Re: TRANSACTION MONITORS

Couple of comments.

First, X/Open adopted (or more precisely "adopted and adapted") both the DCE-based (i.e., Encina) Transactional RPC interface ("TxRPC") and the XATMI (i.e., Tuxedo) interface (among others). To the best of my knowledge, just about everyone is saying that they will provide "TxRPC", including many proprietary systems like DEC's ACMS and Tandem. Not every product is including XATMI interfaces. The reason, I think, is simplicity of the programming model: everyone knows what a procedure call is.

As mentioned, the X/Open TP communications specs were published at snapshot level. It may take a while before the X/Open versions have enough experience with these particular specs for people to rely on them. There is more to these specs than the API, and I would caution wholesale reliance
on any of them yet.

To put on my product hat: if you like RPCs, then you'll probably like transactional RPCs and Encina, and we've been concentrating on making those good for years.

The comment about Encina being tightly integrated with DCE is true. I am not so sure that it as much a limitation as other systems which may depend on some special SVR4 feature. Just about everyone and his brother has signed up to DCE: from DOS/Windows to AS/400 to VMS to MVS, and everything in between. The point is that every distributed system needs naming, security, and so on. Why invent (and administer) several
(one for mail, one for your database, one for your tp system, ...) when
you can reuse an existing one. Not only do we believe this, but I think the industry (as driven by users) is coming to the same conclusion. For example, Oracle and Sybase also showed DCE-based versions of their systems recently at the OSF. One infrastructure is better than N.

Finally, one should not mistake layers and modularity. All TP systems
(and databases) have logging systems, locking systems, recovery systems
and so on. With Encina, we have carefully drawn the boxes so you can get at those subsystems if you want. Other products hide these subsystems, requiring you to rebuild them if needed in a distributed system. But whether hidden or documented, everybody will log, lock and so on. Just because we have more lines in the Encina diagram does not mean more layers: it means more documentation and careful consideration of how pieces fit together for reuse.

             -Mark Received on Thu Jun 17 1993 - 22:51:57 CEST