Ann,

> Sorry I accidentally dropped Dmitry's question about implementation.

My

> assumption was that transient data sets would be stored temporary

files

> structured like database files. Indexes & blobs would be stored in

the

> same file with the data. It probably makes sense to have one file per
> session, rather than one per transient data set.

Why not store them in the database file itself?

Just implement a table name 'redirection' routine which would map TDS
name to the physical db table (which would have a special name like
TDB$FFFFFF-FFFFF-FFFFFF-FFFF-FFFFF).

As I see it, it would simplify the implementation since the engine would
see the TDS just like a real table (registered in RDB$RELATIONS), and
all existing table functionality would be supported.

When the engine opened a DB, the engine would perform the necessary
cleanup of any previously existing TDSs.

Regards,


Sean