Re: SQL:2011 application time
От | Vik Fearing |
---|---|
Тема | Re: SQL:2011 application time |
Дата | |
Msg-id | 1b972c50-1b0f-4753-8a33-b6d8cce8e598@postgresfriends.org обсуждение исходный текст |
Ответ на | Re: SQL:2011 application time (Paul Jungwirth <pj@illuminatedcomputing.com>) |
Список | pgsql-hackers |
On 12/2/23 19:11, Paul Jungwirth wrote: > Thank you again for such thorough reviews! > > On Thu, Nov 16, 2023 at 11:12 PM jian he <jian.universality@gmail.com> > wrote: > > UPDATE FOR PORTION OF, may need insert privilege. We also need to > document this. > > Similarly, we also need to apply the above logic to DELETE FOR > PORTION OF. > > I don't think UPDATE/DELETE FOR PORTION OF is supposed to require INSERT > permission. > > Notionally the INSERTs are just to preserve what was there already, not > to add new data. > The idea is that a temporal table is equivalent to a table with one row > for every "instant", > i.e. one row per microsecond/second/day/whatever-time-resolution. Of > course that would be too slow, > so we use PERIODs/ranges instead, but the behavior should be the same. > Date's book has a good discussion of this idea. > > I also checked the SQL:2011 draft standard, and there is a section > called Access Rules in Part 2: SQL/Foundation for UPDATE and DELETE > statements. Those sections say you need UPDATE/DELETE privileges, but > say nothing about needing INSERT privileges. That is on page 949 and 972 > of the PDFs from the "SQL:20nn Working Draft Documents" link at [1]. If > someone has a copy of SQL:2016 maybe something was changed, but I would > be surprised Nothing has changed here in SQL:2023 (or since). -- Vik Fearing
В списке pgsql-hackers по дате отправления: