Ian Chivers writes:
> On Thu, 6 Jun 2002 20:22:29 -0400 Michael Milgram <[log in to unmask]>
> wrote:
[list-directed writes don't always print every significant digit]
> i personally feel that it is a quality issue.
I consider it more of a preference isue than a quality one. It is
not at all clear to me what answer is "better quality here." In my
opinion, list-directed output isn't the thing to use when you are
looking for something specific like every possibly significant
digit. I could easily argue that, in places where I use
list-directed output of reals, printing 17 digits is just an
annpyance. I'd consider it "higher quality" (in the sense of
being closer to what I tend to be interested in seeing) if it
just printed 5 or 6 significant digits even for double precision
reals.
I'd consider it quite high quality (more than one would probably
expect) if it were to do "fuzzy rounding" and conclude that
values differing by only a few bits from "round" values like 0.1
ought to be printed just as 0.1.
Well, I won't actually try to defend the position that my preferences
here represent "best quality". My point is just that this is a matter
of preference rather than of quality. The distinction to me is that
quality is at least somewhat easier to rank as better vs worse,
wherease preferences are just different, one not necessarily being
better or worse than another.
--
Richard Maine | Good judgment comes from experience;
[log in to unmask] | experience comes from bad judgment.
| -- Mark Twain
|