I will
go against the flow, and back withdrawal and re-submission of v6. Once something
is adopted as an ISO standard, more people will start implementing it. We know
from the experience of XSLT and many other specifications that, once a
specification is widely implemented, it is hard to get an upgrade widely
adopted. In the XSLT case, version 2 has huge improvements over version 1, but
because Microsoft has not adopted it, there has been a split in the development
community.
I was
always a bit uncomfortable with putting v5 to ISO when we knew that it could be
improved and made easier to understand and implement. Of course, if we do decide
to withdraw and resubmit, we are back to the problem that there is probably a
solid month's work just to draft the improvements, and historically EML has been
developed and maintained by people who do not have the backing of a large
company. I, for one, have gained no business benefit from my work on EML
over the last couple of years and do not see how I would benefit in the future.
So -
can one of the companies or other organisations on the TC either take on or
support this work, or are we forced to accept second best yet
again and go for John's first option?
Regards
Paul
-----Original Message-----
From: John Borras
[mailto:]
Sent: 07 January 2009
09:41
To: EML TC; James Bryce Clark
Cc: Laurent Liscia;
Mary McRae
Subject: Re: [election-services] Progress and EML
submission status for ISO/IEC JTC1: questions for TC
First let me say many thanks to Jamie for this very full picture of where
we are and the options open to us. We clearly have a very difficult
decision to make now on how to proceed.
There is one other option open to us in addition to those put forward by
Jamie. We could withdraw our JTC1 submission, produce v6 of
EML and then submit that later in the year to JTC1, by which time hopefully
the dust has settled around the current negotiations. We know that v6
will have some significant improvements over v5 but on past experience could
take us most of this year to get it through the OASIS approval process.
That would mean no ISO approval until next year probably.
My personal goal, and I know it is also that of the UK government, has
been and still remains to get ISO accreditation as soon as possible. We
believe having that badge will help tremendously in getting
national jurisdictions and suppliers on board with
using EML, and that is no disrespect to OASIS it's just that election
officials will have heard of ISO but few have the same understanding of OASIS
and therefore are more sceptical. One other key factor from my point of
view is for us to remain in control of EML. The people on JTC1 have
little or no experience in this field and I know that for a fact from my work
with the Council of Europe committee.
So I would boil the choices for us down to two:
- do we accept Jamie's recommended approach which probably means another
2/3 months delay but has the strong possibility of eventually succeeding,
- or do we withdraw the submission, do v6 and then submit that which
means no ISO approval until next year probably.
I favour the first option and whilst that is going on we continue with
our work on the USA localisation for VIP which can then be brought into a
v6 later in the year.
Can you please let me have your views within the next few days. If
there is no common consensus I will call a TC meeting and hold a vote on the
issue.
Regards
John
M. +44 (0)7976 157745
Skype: gov3john
From: James Bryce Clark
<>
To: EML TC
<>
Cc: Laurent Liscia
<>; Mary McRae
<>
Sent: Tuesday, 6 January, 2009
19:08:15
Subject:
[election-services] Progress and EML submission status for ISO/IEC JTC1:
questions for TC
Members of the Election & Voter Services
TC:
SUMMARY: We have several options for how best to
respond,
and adapt, to significant changes in ISO/IEC
JTC1's
procedures and plans. We must make a tactical decision
about
accelerating, changing or decelerating the OASIS
submission
of EML v5.0. We would appreciate hearing the
opinions of TC
members about these issues.
As you
know, with your approval, we submitted EML v5 to ISO/IEC JTC1
late last
year for its approval as an ISO/IEC standard). JTC1 "PAS"
(Publicly
Available Specification) rules permit it to bring in
specifications
approved by accredited external groups and
essentially endorse them by a
re-approval ("transposition"). A
number of standards have completed this
process successfully, and
OASIS has been an accredited PAS submitter since
2004.
JTC1 was slow to respond to our EML submission, but is now ready
to
submit our work to its members (over 35 voting national standards
bodies) for a vote. However, at the same time, we've run into significant
changes in the expectations of JTC1 members about the way that a standard is
controlled and maintained *after* its transposition
approval. Recent events
give us some concern that the terms of submission we proposed, with your
approval, may need to be changed, in order for the vote to succeed.
Please note, the uncertainty is about the "ground rules" going forward
for further revision of the standard, and *not* any dispute about the
substance of your work today.
As described below, it's our advice that
the TC consider modifying our OASIS terms of submission, to anticipate JTC1's
changed expectations. Four options and our recommendation are described
below; we request your input.
BACKGROUND
I put a detailed
background discussion in the next, separate e-mail, so that this one focuses
on action. Here's the very short version: JTC1 is re-examining and
rewriting its own rules, about maintenance of outside submissions, very
slowly. There's been controversy about this. There's still a high
degree of mutual trust and regard between JTC1's members and OASIS. But
at the same time, they are struggling to re-define their rules for
submissions, and we don't want EML to get caught in the downdraft of
uncertainty.
Our default position is that the OASIS TC who creates a
submitted standard continues to control it, and sets its own schedule for
taking in feedback from JTC1, and bringing major completed revisions back to
JTC.
As described in the background summary post, that position has run
into some resistance. We're concerned that any new submission sent in
blindly with the same terms would be rejected, now, based on the procedural
issues. In the current standards politics environment, there are three
new factors:
* We are in mid-negotiation with JTC1 about
possible alternative collaborative maintenance methods.
*
JTC1 as a community has both an unclear, under-revision set of rules for
external submissions, and a genuine ambivalence about those
submissions.
* Finally, within OASIS, our own Board and
management are re-examining how best to handle external
submissions.
(Details in the background post.) The third point is
unlikely to affect the EML submission, as OASIS already approved it and sent
it on to JTC1 in mid-2008, before we started to re-examine our own
processes. But the first two points will weigh heavily on EML.
.
ACTIONS
It's our hope that OASIS, by being a good citizen in
this environment, will help define mutually satisfactory and constructive
solutions. But we still face the immediate question of how to handle the
EML submission. Right now, it's in JTC1's hands as we originally
submitted it, after your TC reviewed the draft, with the same basic terms as
we used for ODF v1.0. See the attached "Appendix A" document, which
comes from that submission.
OASIS has four options I can see.
Input from the TC would be helpful to us on which is best.
1. We
simply could have the submission balloted now, as is, by the new JTC1
leadership, who's apparently happy to launch it at any time. My personal
guess, but it's only a guess, is that it would attract significant opposition,
because of the unresolved issues in the JTC1 polity about maintenance.
Regardless of the work itself.
2. We could submit it now, but
after an amendment adding what we think might be acceptable terms. In
the attached copy of "Appendix A", I've inserted some possible draft
changes. These would require some commitments from the TC, as you can
see, so they would need to be acceptable to you. It has the virtue of
moving faster; and the vice of being uninformed by whatever our conversations
with JTC1 will be in the next 30 days. Personally I think option 3 is
wiser.
3. We could hold the submission back for 30 days, and see
if either our first round of conversations, in late January, or the March
conclusion mentioned in the resolution, will let us more reliably refine the
proposed changes set out in option 2 and the attached markup. In that
way, we'd be sending in something better informed by JTC1 expectations.
This is my first preference, tactically, though the views of others, including
the TC and our management and Board, also may weigh in.
4.
Finally, we could send EML somewhere other than JTC1. Frankly, choosing
another forum is a strong move. I don't think it's off the table, as
it's hard to say how the PAS process will fare over the long run.
However, OASIS probably is in the best position of any consortium, given our
good will and track record, to make PAS work. In my view we ought to do
so. Thus, turning to another authority for this work at this time
probably is premature.
Now, none of the foregoing obviate the fact that
the great work you have done, within OASIS, manifestly deserves the further
promotion and higher profile, and has been languishing for all this
"politics". I apologize again to you, for both the situation, and our
inability so far to break logjams to advance it more rapidly. Still, we try
not to let our good work be rejected or fail elsewhere. It's possible
that we were right, that only a negative outcome could have resulted from this
interregnum and re-appraisal period.
We'd appreciate feedback from the
members of the Committee on these options (and any others you see).
Among other things, the TC's views on whether the early draft of "appendix A"
revisions seems like a plausible way forward. We'd also be happy to join
you at a TC meeting if further discussion or questions would be
helpful.
Kind regards and best wishes for a successful and safe new
year. Jamie Clark
~ James Bryce Clark
~ Director of Standards
Development, OASIS
~ http://www.oasis-open.org/who/staff.php#clark
---------------------------------------------------------------------
To
unsubscribe from this mail list, you must leave the OASIS TC that
generates
this mail. Follow this link to all your TCs in OASIS at:
https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
