On Monday, June 27, 2005, at 02:30 pm, Robert Strandh wrote:
Actually, I think it is time to give McCLIM a major overhaul, though I am not convinced anybody has the time.
I agree with this; at the very least, I think it needs a thorough review, and notes taken where the implementation disagrees with, or extends, the spec. I don't know how many annotations there are on the annotatable spec but I'm sure plenty of things didn't make it onto that document (it was out of order for a while).
Such an overhaul would start by correcting and supplementing the specification where it is wrong and/or incomplete. We might call the document obtained this way the CLIM 2.2 specification. It would be much more precise than the CLIM 2.0 specification, and we would make sure as much as possible that it does not have any contradictions.
Don't both Lispworks and Franz already have 'CLIM 2.2'? Neither of which agree by the way (either with each other, or in several places, with the spec, for those methods (the vast majority) that are listed in the spec). Do we need to take the vendor CLIMs into account at all? I'm currently putting together a document comparing the methods in the spec, the different user guides, and as implemented in McCLIM but it's slow going. Perhaps it will be useful one day. -Duncan
From the CLIM 2.2 specification, chunks of the code base could be refactored, checked, and selectively rewritten.
All of this, of course, requires more manpower, which we do not have.
-- Robert Strandh
--------------------------------------------------------------------- Greenspun's Tenth Rule of Programming: any sufficiently complicated C or Fortran program contains an ad hoc informally-specified bug-ridden slow implementation of half of Common Lisp. --------------------------------------------------------------------- _______________________________________________ mcclim-devel mailing list mcclim-devel@common-lisp.net http://common-lisp.net/cgi-bin/mailman/listinfo/mcclim-devel