On 11/02/2011 23:55, Andrei Alexandrescu wrote:
On 2/11/11 7:07 AM, foobar wrote:
Andrei Alexandrescu Wrote:
I don't find the name "iota" stupid.
Andrei
I want to make a few comments, arising from several different posts in
this discussion.
First, before this discussion that Ary started, I had heard the "iota"
function being mentioned a few times in other NG posts, and although I
wasn't 100% sure of what it did, I did deduce it returned a sequence of
numbers (in range API).
And also I assumed this number sequence was related to some concept in
mathematics called iota (like a Fibonacci sequence, or the factors in a
Taylor series, etc.). I found now that such is not the case, the term
iota as used in D was introduced by the APL language, but is not defined
or well-known in mathematics academia.
Generally speaking I agree with with Ary's idea of having better, more
descriptive names. But I am not sure I agree what would be a better
alternative.
I find the name "iota" to be meaningless, but "range" for example is
worse because its meaning conflicts with the general meaning of D
"ranges". Conflicts in the sense that although iota(...) does return a
range, it does something quite more specific. I would rather introduce a
new term ("iota"), than use a confusing or overlapping one like "range"
(or even "interval").
I might prefer the function to be called "numberSequence" or
"numberSeq", but I don't yet have a strong enough opinion to actually
cast a vote.
--
Bruno Medeiros - Software Engineer
