Tom,
>> -- Average 20% overall performance improvement
>
> Hmm, do we have any numbers backing that up? Some things improved much
> more than that, some not at all, so I'm kinda wondering where "20% overall"
> came from. If you're basing that on one specific benchmark you oughta
> say which one.
TPCC scores, according to Unisys, and the Tweakers.net benchmark. The
actual press release is more specific. The idea was for the community
announcement to be brief.
So:
Higher performance (+20% in OLTP benchmarks)
>
>> -- Warm standby databases
>
> I think we could do that already, it just didn't work as smoothly.
> Maybe "Better support for warm standby databases" would be better.
Ok, "improved"
--Josh
