Hi David,
For now, the goal of the developments in the "headless" branch are mainly triggered by and aimed at our desire to radically clean-up the MonetDB code base, both conceptually and literally. Replacing BATs by COLs indeed is a major part of this clean-up, mainly triggered by the observation that, in practice, BATs are mostly (only?) void-headed. Conceptually, COLs are nothing else but BATs with an implicit (rather than explicit) void (dense) head with seqbase 0.
Yes, we plan to merge the headless branch back into the tip once we are confident that we are on the right track. However, we cannot yet estimate when this will be, and whether there will still be a release of the current tip, i.e., after Mar2011, but before merging the headless branch back into the tip.
We'll keep our user community posted as soon as we know and/or changes become relevant for you.
Kind regards, Stefan
On Mon, Feb 14, 2011 at 04:52:19PM -0800, dsuh wrote:
I noticed a new headless branch that aims to replace BATs with COLs. Could you give me more information on COLs? Is what you are trying to achieve faster queries, less memory usage, or other advantages? Assuming that this is the new direction for MonetDB, do you expect that it will be merged with the Hg default tip for the next release after March 2011? Thanks.
David Suh
The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb
Monetdb-developers mailing list Monetdb-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/monetdb-developers