QUESTION: what does "deprecated" mean?

[Lofton Henderson <lofton@rockynet.com>]

WebCGM TC,

[I'd like feedback before next telecon, so I can start to align next draft 
of text...]

We need to agree on a precise definition of "deprecated".  I took this 
action item at the editors meeting and the 7/13 telecon.

Right now, the editors-draft text says:  "For WebCGM 2.0, deprecation is 
defined as follows. Deprecated features must not be present in conforming 
2.0 content, but must be supported by conforming 2.0 viewers that support 
conforming 1.0 content."  This is approximately the definition of 
deprecation in MathML 2.0 (see end of this message).

It seems to me that the issue in the definition probably revolves around 
"must not be present in conforming 2.0 content", and that there are two 
ALTERNATIVES:

Alt.1:  Deprecated features MUST NOT be present in conforming 2.0 content.
Alt.2:  Deprecated features SHOULD NOT be present in conforming 2.0 content.

Alt.1 means that conforming authoring tools must stop generating the 
feature now, in any valid 2.0 output.  Alt.2 recommends that they stop 
generating it.  In either case, the stuff remains in the 2.0 specification, 
but with the "deprecated" label, and tied closely to our (tbd) precise 
definition of "deprecated".

RECOMMENDATION:  Alt.1.

DISCUSSION:  FYI, here is what is currently deprecated in 2.0 (quoting from 
Appendix A.2):

>-- TILE compression types 0, 1, 2;
>-- BITONAL TILE compression types 0, 1;
>-- the three WebCGM 1.0 object behaviors in the URI fragment -- highlight, 
>view_context, highlight_all -- are deprecated and replaced by a set of 
>behaviors that are atomic, orthogonal and comprehensive in combination;
>-- the 'viewport' param element, unused and unseen in the five years 
>between WebCGM 1.0 and 2.0.
>-- the WebCGM 1.0 Character Set List designation tails for UTF-8 and 
>UTF-16 are replaced by correct derivations in WebCGM 2.0, and the 1.0 
>forms are deprecated.
>
>[new LH Note.  what about picBehav term in fragment of 'linkuri'? It has 
>been called deprecated, in the 'linkuri' description, for some time now, 
>e.g. since 2nd Release of 1.0 (2001).]

Looking at the list, I think the questions will be about the compression 
types.  Do we mean to say that you MUST NOT generate in 2.0 metafiles, or 
SHOULD NOT generate?  Thoughts?

SIDE QUESTION.  Also, what about that ancient "deprecated" wording in 
'linkuri', about the picBehavior term of the fragment.  Thoughts?  (Do we 
want it to be legal but discouraged as of now, or do we want it to become 
illegal?)

MathML 2.0 Deprecation Definition
==========

>7.2.1.2 Deprecated MathML 1.x Features
>
>MathML 2.0 contains a number of MathML 1.x features which are now 
>deprecated. The following points define what it means for a feature to be 
>deprecated, and clarify the relation between deprecated features and 
>MathML 2.0 conformance.
>
>    1.  In order to be MathML-output-conformant, authoring tools may not 
> generate MathML markup containing deprecated features.
>    2.  In order to be MathML-input-conformant, rendering/reading tools 
> must support deprecated features if they are to be in conformance with 
> MathML 1.x. They do not have to support deprecated features to be 
> considered in conformance with MathML 2.0. However, all tools are 
> encouraged to support the old forms as much as possible.

Regards,
-Lofton.

[1] http://www.w3.org/TR/MathML/chapter7.html#interf.deprec