On Nov 14, Shiro Kawai wrote:
> From: Eli Barzilay <eli_at_barzilay.org>
> Subject: Re: Re: [r6rs-discuss] [Formal] "Byte-vector" would be a better name than "bytes" for a data type.
> Date: Tue, 14 Nov 2006 14:46:27 -0500
>
> > sub1, still.
> >
> > IMO:
> >
> > * `bytes' are closer in use to strings than vector.
> >
> > * so `byte-string' is a better name, but then `character-' is implicit
> > in one but not the other. Similar asymmetry to `length' vs
> > `vector-length'.
> >
> > * so a better choice would be `character-string' and `byte-string',
> > but that's inconvenient.
> >
> > * so `bytes' is a good compromise. (I use "byte strings" in text
> > instead of "byteses".)
>
> A bit off-topic...
>
> Hmm.. byte-vector strongly suggests it's an array of bytes, implying
> indexed access is O(1). I think it matches the practical use of
> byte vectors (such as dealing with contiguous region of memory).
(Good point. But before the unicode revolution, "string of
characters" had these features too, so that's not why they didn't end
up being "character-vector"s.)
> On the other hand, a string is a sequence of characters, but
> it doesn't need to be an array of characters---it may
> internally be a list of characters, or a balanced tree, or
> something else.
>
> So byte<->vector and character<->string aren't parallel, IMHO.
>
> To return to on-topic, I'm +1 to byte-vector.
--
((lambda (x) (x x)) (lambda (x) (x x))) Eli Barzilay:
http://www.barzilay.org/ Maze is Life!
Received on Tue Nov 14 2006 - 15:07:47 UTC