[odf-discuss] ODF Interop

[robert_weir at us.ibm.com]

Marbux wrote on 10/31/2007 03:52:26 PM:

> 
> Rob, I think you missed the part where Jody said he wants an 
> interoperable product. Given that five separate Novell proposals to 
> enable interop with MS Office -- three signed off on by 
> Massachusetts ITD -- never even made it onto the TC's agenda list of
> action items, you might have more luck in enticing people to 
> contribute to ODF if you used your new power as TC co-chair and 
> power to out-vote Sun to put the Novell proposals on the agenda. As 
> it is, the TC has run off the only people who were actually working 
> on high fidelity round-trip interop between ODF apps and MS Office. 
> And the decision on preservation of xml:id attributes was the last 
> straw for us. That broke interoperability even among ODF apps. 
> 

Hmmm... were these Foundation proposals or Novell proposals?  I am hearing 
conflicting accounts of this.

In any case, I will make it my special task to ensure that far fewer 
proposals are "lost" by Novell, the Foundation or anyone else.  I think we 
need some better workflow management to track exactly when proposals are 
made and acted on. 

A specific example to consider:

The Foundation's web site (http://opendocumentfoundation.us/index.htm) 
currently has PDF file called "Universal Interop Framework Proposal".  The 
header of this document says "Submitted to the OASIS Office Technical 
Committee by The OpenDocument Foundation October 16, 2007."  However, no 
such proposal was submitted to the ODF TC on that date or any other date. 
Are we going to hear complaints months from now that the ODF TC refused to 
put this on the agenda?  As I said before, we need better tracking of 
proposals, so we can ensure that no more proposals are "lost" in this way.

Also, consider that if the Foundation (or is it Novell?) thinks that their 
proposals are not given a fair hearing, then they should escalate within 
the defined appeals process OASIS.  It is the job of the Chair to make the 
agenda, but this should not be done by filtering proposals according to 
their own corporate interests.  All TC members should have an equal right 
to have their proposals heard, discussed and voted on.  I have not noticed 
any abuse of this kind.  But if you (or is it Novell?) believe this 
principle has been violated, then your first course of action is to 
immediately appeal the decision of the Chair in the meeting (according to 
Robert's Rules of Order) and if that does not give satisfaction then to 
follow the OASIS appeals process to escalate this to the OASIS Board of 
Directors:  http://www.oasis-open.org/committees/process.php#s4.2

Since there are clearly documented ways of appealing TC Chair decisions, 
and you have not availed yourself of them (to my knowledge at least) I'm 
puzzled that you are choosing instead to bluster in public about how you 
have been wronged. If there is merit to your case, then why not press it?

I have nothing to say on Massachusetts signing off on Novell's proposals. 
I was not even aware that Novell was involved in Massachusetts.  Or was 
this the Foundation?


> A visible signal that the TC has a new-found interest in enabling 
> interoperability might establish an incentive for folks who care 
> about interop to work on the spec. But as things stand, I personally
> think Jody would be better advised to tell you to take the 
> proverbial flying leap unless you're ready to commit here and now to
> add an interoperability framework to ODF and to eliminate the 
> discretion for ODF apps to trash each other's metadata. 
> 

What we have done is start an interoperability initiative within the OASIS 
ODF Adoption TC.  As part of this initiative we've had our first workshop, 
in Barcelona, to bring together the major (and minor) ODF vendors to try 
out some interoperability scenarios and discuss approaches to improving 
interoperability.  MS Office plugin's were tested as part of this 
workshop, including Sun's and Microsoft's.  I wish the Foundation had 
participated in this effort.

I'm not a big fan of silver bullets.  Maybe they exist, maybe they don't. 
But I'm more inclined to take a conventional approach to interoperability, 
meaning iterating on the text of the standard to eliminate ambiguities, 
work on test suites and compliance tests and eventually logo 
certification.  That is the approach used for interoperability in a vast 
number of standards.  Are there other approaches?  Sure.  But my 
estimation of the Foundation' approach is that it is unproven and unlikely 
to work.  Of course, if I'm proven wrong then I would be even happier, 
since I would then have 100% compatibility with only 5 easy changes to 
ODF, as the Foundation claims.  I'd like to believe in the Foundation's 
approach.  I'd also like to believe in Faerie Godmothers and Santa Claus. 
But until I see technical details and running code to back up the 
Foundation's extraordinary claims, I'll continue to recommend the 
conventional approach to interoperability.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.opendocumentfellowship.com/pipermail/odf-discuss/attachments/20071101/9d083a1d/attachment.html