On 31 Jul 2015 11:59, "Tatsuo Ishii" <is...@postgresql.org> wrote: > > > On 31 Jul 2015 10:15, "Tatsuo Ishii" <is...@postgresql.org> wrote: > >> > >> > I think it would be nice to have... but not to the point of working on > >> > it myself. > >> > > >> > Might be worth an email to -general to see how many people have > >> > immediate use for it. > >> > >> What I am thinking about is, > >> > >> 1) Implement certain class of updatable views allowed in SQL:1999 > >> (UNION ALL, natural joins) > >> > >> 2) Anything beyond #1 (I have no idea for now) > >> > >> Let me see how people are interested in... > >> > > > > How does the standard define it? Do they also follow the same MVCC > > semantics as normal tables? > > In my understanding there's no such concept like MVCC in the standard. > Anyway in our implementation, we should keep the MVCC semantics of > course. >
Yes I meant our internal MVCC semantics. I will have to look at the way MVCC handles views for exact logic though > > I am concerned that we may end up losing read > > performance for views if we implement this (unless I am missing something) > > Why do updatable views lose read performance? I thought the only > performance concern will be in the update/delete/insert operations. I meant update, sorry. Pre coffee mails tend to be incorrect :)