state table RMD 7D

[Christopher B Ferris <chrisfer@us.ibm.com>]

Regarding the state table [1] cell 7D.

It isn't abundantly clear what it means
for there to be an "unrecoverable error on creation". I must
assume that this is different than cell 6D in that instead of the RMD being
able to continue to function, it enters a FUBAR state. e.g.The RMD could
crash or the hard disk be corrupted, in which case, IMO the state table
reverts to initial state.

I would also point out that an unrecoverable
error can occur at any point during the life of the RMS and/or RMD. I think
that we need to characterize
"unrecoverable" as being just
that, a condition from which there is no possible recovery. In the case
where an implementation retains the sequence state in persistent non-volatile
store, then I think that we must assume that "unrecoverable"
means that the state is lost. All sequences for which this condition is
true should revert to the initial state.

However, what this points out is that
it is possible for the CreateSequenceResponse or CreateSequenceRefused
to NOT be received by the RMS.
Seems that it would be an implementation
detail as to how long an RMS awaits the response for a CreateSequence request
message, but one must assume that 
eventually the Creating state would
timeout and cause a CreateSequence to be resent or generate a fault back
up to the AS. In any event, this is implementation detail that I believe
is out of scope for the TC/spec.

Cheers,

Christopher Ferris
STSM, Emerging e-business Industry Architecture
email: chrisfer@us.ibm.com
blog: http://webpages.charter.net/chrisfer/blog.html
phone: +1 508 377 9295