Re: Comments on XML Protocol Reqs

On Wed, Dec 20, 2000, Henrik Frystyk Nielsen wrote:
> > R700b: This seems to be thinking only about indepenent extensions.
> >      Despite trying hard, in many cases extensions are not 
> > independent.
> >      It should be possible to express e.g. that the message is
> >      successful if either extensions A and B OR extensions C and D
> >      are known. This can not be expressed with a simple mandatory/
> >      optional distinction.
> >      The requirement should make clear what combinations are
> >      needed, and should include combinations as described above.
> 
> The problem with negotiation is that it can only be done within a
> context. In other words, in order to negotiate, you need to know what
> you are negotiating. As the purpose of XP is to *not* know about any
> particular area (which can be expressed as an XP module) all we need is
> a single optional/mandatory bit. That allows for any negotiation, logic
> language or dependency language to be expressed in a layer or layers
> above (neither of these are described by the charter btw.). That is, by
> having a single bit XP can support the model you describe but XP doesn't
> have to define it.

I am not sure that it is really a negotiation problem. I do not see
much difference between "must deal with A" and "must deal with A or
B".

I think that it is more a simplicity problem. An optional/mandatory
bit is easy to define; a combination of and's and or's is more complex
to represent and process.

-- 
Hugo Haas - W3C/MIT
mailto:hugo@w3.org - http://www.w3.org/People/Hugo/ - tel:+1-617-452-2092

Received on Friday, 5 January 2001 08:36:13 UTC