[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]
formic enhancements revisited
- To: <acad!blitzen!rick@xxxxxxxxxxx>
- Subject: formic enhancements revisited
- From: Mark S. Miller <vlad!mark>
- Date: Thu, 28 Jun 90 15:30:18 PDT
- Cc: <xtech@xxxxxxxxxxx>
- In-reply-to: <Rick>,23 PDT <9006281021.AA01271@xxxxxxxxxxxx>
Date: Thu, 28 Jun 90 03:21:23 PDT
From: xanadu!acad!blitzen!rick (Rick Mascitti)
...
Are all of the file-scope functions in febe/*x.hxx and xlatexpp/*x.hxx
pseudoconstructors? Looks that way to me. If this is (and will remain)
the case, then we may not need the type constraint.
It seems to me that both Ceran and Scheran (and eventually CeRand, the
one for Objective-C) simply wants to wrap all external functions
defined at file scope in *x.hxx, whether or not one would consider
them pseudo-constructors. Part of the problem is that the notion of
"pseudo-constructor" is largely in the eye of the beholder.
Good news, though: the $(IF CLASS ! | name: SP2_+ name: CP2_+) stuff
works as advertised. Many thanks to roland, just saved me a bunch
of culling. Modulo pc's and overloads, Scheran is fully automated.
Lots of fun cases for void and IntegerVar valued functions, IntegerVar
and char * args, and some other OBTWs.
Congratulations! Great stuff!
This means we have 4 languages already (Smalltalk, C++, C, and Scheme)
driven off a single source. Language independent programming anyone?