> From: John Fox > > Dear Gregory, Paul, and Jan, > > I recall proposing something like this (that is, a classification of > available functions) some time ago, but it never got off the > ground. The > advantage of using keywords is that package authors would > classify their > own functions, but I don't think that the current set of keywords is > adequate. It would be particularly useful to work out a > hierarchical or > perhaps hyper-linked classification (without restricting particular > functions to just one terminal node).
I guess something similar to GAMS would help a lot: http://gams.nist.gov/ Personally I think "refactoring" of packages is too difficult to manage, but cross-indexing of functions in an easily searchable fashion would go a long way. Best, Andy > Regards, > John ______________________________________________ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-devel
