OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

office message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: RE: [office] Proposal: Align IS 26300 to ODF 1.1 instead of 1.0 maintenance

"Dennis E. Hamilton" <dennis.hamilton@acm.org> wrote on 01/04/2010 
04:22:28 PM:
> I don't propose that we renege on defect processing but that we apply 
> ones we can to a 1.1 errata and bring them back to ISO as the IS 26300
> amendment for 1.1.  We could respond with dispositions but agree not to
> create more corrigenda until we have the 1.1-aligned IS 26300 to apply 
> to.
> This might make some delays but it would have 1.1 be the common baseline
> (however we agree on equivalence) soonest. 
> I'm not sure which is the longest tent pole.  Our six month rule about
> errata (once we produced a 1.0 Errata 02) or the time that it takes to 
> draft corrigenda through SC34 once approved as errata at OASIS.  I have
> assumed that the second would be a greater barrier to accomplishing 1.1
> alignment early.  While we're constrained on when we could have more 
> (after 1.0 Errata 02 or 1.1 Errata 01) to the respective OASIS 
standards, it
> seemed like getting 1.1 into that stream was the most effective thing we
> could be doing, all else being equal.

The time periods are like this:

Approved Errata we can do every 6 months.  In additional to whatever 
editing work needed to create them, there is around 1 month of public 
review and administrative overhead it getting them done.

Technical Corrigenda trigger a 3 months ballot in SC34.  Generally they 
can be triggered at short notice, if we're in the right form.  So I 
wouldn't assume any delay between completion of OASIS Approved Errata and 
the start of a DCOR ballot.  I know there has been delay on this in the 
past, sometimes amounting to almost a year, but that can not be explained 
by any procedural requirement. The delay might have been as little as a 
day or two if all parties concerned wanted to move things along.

An amendment, in addition to whatever time would be required to prepare 
the document, would require additionally that we meet the requirements of 
the OASIS Liaison policy.  On the JTC1 side an amendment requires both a 
ballot in SC34 as well as JTC1.  I believe these are sequential.  So the 
net result is around a 6-8 month delay.

>   1.1 Perhaps you know what the calendar setback would be for submitting
> 1.1-as-amendment before the September 2010 SC34 plenary and could it be 
> earlier than that via SC34 WG6?

The dates of the plenary are irrelevant for submitting an amendment. 
Important thing to note is that there is no PAS process for amendments. So 
you need to go through the NP, WD, FDPAM, FDAM and DAM ballot procedures 
at the WG, SC and JTC1 levels.  I don't believe you can do an NP vote at a 
meeting.  So NP ballot=3 months, FPDAM ballot is 4-6 months, FDAM ballot 
is 2 months.  Add it all up and I think that ODF 1.2 will pass it in 
processing, since PAS rules send it directly to JTC1 for a 5-month ballot 
under the new rules.

But to answer your questions in general, producing material for discussion 
at a meeting would need to be done 4 weeks in advance.

>   1.2 Also, I couldn't find anything in the JTC1 procedures that helped 
> understand what the checkpoints and lag times are for processing of a 
> submission.  Do you have some staging information that applies to that 
> once OASIS makes the submission?

Keep in mind that JTC1 procedures are being rewritten in the areas of Fast 
Track and PAS.  But generally it looks like this:

1) OASIS process for Committee Draft, Public Review Draft and Committee 
2) OASIS ballot (1 month)
3) Requirements of OASIS Liaison Policy 

If we submit when the current rules are active, then it looks like the 
JTC1 PAS ballot is 6 months.  If the new rules are active, it would be a 5 
month ballot.  It looks like the rules will likely change in July.

Obviously you can receive comments during the ballot and a BRM may be 
necessary to discuss those comments.  But this is something that can 
happen for an amendment as well.  So I'd consider that to be an additional 
4 month end-game activity either way.

The administrative delay from submitting a PAS submission to JTC1 and the 
ballot issuing should be small, less than a month.

>   1.3 With regard to the OASIS policies and procedures for submissions 
> another standards body, the question seems to be whether submission of 
> amendment for 1.1 alignment triggers that process, especially provision 
> on conduct of an OASIS Interop Demonstration.  I agree this might be a
> show-stopper.

My understanding is that requirement is triggered.  We obviously need to 
consider this for ODF 1.2, in approximately the same time frame.  So we 
need to solve this one way or another.  Maybe we can have a single Interop 
Demo to satisfy the requirements for both ODF 1.1 and ODF 1.2?

>   2.1 I'm assuming that creating a version of 1.1 that has the errata
> applied is a production matter and not something that requires 
processing of
> a new committee specification and taking the update through the OASIS
> Standard approval process.  Even if it were to require an OASIS ballot, 
> is apparently a thirty-day deal if I am reading the TCScheduler 
> correctly.

OASIS TC Process 3.5:

"A TC may approve a set of Errata to an OASIS Standard as "Approved 
Errata" to the corrected specification by: 
(a) Adopting the set of proposed corrections as a Committee Draft, in the 
form of a list of changes, and optionally accompanied by a copy of the 
original specification text marked to incorporate the proposed changes."
So including the errata in a reprinting of the standard seems to be 
allowed without additional approval.  But our issue is that we have no 
Approved Errata for ODF 1.1.

>   2.2 I don't understand the exact process for taking such a 1.1 to SC34 
> an amendment to IS 26300, and I'm not clear how a "diff" is handled, 
> we mean some sort of change-tracking version that has been "diff"ed 
> IS 26300 (that is, 1.0 edition 2?) so we see deletions and insertions
> against 26300?

If I were doing it, I would simply do this:

a) Create Approved Errata for ODF 1.1 corresponding to all corrections 
made to ODF 1.0 via Approved Errata
b) Apply all of these errata to the ODF 1.1 text
c) Apply all of the ODF 1.0 errata to the ODF 1.0 (second edition) text
d) The diff between c and b is your amendment.

>   2.3 I agree that this might be constrained by the OASIS policy (see 
> above) and we should find out how that impacts the presumption of 
> for this approach.
>   3.1 My wildly-optimistic trial calendar for approval of ODF 1.2 
> that we couldn't be making a PAS submission of an ODF 1.2 OASIS standard 
> JTC1 before October, 2010, and I didn't even consider the
> three-independent-implementations requirement.  I don't understand the
> timeline for the submission within JTC1 so I don't feel comfortable 
> allowance for JTC1 procedural requirements and how they fit on a 
calendar of
> SC34 plenary cycles, etc.  I do feel quite safe in assuming we wouldn't 
> an ISO/IEC version before 2011. 

The requirement is for three "Statement of Use" by OASIS member 
organizations.  This does not necessarily require that there be three 
implementations, nor that they are independent.

In any case my estimates are a bit more optimistic than yours, by 3-4 
months.  But it still comes down to this:

ODF 1.2 end date = remaining TC work + 2 month review + 1 month ballot + 5 
month ISO ballot =
ODF 1.1 amendment = remaining TC work + 1 month Errata for 1.1 + 2 month 
NP + 4 month FPDAM + 2 month FDAM

So if we had the texts complete today, dropped from heaven, the ODF 1.2 
process would take 8 months minimum. The amendment procedure would take 9 
months minimum.  So we're really close here in terms of expected 
completion dates. 

Of course, every week we discuss this takes us closer to the ODF 1.2 
completion, but no further to starting the amendment.  WG6 isn't able to 
even have a phone call on this topic for another 3 months.  So this does 
not look good.  Short of a general acclimation from the TC that they want 
to drop what they are doing and start work on the ODF 1.1 amendment, I 
don't think this is going to work.  From my own narrow corporate 
interests, I see far more benefit to having an OASIS ODF 1.2 standard than 
from having an ISO ODF 1.1 standard.  We are moving Lotus Symphony to ODF 
1.2 very soon.  Be sure of that.  So my attention and efforts are 
accordingly applied.  Others will need to make their own call on their 

>   3.2 I don't know how to compare an amendment timeline with that for a 
> PAS submission, but I speculate that a 1.1 amendment could be a year 
> of ODF 1.2 being approved and published at JTC1.  Maybe more, unless the 
> submission of a new version of an existing standard is very streamlined.
> Any information you or others have on that would be very helpful.

Hopefully the above numbers give you some indication. The important point 
is that ODF 1.2 gets accelerated processing via the PAS procedures, while 
an amendment is on the normal ISO slow track. 

Keep in mind JTC1 Directives 12.2.1:

"The social and economic long-term benefits of an IS should justify the 
total cost of preparing, adopting and maintaining the standard. The 
technical consideration should demonstrate that the proposed standard is 
technically feasible and timely and that it is not likely to be made 
obsolete quickly by advancing technology or to inhibit the benefits of 
technology to users."

So what if ODF 1.2 overtakes the amendment?  Isn't that a bad thing?  If 
it is, then wouldn't it also be bad if ODF 1.2 was approved only 1 month 
after the ODF 1.1 amendment was approved?  Or 2 months?  Or 3 months? 
Remember, ODF 1.1 is fully documented as an OASIS Standard.  The value in 
transposing it to an International Standard is not to benefit 
implementers.  We implementers have all the information we need.  The 
benefit of an International Standard is primarily on the adoption side. 
For this to have any real value, the standard needs to be there long 
enough for it to actually be adopted.  I've seen these processes around 
the world, and formal adoption does not happen quickly.  If you are just 
going to dangle the 1.1 standard for 6 months and then replace it with 
1.2, it will benefit no one, and may confuse many.

I'll grant you that if we had known about the possibility of an amendment 
back in 2007 this would have been the logical way forward and it would 
have had my unreserved support.  But a good idea in 2007 is not 
necessarily a workable idea in 2010. 


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]