W3C home > Mailing lists > Public > public-multilingualweb-lt-comments@w3.org > April 2013

[ISSUE-86][ACTION-408] normative text in notes and formatting normative keywords

From: Dave Lewis <dave.lewis@cs.tcd.ie>
Date: Tue, 02 Apr 2013 15:45:48 +0100
Message-ID: <515AEF1C.3050708@cs.tcd.ie>
To: public-multilingualweb-lt-comments@w3.org
Hi
Issue-86 started with a comment from Richard on normative key word in 
notes, and Felix had an action, also passed to me to check for normative 
text in notes:
https://www.w3.org/International/multilingualweb/lt/track/actions/408

Felix already checked there was no rfc2119 annotated text in the notes, 
which there wasn't. However should we also check for and address 
unadorned usage of these keyword and change the text correspondingly for 
removal of doubt?

I found the following occurrences of 'must' in notes. Should we change 
these instances to 'ought to'?

There are also many instances of the use of 'should' and 'recommend' in 
notes, but i suspect this isn't likely to cause an issue in the same way 
as 'musts':


In section 6.2; Global Rules
http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#html5-global-rules

*
"Note:*

HTML5 parsing algorithm automatically puts all HTML elements into XHTML 
namespace (|http://www.w3.org/1999/xhtml|). Selectors used in global 
rules _must_ take this into account."

*In section 8.2.2
"Note:*

It is not possible to override theTranslate 
<http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#trans-datacat>data 
category settings of attributes using local markup. This limitation is 
consistent with the advised practice of not using translatable 
attributes. If attributes need to be translatable (e.g., an 
HTML|alt|attribute), then this _must_ be declared globally."


        In section 8.14.2 Implementation

*"Note:*

The source node and the target node may be of different types, but the 
target node _must_ be able to contain the same content of the source 
node (e.g. an attribute node cannot be the target node of a source node 
that is an element with children)."

----- end of note topic

While doing this I also found the following occurances of unadorned 
rfc2119 keywords (underlined below) that seemed to be used in a 
normative way in normative sections. Editorially, should these be made 
rfc2119 annotated 'MUST' etc?

Also, section 6 (using ITS markup in HTML) and 7 (Using ITS Markup in 
XHTML) are marked as neither normative or informative (I presume it is 
the former).


      3.1 Notation

It is _recommended_ that XML implementations of this specification use 
this prefix, unless there is existing dedicated markup in use for a 
given data category.


          5.3.2.1 Absolute selector

The absolute selectorMUST 
<http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#rfc-keywords>be 
an XPath expression which starts with "|/|". That is, it _must_ be 
anAbsoluteLocationPath 
<http://www.w3.org/TR/xpath/#NT-AbsoluteLocationPath>or union 
ofAbsoluteLocationPath 
<http://www.w3.org/TR/xpath/#NT-AbsoluteLocationPath>s as described 
inXPath 1.0 
<http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#xpath>.


        5.3.5 Variables in selectors


        The|param|element has a _required _name attribute.


        5.4 Link to External Rules

One way to associate a document with a set of external ITS rules is to 
use the optional XLink[XLink 1.1] 
<http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#xlink1>|href|attribute 
in the|rules|element. The referenced document _must_ be a valid XML 
document containing at most one|rules|element.


      5.8 ITS Tools Annotation

has three unadorned '_shoulds_'


        8.3.1 Definition

Two types of informative notes are needed:

  *

    An alert contains information that the translator _must_ read before
    translating a piece of text.


        8.17.1 Definition

'_recommended_' in notes colum for profile refernece attribute


cheers,
Dave
Received on Tuesday, 2 April 2013 14:43:08 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 2 April 2013 14:43:09 UTC