Re: period data type

From: Michael Glaesemann <grzm(at)seespotcode(dot)net>
To: Warren Turkal <wt(at)penguintechs(dot)org>
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: period data type
Date: 2007-02-05 04:38:42
Message-ID: 4DA5E1C5-1DDC-4091-A532-4D57CAB9BBB4@seespotcode.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers


On Feb 5, 2007, at 12:47 , Warren Turkal wrote:

> Is anyone working on a period data type as described in Dr. Richard
> Snodgrass'
> book _Developing Time-Oriented Database Applications in SQL_[1]? I
> did not
> see a relevant project listed in the TODO. I would like to contribute
> (possible funding and/or coding) the development of a conforming
> implementation of this data type.
>
> [1]http://www.cs.arizona.edu/~rts/tdbbook.pdf

This is an area I'm actively working on. I've worked out an example
using composite types and PL/pgsql, though it's not very performant.
I'm currently looking at making a contrib module, but have been
slowed a bit by my lack of C experience. Regarding conformance, as
there's currently no standard, I'm leary of cleaving too closely to
the Temporal SQL proposal to ward of making something that would be
close but not quite conformant to a future addition to the SQL
standard. Also, I hope to make a more general interval/range/period
constructor that would allow you to make "periods" of other types
(say, integers) as well. I've found Date/Darwen/Lorentzos' Temporal
Data and the Relational Model[1] very helpful.

Interested in possibly working together?

Michael Glaesemann
grzm seespotcode net

[1]http://books.elsevier.com/uk/mk/uk/subindex.asp?isbn=9781558608559

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Markus Schiltknecht 2007-02-05 06:44:21 Re: [PATCHES] Fix "database is ready" race condition
Previous Message Jim Nasby 2007-02-05 04:23:15 Re: Performance penalty of visibility info in indexes?