| From: | Thomas Lockhart <lockhart(at)fourpalms(dot)org> |
|---|---|
| To: | Peter Eisentraut <peter_e(at)gmx(dot)net>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | "Ross J(dot) Reedstrom" <reedstrm(at)rice(dot)edu>, Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us>, Thomas Lockhart <lockhart(at)alumni(dot)caltech(dot)edu>, pgsql-hackers(at)postgreSQL(dot)org |
| Subject: | Re: Re: [GENERAL] +/- Inf for float8's |
| Date: | 2001-06-05 03:07:04 |
| Message-ID: | 3B1C4CD8.4301C97C@fourpalms.org |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
> > -infinity < normal values < infinity < NaN < NULL
> I was thinking about making NaN equivalent to NULL. That would give
> consistency in ordering, and probably also in arithmetic. Additionally,
> if the platform supports it we ought to make the Invalid Operation FP
> exception (which yields NaN) configurable: either get NULL or get an
> error.
I'd like to see the distinction between NaN and NULL retained, since the
two "values" arise from different circumstances and under different
conditions. If a particular app needs them to be equivalent, then that
is easy enough to do with SQL or triggers.
- Thomas
On a modestly related note, I'm come over to the notion that the
date/time value 'current' could be ripped out eventually. Tom, isn't
that the only case for those types which bolluxes up caching of
date/time types?
| From | Date | Subject | |
|---|---|---|---|
| Next Message | chris.bitmead | 2001-06-05 03:08:58 | Re: Question about inheritance |
| Previous Message | bpalmer | 2001-06-05 02:24:51 | Re: Mirrors not tracking main ftp site? |