| From: | Thomas Lockhart <lockhart(at)alumni(dot)caltech(dot)edu> |
|---|---|
| To: | Don Baccus <dhogaza(at)pacifier(dot)com> |
| Cc: | pgsql-hackers(at)postgreSQL(dot)org |
| Subject: | Re: [HACKERS] minor bug... |
| Date: | 2000-02-10 21:45:53 |
| Message-ID: | 38A33191.AE082DC8@alumni.caltech.edu |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
> I assume not having a type added to hash ops isn't fatal, because
> "numeric" isn't there and Jan strikes me as being a very thorough
> guy...
A hash index is probably even less useful than the btree index for
this type, unless it can be used with multi-column indices. Because
the hash will chain duplicate values into a list of some kind, and
you'll get *long* lists.
Find and steal the code for "char" (the real one-byte character type).
But a one-bit hash is what you really want, so it may be better to
implement your own.
- Thomas
--
Thomas Lockhart lockhart(at)alumni(dot)caltech(dot)edu
South Pasadena, California
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Thomas Lockhart | 2000-02-10 21:51:41 | Re: [HACKERS] createdb default arguments |
| Previous Message | Ross J. Reedstrom | 2000-02-10 21:03:35 | Re: [HACKERS] make_ctags script |