Re: Can't PL/SQL do the simple stuff that Transact-SQL can?

Having worked on SQL-Server's Transact-SQL and then moving to PL/SQL I've encountered similar moments of disbelief.

PL/SQL is a very powerful language. I love its error-handling capabilities and its innate structure-building tendencies. But it has limitations. I suspect you will discover (if not already) the severve limitations Oracle PL/SQL imposes on temp tables. When you look at the functionality for temp tables you will discover you cannot use SQL against these tables! This really annoys me. Being forced to loop through embedded cursor after embedded cursor and write all the 'FETCH INTO ...' syntax adds untold hours to my code creation and debugging efforts. Like you I've learned to avoid cursors whenever possible due to perforamance reasons but with Oracle you have no choice. The whole purpose of SQL was set operations! Instead I'm (finding myself) dealing with each individual record via these fetch statements..

I've been working with PL/SQL for 4 months now. I currently have a distinctly mixed attitude about it. I've worked with SQL-Server and Informix backends before this. I hope after another four months I will discover more 'attractions' then 'turn offs.'

Jonathan Tew wrote:

> Wow, thanks for posting a solution to the problem. That strikes we as
> really complex code though. I've been told that cursors are very
> inefficent things and should be used as little as possible. Under MS
> SQL Server often a cursor will blow up in a stored proc causing
> problems. Does Oracle have such problems with cursors and would this be
> a bad thing performance wise? I'm under the impression that the kinds
> of things that I'm trying to do are simple and common... why hasn't
> Oracle taken care of this problem with an exists() construct?
>
> Thanks,
> Jonathan Tew
Received on Fri Feb 19 1999 - 09:29:51 CST