John Lloyd wrote:
> The applications that can be handled by SQL aren't at all interesting
> for us. I'm talking about serious knowledge base applications involving
> sophisticated query answering, uncertainty, non-monotonic reasoning, and so on
> - just the kind of applications that LP languages are good for. The technology
> for doing this effectively on the Web hardly exists at the moment, but some
> LP researchers have begun to move in the right direction. I think
> we have a real opportunity to find a significant niche here.
I fully agree. I just came back from ICLP'97 where there was
an interesting tutorial by Manuel Hermenegildo on WWW-based
LP applications and a workshop on LP and the internet.
As shown by Manuel's Pillow library for WWW applications,
it is not difficult to connect Prolog programs to the internet.
Moreover, Prolog can also be used as a scripting language
for WWW applications, similarly to Perl but with a higher level.
Consequently, an FLP language can do the same but adds nice
features to LP like polymorphism, higher-order features
or demand-driven evaluation. I think this is the main argument -
FLP is just the better LP (from a SW-engineering point of view).
It is difficult to find killer apps particular for FLP - they
are mainly identical to killer apps for LP, but by adding
functional programming techniques to LP we can avoid a lot of
the "problematic" (impure) features of Prolog.
Hence we should concentrate on current and future LP
applications as suggested by John (by the way, all killer apps
mentioned by Phil for FP are also typical killer apps for LP).
As a consequence for the design of Curry, we should not omit
completely the LP programming style, otherwise it is hard
to convince the LP community.
It was also interesting to hear at ICLP'97 that people at
BIM (a Belgian company trying to sell Prolog systems and applications
and which went bankrupt for several reasons) are now trying
to marketing applications based on Mercury. So, it seems
that types become more accepted in LP, which is good for us.
Best regards,
Michael
Received on Mo Jul 14 1997 - 13:41:00 CEST
This archive was generated by hypermail 2.3.0
: Do Feb 01 2024 - 07:15:05 CET