From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Peter Eisentraut <peter_e(at)gmx(dot)net> |
Cc: | pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: [PATCH] Revive line type |
Date: | 2013-06-15 15:02:02 |
Message-ID: | 6899.1371308522@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Peter Eisentraut <peter_e(at)gmx(dot)net> writes:
> Complete the implementations of line_in, line_out, line_recv,
> line_send. Remove comments and error messages about the line type not
> being implemented. Add regression tests for existing line operators
> and functions.
> ---
> This just revives existing functionality, doesn't add anything new.
> One thing that the original code did not settle was how to convert a
> line in form Ax+By+C=0 to the two-points output form. Obviously, you
> can just pick to random points on the line, but I wonder whether there
> is a more standard solution.
ISTM printing a line as two points is an unfortunate representational
choice altogether. If the internal form is Ax+By+C=0, shouldn't the
text form expose A,B,C directly? What you've got here necessarily
suffers a lot of roundoff error during I/O conversions, and so it seems
pretty likely to fail dump-and-reload tests.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Andres Freund | 2013-06-15 15:08:34 | Re: stray SIGALRM |
Previous Message | Sawada Masahiko | 2013-06-15 14:59:01 | Re: Patch for fail-back without fresh backup |