Is the BNF still the spec? The steps to produce are rather manual and mention specific tools (and presumably their behaviour), which makes me uncomfortable... the nice thing about the BNF is that it is textually small enough and in a form which can be analysed and reasoned about, and one can prove that the language has certain properties. This is, needless to say, not feasible with an XML schema, which is why those are generated rather than written. It is desirable to have a formal spec (the BNF + docs) and a set of rules for generating the schema, and I think they (the rules) should be independent of any specific tools.
Of course, it is good to provide hints/tips about how to do it with actual tools too :-)
Pat
On Wednesday 3 March 2004 13:01, Wil O'Mullane wrote:
> Apart from the discussion of cardinality of AND
> I have not heard major complaints about this last effort.
> May we assume this i generally accpetable.
> Or should we stick witht the previous verbose XSD ?
> (I presume not).
>
> I would rather not change 0.7.3 - I think it is tidy and
> has addressed most issues raised earlier.
>
> Vivek has posted steps on its production.
>
> I would like to get on with updateing the paper specification.
>
> Tony you in particular said you wanted the XSD tied up so we can move
> forward. I think Martin our biggest oponent is on board with this
> version. SO may we agree this is good for the rest of this year ?
>
> wil
-- Patrick Dowler Tel/Tél: (250) 363-6914 | fax/télécopieur: (250) 363-0045 Canadian Astronomy Data Centre | Centre canadien de donnees astronomiques National Research Council Canada | Conseil national de recherches Canada Government of Canada | Gouvernement du Canada 5071 West Saanich Road | 5071, chemin West Saanich Victoria, BC | Victoria (C.-B.)Received on 2004-03-03Z22:31:01