saj@iuvax.UUCP (02/15/84)
#N:iuvax:8600005:000:874 iuvax!cjl Dec 8 17:31:00 1983 Many of you annoyed by the ambiguous and imprecise definition of Modula-2 as I did may find it useful to reference the ISO Pascal stan- dard. Many problems in Modula-2 are really inherited from the origi- nal Pascal definition. ISO Pascal standard clarifies many of them. For example, the actual CBR parameters can be any left value expressions (address calculation) including nested array, record component access and pointer dereference. In Modula-2, you will find pointer dereference is not documented. Just like its predecessor Pascal, I hope the Wirth's style of language definition won't hurt the popularity of Modula-2 in the future. Judging as a whole, Modula-2 is still the best one we can find to replace Pascal programming which is old now. C.Jim Lo, Dept. of C and I Sci., IU-PU at Indpls cjl@Indiana@UDEL-Relay
saj@iuvax.UUCP (02/15/84)
#R:iuvax:8600005:iuvax:8600006:000:600 iuvax!cjl Jan 30 21:24:00 1984 It seems a good idea to consult the Ada specifications for the definition of these new features which don't exist in ISO Pascal specification than to consult the existent Modula-2 compilers. For example, When a module is imported the runtime semantics "require" its initialization part be completed before we can proceed further. That automatically precludes the loops in the import-export relation graph. My VAX-VMS Modula-2 compiler from Hamburger doesn't check this kind of error. I cannot find clear definition in the Modula-2 manual neither. CJL, Dept. of CSCI, IUPUI cjl@Indiana@UDEL-Relay
ags@pucc-i (Seaman) (02/16/84)
My Modula-2 manual (Volition Systems) states that cyclic dependencies in IMPORT-EXPORT graphs are permissable so long as the initialization portions do not depend on each other, since they may be executed in arbitrary order. -- Dave Seaman ..!pur-ee!pucc-i:ags "Against people who give vent to their loquacity by extraneous bombastic circumlocution."
ags%pucc-i@jett.UUCP (Seaman) (02/16/84)
References: <122@iuvax.UUCP> Relay-Version: version B 2.10.1 6/24/83; site akgua.UUCP Posting-Version: version B 2.10.1 6/24/83; site pucc-i Path: akgua!clyde!burl!ulysses!mhuxl!ihnp4!inuxc!pur-ee!CS-Mordred!Pucc-H:Pucc-I:ags Message-ID: <207@pucc-i> Date: Thu, 16-Feb-84 14:12:06 EST Date-Received: Fri, 17-Feb-84 09:25:34 EST Organization: Purdue University Computing Center My Modula-2 manual (Volition Systems) states that cyclic dependencies in IMPORT-EXPORT graphs are permissable so long as the initialization portions do not depend on each other, since they may be executed in arbitrary order. -- Dave Seaman ..!pur-ee!pucc-i:ags "Against people who give vent to their loquacity by extraneous bombastic circumlocution."
phipps@fortune.UUCP (Clay Phipps) (02/18/84)
I wish that the Pascal Standard or Wirth's *Modula-2* book had adopted the declaration concepts of "introducing", "establishing", "elaborating", "constraining", "creating", and "initializing", as used in section 3.2, "Object and Number Declarations", in *Proposed Standard Reference Manual for the Ada Programming Language*, July 1980, the predecessor to MIL-STD-1815A (the ANSI Ada standard). These concepts, although somewhat arbitrarily named, are useful in describing the effects of the declaration process. They allow people to use reasonably well defined words, instead of hand-waving, to discuss issues related to declarations, scope, and regions (ANSI/ISO Pascal term). They would have made it straightforward to describe the way scoping works for identifiers in ANSI/ISO Pascal (which is different from how all but 2 Pascal compilers I know of implement it). -- Clay Phipps -- {allegra,amd70,cbosgd,dsd,floyd,harpo,hpda,ihnp4, megatest,nsc,oliveb,sri-unix,twg,varian,VisiA,wdl1} !fortune!phipps