Emma McGrattan reported on the preceding code sprint some details of which are here http://blogs.planetingres.org/notnull/2009/06/10/ingres-code-sprint-09/ and Bill Maimone gave a glimpse of some of the projects that are currently in development both internally and in conjunction with the community.  His live demonstration was hampered by Windows but he showed a new table structure type which running on a laptop out performed MS SQL Server running on a large server environment.  Other developments include MVCC and long object names, more available at http://community.ingres.com/forum/blogs/stephenb/4-whats-happening-dbms-group.html.

In a amended plenary agenda, Rick van der Lans was presented with an award for ‘Contribution to the Ingres Community’ and was available to autograph copies of the newest addition to the Ingres bookshelf for over 20 years http://www.lulu.com/content/paperback-book/the-sql-guide-to-ingres/6608333.

The afternoon breakout agenda contained more presentations than one person could attend.  I chose !EasyIngres, Multiple Simultaneous Assignment, Ingres 9.2 features and Ingres 9.3 and roadmap.

!EasyIngres a simple pre-configured bundle of the components required to start developing with PHP on Windows.  A community development mentored by Bruno Bompar; the project is starting to look at a similar Linux bundle.

Multiple Simultaneous Assignment a presentation on the research being carried out by Adrian Hudnott at Warwick University.  I attended this presentation despite being warned by colleagues that this presentation would make my brain hurt and I can confirm that it did indeed make my brain hurt.

Doug Inkster presented the new features available in Ingres 9.2 and some of the new features in 9.3.  Features such as cached dynamic queries and table auto structure.  Cached dynamic queries takes a prepared query and hashes the query text and uses it to identify a cached query plan removing the need to optimize the query again. Of particular benefit for queries which have long compilation times and short execution times.

Table auto structure implements a different default table structure and enforcement of primary keys.  The default table structure is heap and unique constraints are enforced with a secondary index.  With table auto structure enabled during a CREATE TABLE statement the primary key or unique constraint is used for a B-tree base table structure removing the need for the secondary index.  Previously this would have been achieved using the MODIFY and ALTER TABLE statements after the table had been created.

For the 9.3 and beyond presentation by Christine Normile, I attempted to keep notes on the community wiki http://community.ingres.com/wiki/UK_IUA_2009#9.4_Features and live on IRC #ingres logs available http://irc.planetingres.org/%23ingres.2009-06-09.log.html.

The conference closed with the announcement that next year’s conference will start on June 08 2010 and maybe a 2 day event.

I hope to see you all there.

Related posts

• Ingres Documentation site relaunched
• Heraldic Enterprise Content Management
• Ingres Engineering Summit 2008
• Data Access Message Protocol – Wireshark dissector
• Ingres Migration Toolset