Re: [oiic-formation-discuss] Proposed Use case -- Interoperability in vertical and horizontal ODF markets

[jose lorenzo <hozelda@yahoo.com>]

--- On Sat, 6/28/08, Simon Calderson <caldersons@yahoo.co.uk> wrote:

> From: Simon Calderson <caldersons@yahoo.co.uk>
> Subject: Re: [oiic-formation-discuss] Proposed Use case -- Interoperability in vertical and horizontal ODF markets
> To: oiic-formation-discuss@lists.oasis-open.org
> Date: Saturday, June 28, 2008, 9:45 AM
> marbux <marbux@gmail.com> wrote:
> > I would appreciate a link to the relevant discussion
> because because
> > any such argument that my proposals have been out of
> scope is plainly
> > erronenous as a matter of law.
> 
> Firstly, you're misrepresenting what I said. I said
> legal discussion of your proposals was out of scope, which
> it clearly is because no-one here apart from Andy is in a
> position to give legal advice, including yourself. Where
> legal advice is required, it is given by OASIS, viz.:
> 
>    
> http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00660.html
> 
> It is not for anyone else on this list to argue legal
> points or reach legal decisions. Of course, your proposals
> themselves are not out of scope, to the extent to which
> they address a potential charter for an interoperability TC
> for ODF (which they frequently don't).
> 
> > >Your use case's technical merits have also
> been discussed, time and again.
> > 
> > Not in the thread in which I made the use case
> proposal, this thread,
> > other than my own discussion of it.
> 
> Yes, if we limit our scope to a two-day old discussion of
> you beating the same horse again, sure, it hasn't been
> discussed. However, in reality, it has already been
> discussed previously, which is likely why no-one is
> addressing the points you raise: because you've raised
> this canards before.
> 
> For example, you stated your use-case as:
> 
>     "How may how may(sic) all conformant ODF
> implementations hold 
>     two-way conversations with each other without data
> loss?"
> My emphasis. Back on the 18th June, I had already raised
> this point to the list:
> 
>     "the utility here is being able to say "you
> don't have to implement all of 
>     ODF; you could implement these small parts, and
> here's how your app 
>     can deal with the bits it doesn't know without
> trashing the document". 
>     Surely that's what we want the TC to
> enable[..]?"
> 
>  
> http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00434.html
> 
> So your use case has in fact been recognized and discussed
> previously.
> 
> The technical merits of your proposal have also been
> discussed, time 
> and again. You said in your e-mail starting this thread:
> 
>     "I point out that my CDRF proposal, as elaborated
> to
>     include a compatibility framework, was carefully
> designed to solve
>     that issue [..]"You keep bringing up CDRF, but you
> ignore all of the basic technical questions that your
> proposal doesn't address, which have been put directly
> to you many times, amongst others:
> 
>    
> http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00465.html
> (19th June)
>    
> http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00538.html
> (20th June)
> 
> I repeat again: CDRF does not address your use-case, it was
> not designed to do so nor does it in actuality, and indeed
> the claims you make for it are wild, speculative and do not
> match the standard as published.
> 
> > My use case's technical merits have not been
> discussed in the thread
> > in which I made the proposal.
> 
> That's irrelevant since they have already been
> addressed before and you haven't answered. I'm
> certainly not going to bother reading further proposals
> from you: this mailing list is *not* a write-only medium
> for you. If you start reading and answering, I will then
> think about spending time looking over your latest
> propoals.
> 
> As you will note, I have specified the precise mails as
> they are in the record of this mailing list. There are many
> other such occasions, but I feel starting at the basic
> points you've missed would not be a bad idea.

I read the two links you provided.

In the first, you seem to have missed entirely the "bullet point": the concept of having superprofiles not contaminate files of subprofiles. This helps towards making round-tripping possible.

In the second, you recognize that concept but think it doesn't apply to interop and interop is about so much more. Well, in practice interop has a lot to do with testing at various levels. But the original poster is speaking about first getting the spec correct and without leaving room for variation. If the law says, "you may kill (or you may not kill)," then, sure, some that don't want to kill will misimplement and others that want to kill will misimplement, but the fact remains that with that law everything is possible and legal. Interop in that world is NOT possible except by accident (all that want to kill misimplement while all that don't want to kill implement correctly).

The OP is talking about getting the rules right firstly. That is one step before worrying about getting the implementations right. I have to believe you missed something here since I can't imagine anyone would think that the wording of the rules has nothing to do with getting interop right.

So as to try to avoid being misinterpreted and getting unwarranted flame, I am not disagreeing or agreeing with you or with the OP on all counts, but I did want to provide some sort of answer to you in case you still didn't much know what the OP (marbux) was talking about. [apologies if I misinterpreted your view]