On Thu, Mar 16, 2017 at 10:21 PM, Michael Paquier
<michael.paquier@gmail.com> wrote:
> On Fri, Mar 17, 2017 at 11:17 AM, Robert Haas <robertmhaas@gmail.com> wrote:
>> I understand that the point of renaming pg_clog to pg_xact is that
>> pg_clog contains the dreaded letters l-o-g, which we hypothesize
>> causes DBAs to remove it. (Alternate hypothesis: "So, that's what's
>> clogging my database!")
>>
>> Renaming pg_subtrans to pg_subxact has no such redeeming properties.
>>
>> More, with each of these renamings, we're further separating what
>> things are called in the code (xlog, clog, subtrans) with what they're
>> called in the filesystem (wal, xact, subxact).
>>
>> So if we must rename pg_clog, OK, but can't we leave pg_subtrans
>> alone? It's not hurting anybody.
>
> The only argument behind the renaming of pg_subtrans is really
> consistency with pg_xact, because both deal with transactions. I don't
> personally mind if this portion of the renaming is left off, as you
> say anything labelled with "log" is at the origin of this thread.
Fine! I've committed the pg_clog renaming, but I'd really like to
draw the line here. I'm not going to commit the pg_subtrans ->
pg_subxact naming and am -1 on anyone else doing so. I think that
having the names of things in the code match what shows up in the file
system is an awfully desirable property which we should preserve
insofar as we can do so without compromising other goals. Not
invalidating what users and DBAs think they know about how PostgreSQL
works is a good thing, too; there are a lot more people who know
something about the directory layout than will read this thread.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
