Horton

Re: "serializable" in comments and names

Tom Lane <tgl@sss.pgh.pa.us>

View thread
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Alvaro Herrera <alvherre@commandprompt.com>
Cc: Kevin Grittner <kevin.grittner@wicourts.gov>, Dan Ports <drkp@csail.mit.edu>, Robert Haas <robertmhaas@gmail.com>, "<pgsql-hackers@postgresql.org>" <pgsql-hackers@postgresql.org>
Date: 2010-09-08T17:02:08Z
Lists: pgsql-hackers 
Alvaro Herrera <alvherre@commandprompt.com> writes:
> Excerpts from Tom Lane's message of mi sep 08 12:12:31 -0400 2010:
>> AFAIR it doesn't keep the first snapshot around.  If it did, most of
>> your work on snapshot list trimming would have been useless, no?

> That's my point precisely.  The name "IsolationUsesXactSnapshot" makes
> it sound like it applies to any transaction that uses snapshots for
> isolation, doesn't it?

I don't think so, at least not when compared to the alternative 
IsolationUsesStmtSnapshot.

> How about IsolationUses1stXactSnapshot

This just seems longer, not really better.  In particular, we have
*always* adhered to the phraseology that a "transaction snapshot"
is the first one taken in a transaction, so I don't see exactly
why it's confusing you now.

			regards, tom lane