| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Vadim Nasardinov <vadimn(at)redhat(dot)com> |
| Cc: | Sebastiaan van Erk <sebster(at)sebster(dot)com>, pgsql-jdbc(at)postgresql(dot)org |
| Subject: | Re: \0 and IllegalArgumentException |
| Date: | 2004-10-29 15:36:38 |
| Message-ID: | 16500.1099064198@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-jdbc |
Vadim Nasardinov <vadimn(at)redhat(dot)com> writes:
> I think you do have a point though when you say that \0 should be
> allowed (and handled properly).
Don't hold your breath.
Moving away from the zero-terminated-string approach would require
protocol changes (IIRC, the v3 protocol still depends on that in places),
a massive rewrite effort to change every datatype's I/O functions, plus
writing our own locale library because we couldn't use strcoll()
anymore. And that's just the impacts I can think of before having
had any coffee ;-)
While the private locale library might eventually happen anyway, the
other two things are still daunting, mainly because they break a ton of
user code not only the backend.
Basically, the bang for the buck doesn't seem to be there...
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Markus Schaber | 2004-10-29 16:57:58 | Re: PGobject overhaul (was Re: tightening up on use of oid |
| Previous Message | Jon Orris | 2004-10-29 14:36:57 | Re: \0 and IllegalArgumentException |