[dsdl-discuss] Re: Strawman: bringing the framework inside schemas

Eric

> Attached is a new version of my strawman including:
>
> - an example of implementation of Rick's unit/number proposal
> - an example using STX
> - some first considerations on implementation (I think that it
> fits nicely with James' derivative algorithm.
>
Among the problems:

You can''t use integer to validate dates - you have to use specific
datatypes, with specific ranges. The key factor is that the ranges interact,
so if the month is 2 then you can't have 30 or 31 in your day field, and you
can only have 29 if the year is divisible by 4 but not by 100, etc.

You need to clearly indicate which part of DSDL each of the unqualified
elements belongs to. (I presume the ie: entries in the Intro equate to the
if: ones later in the spec, but am not sure where many of the others come
from.) Where do fragmentRules, applyTo and produce all sit in the DSDL
scheme of things?

The whole thing starts to look messy once you let any part of any standard
nest. It would be neater to have the definitions in separate documents which
get referenced where needed. This would seem to fit more neatly with OO
principles.

Martin

--
DSDL members discussion list
To unsubscribe, please send a message with the
command  "unsubscribe" to dsdl-discuss-request@dsdl.org
(mailto:dsdl-discuss-request@dsdl.org?Subject=unsubscribe)