Re: BETWEEN Node & DROP COLUMN
От | Tom Lane |
---|---|
Тема | Re: BETWEEN Node & DROP COLUMN |
Дата | |
Msg-id | 10705.1025878493@sss.pgh.pa.us обсуждение исходный текст |
Ответ на | Re: BETWEEN Node & DROP COLUMN ("Christopher Kings-Lynne" <chriskl@familyhealth.com.au>) |
Список | pgsql-hackers |
"Christopher Kings-Lynne" <chriskl@familyhealth.com.au> writes: >> What happens if I drop an inherited column in a child table? Maybe it >> works, but what happens when I SELECT the column in the parent table? > Well, what happens if you rename a column in a child table? Same problem? Ideally we should disallow both of those, as well as cases like changing the column type. It might be that we can use the pg_depend stuff to enforce this (by setting up dependency links from child to parent). However that would introduce a ton of overhead in a regular DROP TABLE, and you'd still need specialized code to prevent the RENAME case (pg_depend wouldn't care about that). I'm thinking that it's worth adding an attisinherited column to pg_attribute to make these rules easy to enforce. regards, tom lane
В списке pgsql-hackers по дате отправления: