rkr@g.g.oswego.edu (Rajendra K Raj) (10/30/90)
I'd appreciate hearing from people who have used Modula-3 (as opposed to
Modula-2) as a teaching language, either for introductory programming or
for more advanced programming.
Btw, what happened to the call for discussion for creating a new
newsgroup comp.lang.modula3?
- R. K. Raj
rkr@g.oswego.edu
Department of Computer Science
State University of New York
Oswego, NY 13126muller@src.dec.com (Eric Muller) (10/31/90)
In article <RKR.90Oct30082422@g.g.oswego.edu>, rkr@g.g.oswego.edu (Rajendra K Raj) writes: > Btw, what happened to the call for discussion for creating a new > newsgroup comp.lang.modula3? The discussion (on news.groups) is rather quiet. The voting will start this friday. -- Eric Muller. ------------------------------------------------------------------------------ System Research Center - 130 Lytton Av. - Palo Alto, CA 94301 - (415) 853 2193
pr@cl.cam.ac.uk (Peter Robinson) (11/05/90)
[Apologies if this is a duplicate, but my original posting seems to have
had a spell cast on it by the satanists.]
I am currently teaching a course in Modula-3 to Computer Scientists in the
second year of the three-year undergraduate course at the University of
Cambridge in England.
We teach ML as an introductory language at the beginning of the freshman
year and, after brief historical excursions into Fortran and COBOL,
introduce imperative programming with Modula-3 at the beginning of the
second year. We intend to bring this forward to the end of the first year
in future. The programming strand of the course subsequently moves on to
C, Prolog, LISP and MIPS assembly code.
The course is delivered as 16 one-hour lectures and five practical sessions
each taking one to two hours. That constitutes about one fifteenth of the
year's work. Assessment is by two closed-book exam questions at the end of
the year and a practical exercise (involving writing about 100 lines of
code) undertaken during the year. It is envisaged that Modula-3 will also
become the language of choice for many of the projects undertaken by
students in their final year.
The syllabus covers most of the language and tries to motivate the design
and discuss possible developments, which means that it proceeds at a fairly
brisk pace. I could supply the rough breakdown of material to lectures if
anyone is interested. The practical sessions work through a sequence of
elementary programming problems starting with 'Hello world' and ending with
prime number generation using a collection of concurrent sieves.
We use a collection of 50 DECstation 3100s running Ultrix and the DEC SRC
Modula-3 system for practical work. The workstations each have 12 Mbyte of
RAM and a small local disc that is used for swapping and temporary files.
Most system files are delivered by NFS from a pair of DECserver 5400s and
other files from a variety of NFS hosts around the Department. The
workstations and two main servers are on a private ethernet which is
bridged to the rest of our communications network. The performance is
modest - it can take two minutes to compile and link a 'Hello world'
program. This is not a reflection on the SRC system - the equivalent
operation takes about 6 seconds on an idle server machine with local discs.
We have uncovered a few bugs in the system, which is not altogether
surprising given that we have launched 60 people with wildly different and
idiosyncratic programming styles at it. However, DEC SRC have been
extraordinarily helpful throughout.
I have enjoyed teaching the course; Modula-3 is certainly more fun than
Modula-2 or Pascal which were its predecessors in this role. The students
seem to have enjoyed it too.
In case you were wondering, Mick Jordan of DEC SRC and I are hoping to
deliver a text book based on the course some time next year.
- Peter Robinson.
University of Cambridge Telephone: +44 223 334637
Computer Laboratory Facsimile: +44 223 334678
New Museums Site Telex: 81240 (CAMSPL-G)
Pembroke Street
Cambridge E-mail: pr@cl.cam.ac.uk
England CB2 3QG pr:ComputerLab:CambridgeUniv