Re: BUG #5490: Using distinct for select list causes insert of timestamp string literal to fail
От | Greg Stark |
---|---|
Тема | Re: BUG #5490: Using distinct for select list causes insert of timestamp string literal to fail |
Дата | |
Msg-id | AANLkTinvpWLi3CoBWYgNPxVdFbbLhEHZuYdp9buvDubV@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: BUG #5490: Using distinct for select list causes insert of timestamp string literal to fail (Farid Zidan <farid@zidsoft.com>) |
Ответы |
Re: BUG #5490: Using distinct for select list causes insert of timestamp
string literal to fail
|
Список | pgsql-bugs |
On Fri, Jun 4, 2010 at 11:15 PM, Farid Zidan <farid@zidsoft.com> wrote: > Now this not rocket science, it's simple insert statement where we do not > want duplicates inserted. Works on 10 other DBMSs. > I find usually when one person is arguing something is complex and someone else is arguing it's simple it's the person who's claiming it's simple who is wrong. The other databases are not, I believe, preventing duplicates from being inserted as you describe. They are removing duplicates from the string constants and then silently converting to a different datatype before inserting. When postgres removed these default casts to text it turned up many instances where users had buggy code and Postgres had been hiding from them by silently using string operators which was not what users were expecting. In other words, while it might not matter in this case, in general if you code in this style your code is buggy and these other database implementations are not doing you any favours by making it appear to work correctly most of the time. -- greg
В списке pgsql-bugs по дате отправления: