TODO item: remove postmaster -o option
Hi,
attached is a patch against HEAD that removes the postmaster's -o
option. It passes make check on i386 Ubuntu 5.10. I sent an e-mail
earlier today asking about the status of this TODO item to -hackers.
No answers so far :)
Markus Bertheau
Attachments:
pg-remove-postmaster--o.patchtext/x-patch; charset=us-ascii; name=pg-remove-postmaster--o.patchDownload+5-50
I not gotten to reply to your earlier email. Removal of -o requires
that all the postgres flags be supported by the postmaster. I think
Peter Eisentraut is working on a patch to merge the two binaries and the
conflicting options.
---------------------------------------------------------------------------
Markus Bertheau wrote:
Hi,
attached is a patch against HEAD that removes the postmaster's -o
option. It passes make check on i386 Ubuntu 5.10. I sent an e-mail
earlier today asking about the status of this TODO item to -hackers.
No answers so far :)Markus Bertheau
[ Attachment, skipping... ]
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
--
Bruce Momjian http://candle.pha.pa.us
SRA OSS, Inc. http://www.sraoss.com
+ If your life is a hard drive, Christ can be your backup. +
[Apologies to Bruce for the double send]
2006/3/5, Bruce Momjian <pgman@candle.pha.pa.us>:
I not gotten to reply to your earlier email. Removal of -o requires
that all the postgres flags be supported by the postmaster. I think
Peter Eisentraut is working on a patch to merge the two binaries and the
conflicting options.
The docs say
"The use of this option is obsolete; all command-line options
for server processes can be specified directly on the
postmaster command line"
Markus Bertheau
"Markus Bertheau" <mbertheau.pg@googlemail.com> writes:
attached is a patch against HEAD that removes the postmaster's -o
option.
Removing the option without providing equivalent substitute
functionality is not what the TODO item has in mind :-(
Peter is working on something to clean up the problem properly:
http://archives.postgresql.org/pgsql-patches/2006-01/msg00239.php
regards, tom lane
Markus Bertheau wrote:
[Apologies to Bruce for the double send]
2006/3/5, Bruce Momjian <pgman@candle.pha.pa.us>:
I not gotten to reply to your earlier email. Removal of -o requires
that all the postgres flags be supported by the postmaster. I think
Peter Eisentraut is working on a patch to merge the two binaries and the
conflicting options.The docs say
"The use of this option is obsolete; all command-line options
for server processes can be specified directly on the
postmaster command line"
Uh, is this true? Peter?
--
Bruce Momjian http://candle.pha.pa.us
SRA OSS, Inc. http://www.sraoss.com
+ If your life is a hard drive, Christ can be your backup. +
Markus Bertheau wrote:
The docs say
"The use of this option is obsolete; all command-line options
for server processes can be specified directly on the
postmaster command line"
First, note the use of "obsolete" -- there are better ways to achieve
what you want -- as opposed to "deprecated" -- the use is recommended
against (for some reason to be specified). If we make a plan to remove
the -o option sometime, we should use the latter and announce our
intentions. That needs to be discussed, though.
The above documentation is only in the development branch, so deriving a
justification for removing that option from that is premature anyway.
--
Peter Eisentraut
http://developer.postgresql.org/~petere/