Another updated patch.
* Full GiST support
- Thanks to Alexander Korotkov for sending me a new picksplit
algorithm for my "temporal" project on pgfoundry. I modified it for use
with range types, including a (hopefully) intelligent way of handling
empty and unbounded ranges.
* Quite a few tests added, some cleanup done
* naming issues:
- period -> tsrange ?
- periodtz -> tstzrange ?
- intrange -> int4range
* add int8range
* Documentation improvements
- CREATE TYPE
- Data Types section
* Thom Brown and Tom Lane pointed out that the type inferencing
should be able to promote int4 to numeric for queries like:
select '[18,20]'::numrange @> 19;
* Should the SQL function length(), which relies on polymorphic "-",
be marked immutable, stable, or volatile?
* representation or alignment issues
* parser should be improved to handle spaces and quoting better
* Should btree_gist be pulled into core to make it easier to use
exclusion constraints with range types?
* Typmod (optional)
In response to
pgsql-hackers by date
|Next:||From: Robert Haas||Date: 2011-01-30 22:53:39|
|Subject: Re: Review: Fix snapshot taking inconsistencies|
|Previous:||From: Tom Lane||Date: 2011-01-30 22:47:07|
|Subject: Re: autogenerating error code lists (was Re: [COMMITTERS] pgsql: Add foreign data wrapper error code values for SQL/MED.) |