Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!sdd.hp.com!hplabs!hpcc05!hpyhde1!hpycla!hpcuhc!dhepner
From: dhepner@hpcuhc.cup.hp.com (Dan Hepner)
Newsgroups: comp.databases
Subject: Re: XA Compliant RDBMSs
Message-ID: <2060010@hpcuhc.cup.hp.com>
Date: 1 Apr 91 22:19:40 GMT
References: <1120@tetrauk.UUCP>
Organization: Hewlett Packard, Cupertino
Lines: 34

From: nick@tetrauk.UUCP (Nick Rumble)

>Does anyone know of any other RDBMSs that are currently
>or will soon be (next 12 months) XA compliant?

If you're looking for plug-n-play, you're a bit early.

There are three different definitions of what "XA compliant"
means.

1.  A de-facto standard set by Tuxedo/T
2.  The Preliminary Specification of XA, published by X/Open
3.  The Developer's Specification of XA, as yet unpublished by
    X/Open.

They aren't all the same. 

It is unrealistic at this time to make too much of a claim that 
certain RM products are described as "XA compliant".   The
only real informational content is that certain TM-RM combinations, 
e.g.  Tuxedo/T - Informix,  are available.  It's also quite
possible to make use of Tuxedo/T - DBMS combinations which 
have no utilization of the XA interface at all.

For newcomers, "XA" is a defined interface between "Transaction
Managers", and "Resource Managers", whereby the TM and RM
exchange information about global transactions, and implement
a two phase commit protocol allowing atomic completion of such
transactions over multiple machines / databases. 

Could you be encouraged to explain to us the class of problem for
which you've found /T to be a valuable solution?
 
Dan Hepner