Re: PF Compiled Files with Dictionaries vs. SQL-Created Tables

On 2/3/11 8:48 AM, caura wrote:

People create databases only for 1 reason:
To exploit those databases.

So first of all, the database should do what it's supposed to do.
Secondly performance comes in.

One major topic in exploiting databases, is editing the data into a
usable and user-friendly form.
We are waiting for over 20 years for any support of editing in SQL.
As far as I know, V7R1 or any SQL for DB2 up to V10 still does not
support editing in any possible way.

The simplest editing in SQL ends up in esotheric casting and
substringing. Sorry, but I have never understood this huge black
hole in a "modern product".

PS: DDS has never supported any editing on "character-data" either.

I think that a new "Subject:" thread would be more appropriate to discuss whatever is meant; hopefully to include some specific examples. While I suppose that both CREATE TRIGGER and CREATE FUNCTION LANGUAGE SQL and CREATE PROCEDURE LANGUAGE SQL might be considered limited for supporting too few scalar character "editing" functions beyond the simplistic SUBSTR, LOCATE, REPLACE, etc, and that even having CAST capabilities\functions along with those scalar functions would still be insufficient for [being considered] robust string processing. However EXTERNAL functions and procedures, and "native" triggers can be used to access any various robust string manipulation code\language which is not an integrated feature of the SQLanguage itself.

One must also consider that the SQL is defined by standards, which the IBM DB2 follow [ANSI], as well as for which IBM provides input. The standards organizations would probably need to be lobbied for any specific support, since they apparently have been and are unlikely to propose whatever has been long awaited and presumably still not forthcoming.

Regards, Chuck