Re: [HACKERS] Optimizer is fixed, and faster

Bruce Momjian wrote:
> 
> This is exactly what I need.  My testing is very limited.  I basically
> test the functionality, but not real-world samples.  I am still working.
> I will let everyone know when I am done, and you can throw any queries
> at it.
> 
> Was there a speedup with the new optimizer?  Was the new plan faster?
> The new optimizer uses 'cost' much more reliably.  I hope our cost
> estimates for various join types is accurate. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
No. It's ok for netsloop only.

Vadim