On Wed, Dec 19, 2018 at 08:32:28AM -0300, Alvaro Herrera wrote:
> Hi,
>
> On 2018-Dec-19, Andres Freund wrote:
>
> > Well, that depends on what "non-heap layouts" you're thinking of. I
> > think there'd be some further work needed to make efficient IOTs
> > possible, but the patchset gets us a long way to be able to do that in a
> > pluggable fashion. Biggest problem would probably be extending the
> > existing index AMs, for secondary indexes, to point to a key wider than
> > a tid, without loosing too much efficiency.
>
> I think the important question is will we eventually get the option to
> do "CREATE TABLE ... STORAGE indexorg" (or whatever name) rather than
> are we already getting that feature, and I think the answer is clearly
> yes, so we should keep using the word "heap" in the name as the primary
> feature of the historical implementation.
>
> The "zheap" name makes it clear that it is still a heap; the main
> difference (if I understand correctly) is how does tuple
> updating/deletion work.
>
> The current heap implementation is for "non-overwriting storage
> management", but that's a mouthful and acronyms based on
> "non-overwriting" don't seem great ("noheap" seems a bit silly. Maybe
> "nowheap" is better? How about "nosheap"?)
>
> Maybe we can take the easy way and use something like "stdheap".
>
> Or just "nheap".
oheap for "original?"
Best,
David.
--
David Fetter <david(at)fetter(dot)org> http://fetter.org/
Phone: +1 415 235 3778
Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate
