Archives of the TeradataForum

Message Posted: Wed, 05 Oct 2011 @ 13:24:00 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Thanks Dieter and Eric Barner, your comments are obviously based upon sound (and no doubt sometimes bitter) experience.

I accept all your points .... but I'm still left thinking that it would be nice if Teradata allowed for more than just C/C++ and COBOL as the languages that can be used for UDF's.

After all, I'm a bit long in the tooth for learning C, my thinking being that it could be a recipe for disaster to mess with C unless you really know what you're doing.

However, why Teradata couldn't allow a fast running scripting language (something like Python) is beyond me. Ok, they're not as fast as native C, but they're not far behind.

If I was being critical of Teradata (and don't get me wrong, I'm a fan of Teradata), there are three things it should have provided before now :-

i) UDF's in a High Level language (other than Cobol), even if it's one that is Teradata's own.

ii) Architecture aside, make Cursor processing reasonably fast (with the caveats that it's never going to be as fast as parallel Set Based processing)

iii) More flexibility within BTEQ, such that you can return variables from Stored Procedures (for example) and proper IF - THEN processing.

Others may disagree but that's where I've been most frustrated with Teradata, at least in what I've personally been involved with.

Regards

David Clough
Senior Database Designer
Database Design Group

	<Prev	Next>		<<First	<Prev	Next>	Last>>