chase@Ozona.orc.olivetti.com (David Chase) (02/07/90)
Archive-name: modula-3-info/29-Jan-90 Original-posting-by: chase@Ozona.orc.olivetti.com (David Chase) Original-subject: Re: Modula-3 (was: Modula-2 compilers/environments for UNIX) Archive-site: orc.olivetti.com [129.189.192.20] Archive-directory: pub/modula-3-info Reposted-by: emv@math.lsa.umich.edu (Edward Vielmetti) In article <1990Jan29.164508.15205@cs.rochester.edu> ken@cs.rochester.edu writes: >Just curious. Is the DEC SRC Modula-3 that was advertised as available >for ftp from gatekeeper.dec.com a superset of Modula-2, i.e. can one >use that to compile M2 programs? Afraid not, but the differences aren't extreme. I think you'd probably find the lack of variant records and the change in CARDINAL most vexing (CARDINAL is now a subtype of INTEGER -- if you want to get at all 32 bits, you have to use the Word interface). Another change is that WITH expr DO ... END has been replaced with WITH id = expr {, id2 = expr2} DO ... END The precedence of the expression operators has been changed, also, but it is unlikely that change will break many programs (that was one argument in its favor). There's restrictions on what things can be OPAQUE types -- no more assuming that an integer and a pointer are the same size. Of course, there's lots of gunk that's been added, but that's not tremendously important if you're just porting old code. If you have a postscript printer capable of handling large files (apparently an Apple Laserwriter is not normally up to the task) you can print your very own copy of the report (71 pages). The compressed PostScript is 110Kb, and can be obtained via anonymous ftp to orc.olivetti.com [129.189.192.20] from the directory ~ftp/pub/modula-3-info. You can also lay your hands on a copy of the (Olivetti) license in the same directory, or write to Modula-3 Olivetti Research California, STL 2882 Sand Hill Road Menlo Park, CA 94025 for more information. David