A (actually, the first) CMIS TC F2F in Redmond is coming to an end (ok, at the time of this writing, folks are still considering the issues that have been raised so far (see the CMIS Jira at http://tools.oasis-open.org/issues/browse/CMIS)). Many valuable contributions and discussions evolving around controversial aspects have taken place. Not surprising, one of the important results is the aggressive schedule set for the release of CMIS 1.0. The preliminary schedule leading to a first public draft can be found here: http://www.oasis-open.org/committees/download.php/30951/CMIS%20strawman%20schedule.xls. This schedule requires some functions to be restricted to a common denominator or to be dropped entirely. Thus, ACLs (particularly, the distinction between reading and writing ACLs), batch processing, and mix-ins (or aspects, think decorators when not familiar with AOP) will need to be further refined.
According to the prelimary schedule, the next F2F shall take place sometime around AIIM in April. A plug-fest is due at that time, basically to prove interoperability between the different implementations of the draft spec at that time. Btw, this also points to one of the more obvious problems right at the moment: the lack of publicly available implementations to play around with. Take a look at the JackRabbit CMIS sandbox (and get involved!) to learn more: http://wiki.apache.org/jackrabbit/SandboxCMIS
On side note: since participation on the phone made it hard to understand each and every part of the discussion, corrections and / or amendments to the above brief summary are welcome.