Re: [office] Please read: ODF 1.2 process proposal

[robert_weir@us.ibm.com]

Michael.Brauer@Sun.COM wrote on 12/05/2008 07:00:28 AM:

> 
> If you have any questions or amendments regarding the proposal then 
> please don't hesitate to send them to the list, so that we may address 
> them before or in the call on Monday. I believe it would be good if we 
> in the call on Monday could agree on a specific way how to move forward, 

> which may be this proposal or an amended one. This would allow us to 
> spend the rest of this year and the first weeks of the next one with 
> actually discussing and approving proposals, and would further give us a 

> clear perspective regarding ODF 1.2 before the holiday break.
> 

Thanks for putting this proposal together, Michael.  I do have a few quick 
comments and observations I'd like to share.

First, by way of information, I'd like to point out a spreadsheet Mary 
McRae has put together that does the date math for calculating the public 
review and ballot requirements for an OASIS standard.  You can read more 
about it on Mary's blog here: 
http://oasistcadmin.wordpress.com/2008/08/22/how-long-will-it-take/

You enter in a date on which the specification is ready for public review. 
 So all the technical content is done, the required OASIS material is 
done, the editing is done.  Your estimate was that it would take 18 TC 
calls to discuss and approve all 47 new proposals.  That would take us out 
to May 4th for discussing and approving proposals.  There would be further 
work required for accessibility proposals, for conformance and scope 
clause work, for processing public comments, etc.  That would likely take 
at least another 4 meetings.  So that hypothetically comes out to June 1st 
for the TC voting to approve ODF 1.2 ready for sending out to public 
review.  Plug that into the spreadsheet and out comes November 1st, 2009 
for final approval as an OASIS Standard.  And that is assuming that we do 
not make any changes based on the public review.  If we do make changes, 
then additional editing work is required, as well as an additional 15-day 
public review takes us out to December or later.  And I can almost 
guarantee you that with a standard as long and complex as ODF we will 
catch things during the public review that we will need to fix.

I'm hoping that TC members will agree that completing ODF 1.2 in November 
or December 2009 does not match their expectations or desires.  We have 
too many new killer features in ODF 1.2, like the metadata work, as well 
as items like OpenFormula and presentation tables that users have been 
waiting for.  We need to get this revision out to be reviewed, approved 
and implemented.

So what do we do we need to do?

First, we could simply rush through the 47 proposals by an arbitrary 
deadline.  But I don't think that works out well, since quality would 
suffer.  If you were writing code for a product, and a deadline was coming 
up, would you rush to get all the features in by the deadline?  Certainly, 
I've seen that happen before.  But it usually leads to more problems.

The better solution is to prioritize the remaining work and do only so 
much as you can do well.  What isn't done in ODF 1.2 is then proposed for 
the next release of ODF.

There is obviously many ways in which we can prioritize the remaining 
proposals.  Do we do 46 of them?  45 of them?  10 of them?  5 of them? 
Each one would require a different amount of time and you can calculate 
the schedule impact of these choices based on Mary's spreadsheet.  Michael 
is suggesting that we process around 5 more proposals, so the work would 
completed by January 12th, and then likely going out for public review 
shortly thereafter.  The exactly date would depend somewhat on Patrick, 
but it would set us on a trajectory for ODF 1.2 being approved as an OASIS 
Standard in July.

So I think we want to keep two things in mind when considering Michael's 
proposal.  First, is the TC in agreement that a December end-date for ODF 
1.2 is unacceptable and a July date is preferred, and that it is not worth 
considering intermediate dates such as August or September?

If so, does the TC agree with the details of Michael's proposal for 
getting us there, including the detailed deadlines over the next few 
weeks?  If not, are their amended or alternative proposals that would get 
us to the same dates?

Personally I think Michael's proposal will work, but only if TC members 
make an effort to prioritize their own proposals so we discuss and approve 
only 5 or so additional proposals for ODF 1.2.  If everyone tries to get 
all 47 proposals discussed and approved on this schedule, then we'll get 
poor quality, chaos, or both.

I'd be interested in your thoughts, either on the list or on tomorrow's TC 
call.

Regards,

-Rob