This is gathering some threads that arose from my comp.lang.lisp post
'On the Strange Weakness of Graphical User Programming Languages' -
threads that link up to other essays I've written.

Really I think the problem is that programming has outstripped CS
theory for a long time. We have these theories of computability
dating back to the 1930s developed by people who didn't even have
access to a computer. Elsewhere I noted in 'Lisp for the C21' (L21)

www.lambdassociates.org/lC21.htm

that there is this vaguer but equally important idea of computational
adequacy. When I wrote that piece I didn't see how important it was
to be able to define that concept in a clear and elegant way to get
clear and elegant solutions to modern systems programming problems.
I do now.

The truth is that our formal 1930s theory has long been left behind by
the pace of development of commercial software. Its odd really that
this has not been a focus at university level for longer or that
university CS has not made a bigger issue out of it. I think that part
of the problem is that universities have gone into decline at the same
time that this explosion in innovation has taken place.

On 15 Oct, 11:23, Mark Tarver ukonline.co.uk> wrote:

One thing I have noticed, especially with cross-posted messages
between c.l.l and c.l.f, is that it tends to provoke "real world vs.
academic" arguments which rather supports your view that nothing will
change.

For example, I find it hard to listen to academics at conferences
telling me how I should be writing software when they've never written
a real program for end users before.

Similarly I find it infuriating listening to industry programmers
telling me how amazing these 'new' anonymous functions are in C#.

On Oct 15, 9:41 am, philip.armit...@gmail.com wrote:

Indeed - I too have long been troubled by the increasing trend of
commercialization in higher education, though I have experience only
with the student side of the picture.

In simplest terms, a proper university research system can exist only
when there exist funding sources that are not constrained by practical
results achieved. The constraint must be quality of work performed,
which is a very different metric. You don't want poor researchers,
but you DO want researchers free to pursue directions that have no
apparent hope of return on investment. Only in that fashion can
really NEW directions be found. Knowledge must first be valued in and
of itself, as an end. Oftentimes it becomes a means as well, but at
the university research level that should not be assumed or even
desired. That's what the commercial world is for.

On 15 Oct, 16:41, C Y yahoo.com> wrote:

On 2007-10-15 06:23:48 -0400, Mark Tarver ukonline.co.uk> said:

Are you familiar with the work of Peter Wegner and his co-authors along
these lines?

<http://www.cse.uconn.edu/~dqg/papers/cie05.pdf>

<http://www.cse.uconn.edu/~dqg/papers/turing04.pdf>
(especially section 3)

Also see

Dina Golden et al. (eds.)
Interactive COmputation: The New Paradigm
Springer, 2006

which begins with an opening address by Robin Milner.

Raffael Cavallaro
il-vous-plait-mac.com> writes:

When reading the above, I think it is useful to keep in mind the
commentary that others in the scientific community have to offer on
this stuff. In particular, see Lance Fortnow's opinion and the ensuing
discussion here:

http://weblog.fortnow.com/2006/07/principles-of-problem-solving-tcs.html

On 2007-10-15 18:31:47 -0400, "David B. Benson" eecs.wsu.edu> said:

<http://www.amazon.com/Interactive-Computation-Paradigm-Dina-Goldin/dp/354034666X/sr...>

for

those with deep pockets

Mark Tarver ukonline.co.uk> writes:

On 15 Oct, 23:19, Raffael Cavallaro vous-plait-mac.com> wrote:

Thankyou - very interesting. Yes, the authors are right that the
Turing machine model is not sufficient for modelling computational
processes. Some of their examples are a bit dodgy and I think that
the main point may have been lost (or buried somewhere). As is usual
the authors spend too long coming to the point.