL.....)), so I added CMUCL support
and additional pragmas to warn if you were compiling with some other
lisp variant. This highlited two things for me

        1. Writers of packages which contain implementation specific code
       should put compile directives in that will either throw and
       error or warn the user if the variant they are using is not
       supported. 

        2. The main reason other lisp implementations are not supported is
       because the package depends heavily on sockets and there isn't
       a good standard abstraction for sockets - every implementation
       does it differently. We need a good standardised socket library
       IMO. 

One of the most difficult problems I've encountered with lisp is in
finding the appropriate function for doing some task. Many times I've
looked for some functionality which I knew was likely to exist, but
couldn't find the appropriate function. Some of this was due to not
being familiar with lisp nameing conventions/style and part of it is
due to the large number of functions available in CL. While the
hyperspec is great once you know what your looking for, its not so
good when you have an idea or know the general area, but not the
name. I've often ended up re-inventing the wheel only to stumble
across a built in function that does the same job. While this can be a
good way to learn, it really slows down your development time and that
can be somewhat discouraging at times. 

I'd really like to see a good CL reference which is organised by
functional groups rather than alphabetical etc. Lots of people suggest
the reference/index in Paul Grahame's book, which I'm sure is
good. However, as you can only get the book in postscript form, its
not that useful re: electronic reference. I think people new to lisp
would find such a reference very useful in speeding up their learning
of the language and becomming familiar with the standard CL
libraries/functions. 

Just my 2 cents worth.

Tim
_______________________________________________
Gardeners mailing list
[email protected]
http://www.lispniks.com/mailman/listinfo/gardeners

Reply via email to