Re: Optional arguments, assertions, type annotations

[Noel Welsh <noelwelsh@yahoo.com>]

My take on this is that there are several different
concepts here.

I started developing a package called Schemedoc
(similar to Javadoc) with similar goals (e.g. insert
pre- and post-conditions and provide documentation)
got  some good responses from the list.  You might
want to do some searches in the archives to read them.

As far as pre- and post-conditions go, I believe they
can be subsumed by a sufficiently expressive type
system (where sufficiently expressive means solving
AI).  There is work in type systems for XML (XDUCE)
and dependent types (Cayenne) which expand what can be
typed in interesting ways.  Shriram mentioned type
systems for regexps not too long ago.

That the future.  For the now, I think specifying pre-
and post- conditions separately from the main code is
cleaner than weaving them into the code.  Something
like aspect-orientated programming could be used to
weave them together as required.

HTH,
Noel

> > From: "Thomas Link" <t.link.tmp200101@gmx.net>
> > Date: Tue, 18 Sep 2001 18:13:56 +0200
> > 
> > Hello,
> > 
> > May I ask which approach you favour with respect
> to optional arguments, 
> > assertions (pre-conditions, post-conditions), or
> type annotations (for 
> > security reasons also)? As far as I know there is
> no standard solution 
> > for these issues, which is why different
> distributions have chosen 
> > different roads. As I recently implemented my own
> rather simple 
> > package, which definitly isn't the answer to the
> ultimate question, I 
> > would be interested which solution you prefer. Or
> is there some reason 
> > for why this hasn't been standardized yet?
> > 
> > Cheers,
> > Thomas.


__________________________________________________
Terrorist Attacks on U.S. - How can you help?
Donate cash, emergency relief information
http://dailynews.yahoo.com/fc/US/Emergency_Information/