Ann,

> There's no blr_suspend, but there is a blr_stall. Even very simple
> execute procedures include one blr_stall. Select procedures seem
> to have at least 2.

I had considered examining the BLR for these blr_stall tokens but it seems
that the byte code of BLR doesn't guarantee that you can isolate the
existence of the blr_stall command from other arbitrary bytes. The BLR would
have to be parsed and the context considered before you could declare that
the byte for blr_stall was actually a suspend. It could also be a binary
portion of a text identifier or number, etc.

It's easy to get familiar with BLR using IB_WISQL. See the Connection's
Browse form where Procedures are displayed. There is a tab that shows the
BLR source for all stored procedures. On can see what Ann is talking about
for themselves.

Just a curiosity question, is it possible to directly update the BLR
contents of the RDB$PROCEDURES BLR column and have the other necessary
validations performed and dependencies, etc. propagated? I suppose I could
meander through the system triggers to see if that is what they are hinging
upon rather than the higher DDL interface. Better yet, I can await the
release of the source...

Wouldn't it be cool if we could tame BLR a little in an editor and write
some procedures using BLR? Perhaps GDML would be more appropriate to
resurrect... Seems to me there are a lot of things that could be
accomplished by doing this. Fortunately the SQL interface has been
sufficient for my needs so far... although I have had to do some pretty
clumsy stuff in SQL...

PS. It would be great if when the subject of a thread wanders to
appropriately modify the subject line.

FWIW,
Jason Wharton
CPS - Mesa AZ
http://www.ibobjects.com