PERSONAL, WAS Re: [office-comment] xml:ids discardable - what!? (ODF 1.2CD01)

[marbux <marbux@gmail.com>]

On Sat, Mar 7, 2009 at 2:01 AM, Alex Brown <alexb@griffinbrown.co.uk> wrote:
> Dear all,
>
> In 18.920 we learn:
>
> ----b
> All implementations that are OpenDocument consumers and producers should preserve any xml:id attribute and its value when present on any of the following elements [list follows]
> ----e
>
> Why only "should" preserve?!
>
> The implication of this is that processors are free to strip or alter the IDs within an ODF document.
>
> Since many internal references (and, in any serious workflow) many external references will rely on ID values for integrity this behaviour is fatal for any sane notion of document processing.
>
> PROPOSAL. xml:id values, as the primary means of identifying elements in ODF documents, must be respected and preserved by ODF processors, and this needs to be made clear in the ODF spec.

Bingo! You just hit the crowning blow that caused Gary Edwards and me
to leave the ODF TC. For future reference (i.e., at SC 34), there was
a huge flap over this on both the Metadata SC and the ODF TC. The
change was made deliberately on a "let the market decide" and
"flexibility for unforeseen problems" basis.

I had persuaded the Metadata SC to add language making the
preservation of xml:id values mandatory. A few days before the SC's
work product was to be sent to the TC for approval, Sun blew in with
an 11th-hour proposal to change "shall preserve" to "should preserve"
throughout the SC's work.

Can't find my way back to the Sun proposal at the moment. Part of it
is quoted in the following linked page.
<http://lists.oasis-open.org/archives/office/200706/msg00125.html>.
Noteworthy in that post is Bruce d'Arcus' statement that:

"Really, just to be clear: if applications do not preserve xml:id
attributes, fields will break, and any metadata about document
fragments will be made invalid. Is that really in anyone's interest?
They need not support metadata in any explicit way to do this."

Bruce later buckled to the pressure. My suspicion is that Sun told him
he wouldn't get his bibliography code into OOo 3.0 unless he backed
off. (Bruce is an academic but is the volunteer head of the OOo
bibliograpic project.)

The real story behind this is that there are only two ways Microsoft
can bridge the round-trip gap between Office and OOo via ODF without
unacceptable data loss, via the RDFa support or the foreign elements
and attributes. Under either route, OOo must be programmed not to
trash the Office metadata that ODF does not support. Lots more to the
story, but the allowed destruction  of xml:id attributes was rather
obviously intended to block the OpenDocument Foundation's ODF plug-ins
for Office from achieving high fidelity round-tripping between Office
and OOo.

Best regards,

Paul


-- 
Universal Interoperability Council
<http:www.universal-interop-council.org>