Meanwhile, did you try the information I posted? It helped the person asking
for this at the time...
Also, I think I'm missing something... You're looking for blob length()= 0,
but you mentioned that SELECT ... FROM table WHERE blob_column IS NULL will
not solve the problem.
At this moment I cannot see why?

Regards.

On Tue, Mar 1, 2011 at 10:54 PM, pretzel <davidegrove (AT) gmail (DOT) com> wrote:



--
Fernando Nunes
Portugal

http://informix-technology.blogspot.com
My email works... but I don't check it frequently...

Thank you, Mr. Nunes. I am in process of implementing it. I tried it
VERY quick and dirty, by attempting to use the precompiled blade, but
didn't work. I'm thinking, based on the error message (couldn't find
a library file) that it may be due to a dependency on a library object
from the Sun C compiler, which we don't have. So, now, I'm taking the
more scenic route and getting gcc (which wasn't on this machine), and
recompiling, etc.


I don't represent myself as an expert, so could be totally full of
beans. But, my hunch is that the blob column isn't empty, but, in
fact, has a proper blob handle (meaning, it is properly not NULL).
It's just that the blob to which the handle points has zero length.
I'm guessing that it could have gotten that way by execution of an SQL
"INSERT filetolo(<filename>... ) ..." with an empty file supplied as
the argument.

Anyway, I'm going to try the blade.

Thank you, once again.

DG

Got this from tech support:

"I was following up on your request to extend the LENGTH()
functionality to BLOBs and CLOBs, but is so happens that there is one
Feature Request for that matter already, and we hope to see that
implemented in future releases"

DG