Hi Helen,

> > > > We *should* be able to use this syntax in DSQL:
> > > >
> > > > UPDATE blob_test
> > > > SET blb_text = :zipped_file
> > > > FILTER FROM -2 TO 1;
> >
> >Naah, won't do. Think of multiple SETs and INSERTs as
> >well. A more generalized syntax would be better.
> >
> >How about:
> >
> >SET blb_text = FILTER(:zipped_file, -2, 1),
> >
> >INSERT INTO ... VALUES (FILTER(:zipped_file, -2, 1), ...)
>
> What would be the purpose of introducing a totally different syntax that

is

> inconsistent with ESQL, which is the SQL "super-set" of DSQL?
>
> Why do you think the existing syntax won't handle multiple column inserts
> and updates?

Obviously, I haven't looked at ESQL - I never use it :-)

(checking docs now)

...

(checking some more)

...

I cannot find out how to do this for multi-column inserts/updates,
when filtering from/to multiple sub types. Can you give an example?

Anyway, IMO, my own proposed syntax is clearly on a per-column
basis and rather easy to read/use. It could also be used in a SELECT
statement:

SELECT FILTER(MyBlobColumn, -2, 1)
etc...

With regards,

Martijn Tonies
Database Workbench - developer tool for InterBase, Firebird, MySQL & MS SQL
Server.
Upscene Productions
http://www.upscene.com