From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | kislo(at)athenium(dot)com |
Cc: | Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us>, Joel Burton <jburton(at)scw(dot)org>, igorr(at)ifi(dot)uio(dot)no, pgsql-general(at)postgresql(dot)org |
Subject: | Re: Unanswered questions about Postgre |
Date: | 2000-12-12 01:43:54 |
Message-ID: | 14366.976585434@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Joe Kislo <postgre(at)athenium(dot)com> writes:
> ... this lack of BLOB support. I understand that the C
> API can read/write -files- off the server's filesystem and load them
> into the database. Unfortunately we would absolutely require true
> over-the-wire blob support through JDBC. AFAIK, even with these "toast"
> columns, it still wouldn't fill that need.
This is a misunderstanding. You can still use the old-style large
objects (in fact 7.1 has an improved implementation of them too),
and there's always been support for either over-the-wire or
server-filesystem read and write of large objects. In fact the former
is the preferred way; the latter is deprecated because of security
issues. In a standard installation you can't do the server-filesystem
bit at all unless you are superuser.
The JDBC support for over-the-wire access to large objects used to
have some bugs, but AFAIK those are cleaned up in current sources
(right Peter?)
Adding a similar feature for TOAST columns will certainly be a
notational improvement, but it won't add any fundamental capability
that isn't there already.
> 2) Postgre does not record rollback segments.
We know this is needed. But it will not happen for 7.1, and there's
no point in complaining about that; 7.1 is overdue already.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Sandeep Joshi | 2000-12-12 01:51:28 | Re: Unanswered questions about Postgre |
Previous Message | Tim Kientzle | 2000-12-12 01:10:22 | Re: Unanswered questions about Postgre |