RE: DITA default rules

Hi all,

First and foremost: Thanks Yves.

The following enters my mind when looking at this ...

1. To me, this is an example of applying ITS to an existing vocabulary. It serves or can serve at least two purposes:

	a. checking the concepts we have for ITS (in the sense of "testing ITS")
	b. exemplifying the ITS concepts

2. The "translate" attribute seems related to our discussions about mapping/pointing. I guess the difficulty which we are becoming aware of is the following: 

	So far we have been talking about "identical semantics". We thought of identical semantics as involving same number and meaning of values in the host vocabulary as in ITS.

	However, what might get in the way are the different approaches which might be taken to compute "actual values". This computation involves questions such as "What are the defaults?" and "How do values proliferate (are inherited)?" I guess the difficulties which are created by differences in computation are one argument more in favour of having a standard like ITS.

	Currently, the only approach which I can currently see is the following: ITS mapping/pointing needs to come with a caveat along the lines of "only can be used if defaulting and inheritance of the host vocabulary are identical to that of ITS" 

3. I see a need to add some meta data categories to "its:rules"

	a. Description: prose about the rules set
	b. Host vocabulary/vocabularies: describes to which vocabulary the rules pertain
	c. ID: unique identifier for the rule set
	d. Rules set version: version identifier for the rule set

   In code this, could look as follows

	<its:rules xmlns:its="http://www.w3.org/2005/11/its">
	 <its:desc>ITS rules for TEI used by the Open University</its:desc>
	 <its:hostVoc>http://www.tei-c.org/ns/1.0</its:hostVoc>
	 <its:rulesId>98ECED99DF63D511B1250008C784EFB1</its:rulesId>
	 <its:rulesVersion>v 1.81 2006/03/28 07:43:21</its:rulesVersion>
	 ...
	</its:rules>

   The rationale for this set of meta data is quite easy: It will help administer and use rules easily (since it will be possible to check that an outdated version is used, or that a rule set for DITA is used where one for DocBook should be used.

4. I guess some people may want to combine rule sets (for example have one set for "withinText" one for "translate") in order to easily put together specific ones from "modules".

Best regards,
Christian

-----Original Message-----
From: public-i18n-its-request@w3.org [mailto:public-i18n-its-request@w3.org] On Behalf Of Yves Savourel
Sent: Freitag, 31. März 2006 05:56
To: public-i18n-its@w3.org
Subject: DITA default rules

Hi all,

I've tried to make a default ITS rules file for DITA based on the information I have so far.

I'm still struggling with how to handle the 'translate' attribute. I've asked Don and Gershon some clarifications about how DITA
users are expecting it to be handled.

Anyway, if you want to try out some DITA files, the rules are attached.

Cheers,
-yves

Received on Friday, 31 March 2006 08:17:41 UTC