So, where to go from here? I'm acutely aware that we're hard up against the final 9.6 commitfest, and that we discourage major patches arriving so late in a devel cycle. But I simply couldn't get this done any faster. I don't really want to hold it over for the 9.7 devel cycle. It's been enough trouble maintaining this patch in the face of conflicting commits over the last year or so (it's probably still got bugs related to parallel query...), and there definitely are conflicting patches in the upcoming 'fest. And the lack of this infrastructure is blocking progress on FDWs and some other things.
Thanks for working on this; it is important.
I'm disappointed to see you do this because of FDWs, with the "some other things" like parallel aggregation not getting a mention by name.
While I wouldn't mind seeing this go in, what worries me is the multiple other patches that now need to be rewritten to exploit this and since some aren't mentioned would it be reasonable to imagine those other things won't be prioritised for this release? Or will we be deciding to elongate the integration phase to cope with this? Delay or favour forks, which should we choose?
Anyway, glad to see you will now experience the problems of maintaining large patches across multiple releases and/or the difficulty of arguing in favour of patches that still require work going in at the last minute. Not with relish, just so that understanding isn't limited to the usual suspects of feature-crime.
--
Simon Riggs http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services