Pre-Grant Publication Number: 20090119247
Track the progress of public participation in the review of this pending patent application, and view
application details. The menu on the right will help you navigate this patent application. Subscribe to
the community enables you to receive updates on this application via email so that you can easly follow recent activity.
LATEST PRIOR ART
| Date | Title | Reviewer |
|---|---|---|
| 08/25/09 | Hash joins and hash teams in Microsoft SQL Server | Steven Pearson |
| 08/22/09 | Database System Concepts, 5th Ed.; Chapter 13 | Steven Pearson |
| 07/30/09 | US Patent 6253197 | Brian Flynn |
DISCUSSION
Steven Pearson (over 2 years ago)
The proposed 3rd and 4th phases appear most likely to be novel, from the perspective of this one observer. 3rd phase proposes to hash a partition from either left or right table; the techniques I recall would always hash the left partition. 4th phase attacks too-large partitions by subpartitioning and hashing both left and right partitions in a way that combines to product the outer join result. The technique I recall would instead subpartition the too-large partitions by another means (e.g., different hash function), resulting in smaller corresponding subpartitions. The specific approaches in phase 3 and 4 may have prior art, but I am not aware of it at this time. (Phases 1 and 2 though comprise familiar techniques.)
Btw, I think that at [0055] in phase 4 stage 2 description the application has "left" and "right" reversed, perhaps copied from the analogous paragraph in phase 4 stage 1.
I find the claims a bit awkward and not consistent with the overall technique. In particular, Claim 1 appears to try to describe phase 1, and Claim 4 appears to try to describe phase 2. Unfortunately, in doing so, we are left with no claim that covers the complete technique end-to-end. As a result no claim or linked combination of independent and dependent claims properly covers the described technique and instead, if practiced, would produce an incorrect result. I would have expected instead that claim 1 would be higher level, describing a HFUJ technique of four phases, with further dependent claims elaborating on those phases.PEER TO PATENT ACTIVITY
All
Upload + Explain Prior Art
3 submitted
3 submitted
Annotate and Evaluate Prior Art
2 prior art ratings
3 citations
2 prior art ratings
3 citations
Discuss Patent Applications
4 comments posted
Size of Community: 5
4 comments posted
Size of Community: 5
Research Prior Art
0 research notes
0 research notes
WHAT IS THIS APPLICATION ABOUT
