| From: | Andrew Borodin <borodin(at)octonica(dot)com> |
|---|---|
| To: | Andrew Borodin <amborodin(at)acm(dot)org> |
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>, Alexander Korotkov <a(dot)korotkov(at)postgrespro(dot)ru>, Михаил Бахтерев <mike(dot)bakhterev(at)gmail(dot)com> |
| Subject: | Re: GiST penalty functions [PoC] |
| Date: | 2016-09-01 18:01:17 |
| Message-ID: | CAJEAwVGdb92E-XKfMLN3cxM2BWbbA3rrffzDzg8Ki1H5iQEk2Q@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Here is new version of the patch.
Now function pack_float is commented better.
All inline keywords are removed. I haven't found any performance loss for that.
Remove of static's showed 1%-7% performance loss in SELECT computation
(3 test runs), so I left statics where they are.
Actually, to avoid this kind of hacks, probably, it would be better to
fork GiST to GiSTv2 and implement many API changes there:
1. Bulk load API
2. Richier choose subtree API
3. Allow for key test not just NO\MAYBE answers, but SURE answer to
skip key test for underlying tree
And some other improvements require breaking chanes for extensions.
GiST idea is almost 25, modern spatial indices vary a lot from things
that were there during 90th.
Best regards, Andrey Borodin, Octonica & Ural Federal University.
| Attachment | Content-Type | Size |
|---|---|---|
| cube_improved_penalty_v3.diff | text/plain | 5.7 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Jesper Pedersen | 2016-09-01 18:03:42 | Re: Hash Indexes |
| Previous Message | Tom Lane | 2016-09-01 17:55:32 | Re: \timing interval |