ns@cat.cmu.edu (Nicholas Spies) (06/13/88)
Time was, in HyperCard 1.0.1 and 1.1, that you could pass a function as the "with" parameter to sort, as well as chunk expressions. It seems that HyperCard 1.2 doesn't like this, and instead puts up and error dialog saying something to the effect that "xx is an invalid expression on every card". The value of passing a function to sort is that you used to be able to perform parsing and concatenation of strings used as the sort keys. For instance, if you want to sort a stack of names by last name, and yet want the first and optional second and third names to be significant in the sort, and at the same time ignore any instances of "Jr", chunk expressions just won't suffice--you need to write a function to message the names into the desired order before they are looked at by sort, which sorts them. Although using a function as the sort parameter was not documented, it worked fine in 1.0.1 and 1.1, and it is a consistant generalization of using "&" to concatentate chunk expressions for sort (which is how I found out that it worked in the first place). I can't imagine why this capability was removed in 1.2, and sincerely hope that it is reinstanted, and documented, in future versions. -- Nicholas Spies ns@cat.cmu.edu.arpa Center for Design of Educational Computing Carnegie Mellon University
kurtzman@pollux.usc.edu (Stephen Kurtzman) (06/14/88)
In article <1923@pt.cs.cmu.edu> ns@cat.cmu.edu (Nicholas Spies) writes: >Time was, in HyperCard 1.0.1 and 1.1, that you could pass a function as >the "with" parameter to sort, as well as chunk expressions. It seems that >HyperCard 1.2 doesn't like this, and instead puts up and error dialog saying >something to the effect that "xx is an invalid expression on every card". It would be funny if this has really changed. I recall a Mr. HyperTalk (forgot his name, Dan something-r-other) posting that suggested passing the random number function to "shuffle" the cards in a stack. It would be nice if Dan would answer this one. I made up my mind a while back that I would put off buying any book on hypertalk until I found one that documents this function. My reasoning is that it is an obscure usage of sort and anyone who knows about it possibly knows alot. Of course, if this is no longer true, my heuristic is in error.