Subject | Re: [Firebird-Architect] Re: the insert into t select * from t infinite loop |
---|---|
Author | Jim Starkey |
Post date | 2005-07-05T02:45:18Z |
Dmitry Yemanov wrote:
but nothing says that it does have to be implemented that way. Big, big
difference between semantics and an implementation imperative.
>"Not at all. It has the semantics as if it were implemented that way,
>
>>1. open cursor for "select * from a where a.t_val in (select t_val from
>>t)"
>>
>>
>
>And perform a full fetch from the cursor into some temporary space.
>Generally, an open cursor doesn't mean it's materialized.
>
>
>
>>This has the downside that it can be very memory intensive.
>>
>>
>
>Exactly. And this is a major downside.
>
>
>
but nothing says that it does have to be implemented that way. Big, big
difference between semantics and an implementation imperative.