Re: Syntax for partitioning

On Tue, Nov 17, 2009 at 4:31 PM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
> On Thu, 2009-10-29 at 11:15 +0900, Itagaki Takahiro wrote:
>
>> I think syntax support is a good start.
>
> I don't see a syntax-only patch as being any use at all to this
> community.
>
> We go to enormous lengths in other areas to never allow patches with
> restrictions. Why would we allow a patch that is essentially 100%
> restriction? i.e. It does nothing at all. Worse than that, it will
> encourage people to believe it exists in full, when that isn't the case.
>
> The syntax has never really been in question, so it doesn't really move
> us forwards in any direction. This is exactly the kind of shallow
> feature we have always shied away from and that other databases have
> encouraged.
>
> The only reason I can see is that it allows people to develop non-open
> source code that matches how Postgres will work when we get our act
> together. That seems likely to discourage, rather than encourage the
> funding of this work for open source. It may even muddy the water for
> people that don't understand that the real magic happens in the
> internals, not in the syntax.
>
> Why not just wait until we have a whole patch and then apply?

Because big patches are really hard to get applied. Personally, I
think a syntax-only patch makes a lot of sense, as long as the design
is carefully thought about so that it can serve as a foundation for
future work in this area. I don't think "the whole patch" is even
necessarily a well-defined concept in this instance: different people
could have very different ideas about what would constitute a complete
solution, or which aspects of a complete solution are most important
or should be pursued first. Settling on a syntax, and an internal
representation for that syntax, seems like it will make subsequent
discussions about those projects considerably more straightforward,
and it has some value in and of itself since similar notation is used
by other databases.

At least, that's MHO.

...Robert