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


Help: OASIS Mailing Lists Help | MarkMail Help

office-comment message

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

Subject: RE: [opendocument-users] RE: office-comment text:id vs xml:id (ODF 1.2CD01)

Dear all,

[moving this thread back to office-comment, as was indicated is appropriate]

> -----Original Message-----
> From: Dennis E. Hamilton [mailto:dennis.hamilton@acm.org]
> Sent: 08 March 2009 22:30
> To: Alex Brown; 'ODF Users List'
> Cc: 'Makoto MURATA'; 'marbux'
> Subject: RE: [opendocument-users] RE: office-comment text:id vs xml:id
> (ODF 1.2CD01)
> Hello Alex,
> Thank you.  This exchange has been very useful for improving my
> understanding of what is involved, and how experts such as yourself see
> it.
> 1. I have one strong reservation.
>    1.1 I don't think making a clean break, or throwing a big switch on
> the ODF format and schema is a good idea.  This might work for an ODF
> 2.0, but I don't think it is sensible for a 1.x dot-release.
>    1.2 I am also concerned that there is already a legacy of ODF
> documents and processors (whatever their conformance might be) and
> there needs to be some responsibility taken how we preserve those
> documents and interoperability with those processors.  We are long past
> the era when people thought they could throw a big switch on
> information technology (although, ok, bureaucrats still try it).

I'm inclining the other way on this. I hear what you are saying, but think you are basing your thinking on the false assumption that there is a consistent and preservable corpus of ODF documents out there.

Grated, for any one implementation series (like OO.o) this may be the case, but between implementations, the chaos we have noted makes this a near impossibility.

There is also the JTC 1 Factor. I note that the TC considers this a "1.x dot-release", but that kind of constraint does not and cannot feature in a PAS review. In fact telling NBs there are things they can't do is not tactically very wise. Ask Ecma. I also note there are even some members of the TC who expressed horror at the thought that standards work should be constrained by what actually existed!

What I think might happen to ODF if it tried to make it through JTC 1 with embedded "deprecated" features, is what happened to OOXML. A new, separate "transitional" conformance class will be created where all these crappy features can be dumped. Existing implementations can remain conformant to that, but the brave new world (and maybe basis for ODF-Next) will be a new, strict conformance class. To save time, I think the ODF should bite the bullet and do this now. It is the right thing to do, and as engineers I think the members should feel good about doing this properly rather than perpetuating follies in the main body of the spec.

Implementations can up-convert old ODF files on load, based on however they happened to interpret the old spec. By re-establishing a meaningful baseline for IDs we are aiding interoperability, I believe, not hindering it!
> 2. I also don't think we have broken upward compatibility yet, although
> we do need to seriously look at affirming our commitment to not doing
> that and executing accordingly, no matter how much cd01 has breaking
> changes at the moment.

Isn't the advent of OpenFormula one huge breaking change (and in a good way)?

(stuff about MCE snipped, as I haven't thought about that much yet!)

- Alex.

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