Re: WorkUnit guard semantics issue

Nick,

Referring back to the original proposal submissions, or early combined 
drafts of the ORACLE+CommerceOne work may show the original intent, but that 
does not change the fact that the Sept 22nd draft clearly states the fact 
that a blocking workunit, whose guard condition evaluates to false, will 
fail to match and therefore be skipped.

This is the semantics that Steve and I, and possibly some others, have come 
to expect and think is appropriate. You must also bear in mind that I only 
became fully active within the group around the time that this draft was 
adopted, and therefore this became my understanding of where CDL had 
progressed to. So referring back to previous intent is not relevant for me - 
only what was clearly documented and adopted by the group at that time.

The fact that this raises other issues that need to be addressed, does not 
change the fact that we believe this is the most intuitive semantics for the 
workunit guard condition.

Regards
Gary 

Received on Thursday, 9 December 2004 09:53:47 UTC