From: | Philip Warner <pjw(at)rhyme(dot)com(dot)au> |
---|---|
To: | lockhart(at)fourpalms(dot)org, Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> |
Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Re: Patch application |
Date: | 2001-03-20 06:09:29 |
Message-ID: | 3.0.5.32.20010320170929.02059e20@mail.rhyme.com.au |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers pgsql-jdbc pgsql-odbc |
At 05:36 20/03/01 +0000, Thomas Lockhart wrote:
>
>Unless there is a "process improvement" which comes with developing this
>list, I don't really see what the benefit will be wrt existance of the
>list, developing the list, of deciding who should be on a list, etc etc.
>
Totally agree; such formality will be barrier and we will gain almost
nothing. It will also further disenfranchise the wider developer community.
ISTM that the motivation is based on one or two isolated incidents where
patches were applied when they should not have been. Tom's suggestion of
insisting on CVS headers will deal with the specific case in point, at far
lesser social and labour overhead.
The last thing we want to do to people who contribute heavily to the
project is say 'Gee, thanks. And you are now responsible for all approvals
on this area of code', especially in late beta, when they are likely to be
quite busy. Similarly, when we are outside the beta phase, we don't need
the process.
I suspect that anybody who has worked on a chunk of code in a release cycle
will keep an eye on what is happening to the code.
My suggestion for handling this process would be to allow an opt-in 'watch'
to be placed on files/modules in CVS (CVS supports this, from memory).
Then, eg, I can say 'send me info when someone makes a match to pg_dump'.
Similarly, when I start working on a module, I can add myself to the list
to be informed when someone changes it underneath me. This would be
genuinely useful to me as a part-time developer.
As a further development from this, it would be good to see a way for bug
reports to be redirected to 'subsystems' (or something like that), which
developers could opt into. So, using myself as an example, I would like
special notification when a pg_dump bug is reported. Similarly, I might
like to be notified when changes are made to the WAL code...
If you want to make cute web pages, and define domain experts, make it
useful, make it opt-in, and make it dynamic.
----------------------------------------------------------------
Philip Warner | __---_____
Albatross Consulting Pty. Ltd. |----/ - \
(A.B.N. 75 008 659 498) | /(@) ______---_
Tel: (+61) 0500 83 82 81 | _________ \
Fax: (+61) 0500 83 82 82 | ___________ |
Http://www.rhyme.com.au | / \|
| --________--
PGP key available upon request, | /
and from pgp5.ai.mit.edu:11371 |/
From | Date | Subject | |
---|---|---|---|
Next Message | Manal S | 2001-03-20 07:57:22 | Question |
Previous Message | Thomas Lockhart | 2001-03-20 06:01:19 | Re: More on elog and error codes |
From | Date | Subject | |
---|---|---|---|
Next Message | Juhan-Peep Ernits | 2001-03-20 09:25:55 | Trouble with JDBC2 ResultSet.getDate() |
Previous Message | Bruce Momjian | 2001-03-20 05:54:10 | Re: [Patch] Implementation of DatabaseMetaData.getIndexInfo() |
From | Date | Subject | |
---|---|---|---|
Next Message | DI Hasenöhrl | 2001-03-20 10:17:19 | floatingpoint value and MsAccess |
Previous Message | Thomas Lockhart | 2001-03-20 05:36:04 | Re: Patch application |