On 22 May 2012 22:17, Bruce Momjian <bruce@momjian.us> wrote:
> Remove reviewers from 9.2 release notes; improve attributions.
I am unhappy with now being listed as a secondary author of
pg_stat_statements normalisation. I am credited as a co-author within
the docs (alongside the original 8.4 author, Takahiro Itagaki), and am
solely credited within the original commit message. Tom did of course
clean up the patch, but most of my earlier design decisions, many of
which went against the grain until very late in the last commitfest,
remain.
These include:
* The whole decision as to what type of tree to hash. I was strongly
urged to first consider hashing plan trees, and then later the raw
parse tree. I more-or-less stood by my earliest decision to hash a
serialisation of the Query struct (though switched to pre rather than
post rewrite stage hashing).
* The jumble concept (central to how the feature works), macro
infrastructure, etc.
* The decision to not use the query_tree_walker infrastructure when
walking trees.
* The whole idea of using the low-level scanner API to calculate the
lengths of constants, which were not directly available from the core
system.
* The approach to managing sticky-entry decay, and indeed the very
idea of sticky entries.
--
Peter Geoghegan http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training and Services
