🚀 go-pugleaf

Over four years of discussion within the Dutch STOP project have led to a
certain amount of consensus about the "CIP vs. BMF" issue. Now,
moving on to the other important discussion, transformation systems:

In article <1992Jul22.145710.17064@cs.ruu.nl> nico@cs.ruu.nl (Nico Verwer) writes:
>[..] BMF is much conciser, and simpler to learn, than CIP. On the
>other hand, there is not yet a system for automatic reasoning in BMF.
Ahem. I know of efforts by Vogt et al (@cs.ruu.nl :-)), by
Chisholm/Backhouse, and by Lindsey. That no system exists to do
everything one could want it to, well, that's no surprise. See
our "Ustopia" report for a list of interesting requirements.

>There is such a system for CIP, although I recall the demonstration of
>it at CWI, where attempts to transform the program `1+1' to the more
>efficient program `2' ended in a core dump, after some 15 minutes of
>trying the thing to work.
I must concur that the CIP *prototype* system (written at TU Munich in
the early 80's, Riethmayer et al.) has not done much for the popularity
of program transformation systems, to say the least. We asked a student
to work with it for a few months, and he wasn't much more enthusiastic
than Nico is. This may be partly due to the *prototype* character of
the system, and to the exotic machines on which it was first brought
up. Most of the early work on program transformation systems was aimed
at showing the feasibility of the approach, and not as much on the
practicability of the systems, I'm afraid.

Our group is currently experimenting with the "real" CIP-S
transformation system, a modern (really) window-based system
developed more recently by TU Munich and Siemens. Our first
experiences are much more positive than those with the prototype
system. We hope to report more extensively on this later.
--
Eerke Boiten (eerke@cs.kun.nl)
Dept. of Informatics (STOP Project)
University of Nijmegen, NL