When is this faster than the normal routine?
Date: 13 Apr 2009 22:59
Number of posts: 4
RSS: New posts
Is that rhetorical?
I think the main point of this routine variation was to use less variables (and therefore less cleanup).
The problem is that I don't think that's a good reason for the routine to exist (especially considering the other routine, with cleanup, is still smaller). I kind of buy the "easier to understand" reason, if the routine is to be considered a learning tool, rather than something to be put in a program — which is, of course, the more important part.
A good reason is that the alternate routine is faster for integers with 6 or less digits (which one might be able to improve on, with optimization). This means there are situations in which this routine would be better, which in turn lets people actually make an informed choice about which routine to use.
On the other hand, if the alternate list-to-frequency routine is faster and smaller (and the first doesn't appear to have any useful properties the second doesn't), then why is the first one still there?
I would say that the first list-to-frequency routine is still there because you hold a respectable position in the TI community. You made that routine, so people feel obliged to not question it.
For that same reason is it likely that burr's alternate number-to-string routine is still existent.
Having thought all of that, I reckon that if there are better routines to be shown, then why show the lesser ones? Sentimental is not a good reason.
By all means go ahead and delete the unnecessary pages.