2006/6/6, Larry Garfield <larry@xxxxxxxxxxxxxxxx>:
On Monday 05 June 2006 21:12, Martin Alterisio wrote: > > As for the increment, it actually would never have occurred to me to ++ a > > string before this thread, honestly. :-) However, what it appears to be > > doing (and I'm sure Rasmus will correct me if I'm wrong) is using a > > "string > > base" instead of a numeric base. Thus a++ = b, b++=c, etc. z++ "rolls > > over" > > to the next "digit" (which because it's a string goes to the right rather > > than the left), and resets. So just as 9++ rolls over to 10, z rolls > > over to > > aa. > > > > Does that make more sense? > > You misunderstood me, I completely understand how the operators function, > but you're saying it makes sense the way their functionality is assigned, > what I want to know is the reasons you have that support those > affirmations. I completely understand that string comparison is done > alphabetically, but how does having the functionality for the ++ operator > create a sequence that are inconsistent with the comparison operator, makes > sense? Because defining ++ and < and > in such a way as to make them "behave like numbers" would have made them not work for alphabetizing. A string is a string, and comparison of strings is alphabetic (for some definition of alphabet). It's more useful to deal with strings as strings than to make them quack like numbers.
Then, if it's not a math operation, why use a math operator for such functionality? In which way is the ++ operator that generates a string sequence, useful enough to justify the formal inconsistency between the math operators? I still don't see the advantages of having the ++ recognize the string as a sequence, and generate the next item in the sequence. I believe those decisions should be left to the coder, because he knows what the string really represents and which kind of sequence is being used.
