Re: [ACTION-527]: Propose spec change to address extensibility in general

Am 16.05.13 15:25, schrieb Dr. David Filip:
> Inline:
>
> On Thu, May 16, 2013 at 2:13 PM, Felix Sasaki <fsasaki@w3.org 
> <mailto:fsasaki@w3.org>> wrote:
>
>     Am 16.05.13 15:00, schrieb Dr. David Filip:
>>     On Thu, May 16, 2013 at 1:01 PM, Yves Savourel
>>     <ysavourel@enlaso.com <mailto:ysavourel@enlaso.com>> wrote:
>>
>>         Hi David, all,
>>
>>         > [[
>>         > 2-5: Non-ITS elements or attributes found in a rules
>>         element MAY be ignored.
>>         > ]]
>>         > ...
>>         > Besides we allowed foreign attributes in all elements, so this
>>         > should say that non-ITS elements or attributes in all ITS
>>         elements MAY be ignored.
>>
>>         Actually the schema change allows foreign elements only in
>>         its:rules.
>>         But in any case, for this statement, there is no reason to
>>         mention where the non-ITS elements/attributes can be found:
>>         It's just about what to do with them. So we should use a
>>         blanket statement:
>>
>>         [[
>>         2-5: Non-ITS elements and attributes MAY be ignored.
>>         ]]
>>
>>         or
>>
>>         [[
>>         2-5: Non-ITS elements and attributes found in ITS elements
>>         MAY be ignored.
>>         ]]
>>
>>     I like this one
>
>     OK, so we have now three changes that you agreed upon:
>     - schema change that Jirka committed
>     - 2-5 above
>

I did that change too now in the draft.

>     - the text in appendix D "Foreign elements can be used only inside
>     rules. Foreign attributes can be used on any element defined in ITS."
>
>
>>         And the locations where non-ITS elements/attributes are
>>         allowed can be described in
>>         http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#its-schemas
>>         as Felix proposed.
>>
>>         [[
>>         Foreign elements can be used only inside rules.
>>         Foreign attributes can be used on any element defined in ITS.
>>         ]]
>>
>>     It is good to have this text in appendix D but not sufficient IMHO
>>
>>     I understand that making a general blanket statements on one
>>     place is convenient when writing, but is not consumption friendly.
>>     The elements are specified throughout all categories and people
>>     who read only on the categories they are interested in, are
>>     likely to miss the extensibility.
>>
>>     An important question
>>     Is appendix D normative?
>
>     Yes - see the TOC
>     http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html
>     all other appendicies say "non-normative".
>
>
>>     I am not sure (it is not marked non-normative but it also does
>>     not explicitly say that it is normative, as we do elsewhere in
>>     the spec like this:
>>     /This section is normative./
>>
>>
>>     On the assumption that the appendix D is normative (perhaps we
>>     should add /This section is normative./)
>
>     Done for all normative appendices a-d.
>
>
> Perfect
>
>
>
>>     , I am happy to keep the requirement for specifying extensibility
>>     on every element for the second last call, because then it will
>>     be indeed editorial only. provided we have the NORMATIVE appendix
>>     D blanket statement.
>
>     We have that now.
>
>
>>
>>     This is an example to make sure that we are on the same page:
>>
>>     [This is the current text]
>>
>>     GLOBAL: The |termRule| element contains the following:
>>
>>      *
>>
>>         A required |selector| attribute. It contains an absolute
>>         selector
>>         <http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#selectors> which
>>         selects the nodes to which this rule applies.
>>
>>      *
>>
>>         A required |term| attribute with the value "yes" or "no".
>>
>>      *
>>
>>         None or exactly one of the following:
>>
>>          o
>>
>>             A |termInfoPointer| attribute that contains a relative
>>             selector
>>             <http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#selectors> pointing
>>             to a node that holds the terminology information.
>>
>>          o
>>
>>             A |termInfoRef| attribute that contains an IRI referring
>>             to the resource providing information about the term.
>>
>>          o
>>
>>             A |termInfoRefPointer| attribute that contains a relative
>>             selector
>>             <http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#selectors> pointing
>>             to a node that holds the IRI referring to the location of
>>             the terminology information.
>>
>>
>>     [This is the addition I require]
>>
>>       * None, one or more attributes from any namespace.
>>
>
>
>     I disagree with adding the same text to all sections. It makes the
>     spec longer, nothing else. Anyway, you said that "I am happy to
>     keep the requirement for specifying extensibility on every element
>     for the second last call". Could you confirm that the current
>     state resolves this issue? Note that "2-5: Non-ITS elements and
>     attributes found in ITS elements MAY be ignored." is not in the
>     spec yet, but once you agree I add it. Others, if you disagree
>     with adding it please shout.
>
> *As I explained in the other response, this list is ENUMERATIV and in 
> main part of the normative prose text. So the spec is currently in 
> conflict with the schema and appendix D.*
> *It is not optional to specify or not extensibility in an enumerativ 
> list,.*
> *If it is not there it simply means it is not allowed.*
> *
> *

I disagree, see at
http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013May/0172.html
"it" referring to the statement about elements and attributes.
[
Well, we can also put it on the beginning of
http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#datacategory-description
and say "This statement relates to all elements defined in this
specification". In that way it "inherits" to all definitions. Is there a
problem with this?
]

But as I understand you above the issue is now resolved and you will 
open another issue during last call, if needed, right?

Best,

Felix
> Sloppy as it is, I can wait with bringing the elements up to date as 
> an editorial CORRECTION in the second last call, if it seems such a 
> big issue to introduce now.



> *
> *
>
>
>     Best,
>
>     Felix
>
>
>>
>>     Section 5.2.1 should also contain the provision for elements and
>>     attributes from any namespace within <its:rules>
>>
>>         -ys
>>
>>
>>
>
>

Received on Thursday, 16 May 2013 14:06:37 UTC