> But HyperTalk is not C, was not designed for > efficiency, and so got some stuff right.
You can say that again! :) > There is no reason for boolean, integer and/or > character types to be randomly interchangeable Not "randomly", that's for sure! > They are conceptually distinct. Yes, but as you point below, HyperCard automatically translates from one data-type to another when the code being interpreted calls for it. OTOH, enforcing this conceptual distinction by making the scriptor explicitly cast the data as a certain type, like AppleScript and many others do, would be an absolutely grievous injury to the simplicity and forgiveness of HyperTalk. > Why do you think HyperTalk had numtochar > and chartonum? To convert characters to-and-from ASCII. Come to think of it though, perhaps we could make our interpreter automatically cast between a char and its ASCII equivalent (and vice-versa), if-and-when the execution of the script calls for it. > Implicit conversion between strings and other types > is a fundamental part of HyperTalk... You betcha! > ... perhaps unfortunate, perhaps not ... More than fortunate. A God-send really. Explicit casting between data-types is of one the most frustating aspects of scripting with AppleScript. The only downside to auto casting is that it uses up more RAM and slows execution down a little bit. But nowadays, blazzingly-fast RAM and clock-speeds make this downside moot. > but there is no reason for the other types > to get confused... I would accept any change if-and-when there is a solid logic behind it, particularly if it makes things easier and more-forgiving for the scriptors, which in turn will make their wares less buggy and less complex. Obstinately yours, Alain __________________________________________________ Do You Yahoo!? Send your FREE holiday greetings online! http://greetings.yahoo.com _______________________________________________ Freecard-general mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freecard-general
