Re: [oiic-formation-discuss] My perspective. display perferct?

["Sam Johnston" <samj@samj.net>]

On Tue, Jul 1, 2008 at 3:51 PM, <robert_weir@us.ibm.com> wrote:

> On Tue, Jul 1, 2008 at 7:33 AM, Dave Pawson <dave.pawson@gmail.com> wrote:
> I think I'm doing a Rob Weir.
>
> Proposal for consensus: Please respond if you support it.
>
> "There will be no deliverable addressing the ideas of 'pixel perfect' as
> used on this list. "
>
> Please respond or you'll have one mans version of consensus.
>
> worksforme (ODF != PDF)

I would not put a restriction in the charter against such a deliverable.  But I'd agree that we have no basis for including in such a deliverable at present.

I think it was Dave's intention to get some consensus around the 'pixel perfect' discussion, in that it would not be required rather than specifically excluded (at least it was mine).

I think we will want the charter will be written broadly enough to cover current and future versions of ODF.  We cannot know what features will be in, say ODF 2.0.  Nor can we know what profiles the proposed TC will create.  For example, an ODF/Web profile, if it maps a subset of ODF primitives to XHTML/CSS2, can make rendering and fidelity guarantees which equal that of XHTML/CSS2.

One of the first tasks of the TC should be to determine use cases and user demand; a list of ODF directives ordered by user priority would be very useful for implementors and profile writers if nothing else.

My larger concern is that by not talking about the version at all the TC could be subverted by working either too far in the past or too far in the future, and while I'm not a subscriber to conspiracy theories I am well aware that the requirements of the various participants in terms of interoperability vary wildly (eg users vs implementors).

I've talked about this before (and even proposed some simple 'rules') but provided the TC is pointed at the 'most current available version' or similar I'll likely be satisfied. Note that working where the current activity is (likely 1.2, even if unreleased) will almost certainly give the best results; making recommendations about 1.0 could perhaps be equated to closing the barn door after the horse has bolted.

Sam