RE: WG assignments -- WebCGM 1.0 Errata review

Dieter, All --

I'm a little shaky on the answer to Dieter's question, so I'd be happy to 
have someone correct it or supplement it...

At 09:10 AM 9/11/2007 -0400, Weidenbrueck, Dieter wrote:
>[...]
>E02 and E03 look good to me.
>
>One question:
>
><OBJECT DATA="xxx.cgm" TYPE="image/cgm;Version=4;ProfileId=WebCGM"
>WIDTH="200" HEIGHT="100" >
>
>This is the correct way to specify an OBJECT. What happens in the case
>of a cascading profile?
>The ProfileID=WebCGM would be wrong for an ATA or S1000D profile.
>Do we actually need the profile ID here? This is probably confusing
>anyway, because the integrator has no clue as to which files he will
>display in the viewer at the time he has to write this.

Good question.  I'll try to answer from memory, without going to research 
it too much.  Long story short, I think we made a mistake when we 
registered the MIME type a dozen years ago.

The value of the <OBJECT> 'TYPE' attribute is supposed to correspond to the 
registered MIME type.  From HTML 4.01:

[[[
This attribute specifies the content type for the data specified by data. 
This attribute is optional but recommended when data is specified since it 
allows the user agent to avoid loading information for unsupported content 
types. If the value of this attribute differs from the HTTP Content-Type 
returned by the server when the object is retrieved, the HTTP Content-Type 
takes precedence.
]]]

When cgm was registered as a MIME type in 1995, Brian DiAntonio from 
Carberry Technology was the person who (apparently) knew about that stuff 
and drove the registration, participating in the ATA GWG.  Chris was 
involved as well (I was fairly clueless about the topic at that time -- 
still am).

So here is the registration:

http://www.iana.org/assignments/media-types/image/cgm

Notice that the parameters are all "required", none of the parameters is 
"optional".  Chris has commented subsequently that he thinks it could have 
been done better, and I guess I agree.  It seems to me that "image/cgm" is 
all that is really needed (the other info is inside the metafile anyway).

After all, this is about content-type for transfer, right?  The things like 
Version and ProfileId are pretty irrelevant in that context, it seems to 
me.  On the other hand, after transfer, viewers will find and handle 
Version and ProfileId appropriately within the metafile, as they start to 
process the metafile.  (Question.  Does the viewer ever even see this TYPE 
attribute?)

So ... back to your question.  Yeah, technically speaking, a cascaded 
S1000D profile probably wants to change the ProfileId there, and ATA similarly.

What confuses me is this:  what component of the whole Web-distributed 
system -- servers/clients and HTML content/CGM content -- actually sees and 
does anything useful with the ProfileId attribute within the 
MIME-registered content-type?  My impression is that it doesn't actually 
matter in a substantive way to *any* system component.

Thoughts?

-Lofton.

>-----Original Message-----
>From: public-webcgm-wg-request@w3.org
>[mailto:public-webcgm-wg-request@w3.org] On Behalf Of Don Larson
>Sent: Dienstag, 11. September 2007 00:39
>To: Lofton Henderson; WebCGM WG
>Subject: re: WG assignments -- WebCGM 1.0 Errata review
>
>
>Lofton,
>
>I have reviewed E10 & E11 and it looks okay as written.
>
>Don.
>
>  >  Hi All,
>
>  >  I want to make a small assignment for each WG participant, to be
>completed  >  before the Thursday telecon.
>
>  >  I realize that the 1.0 errata document [1] has grown to be bulky,
>and it  >  may be daunting if you're busy.  Therefore, I'd like to have
>each person  >  review two errata, okay?
>
>  >  To remind you of present status... The WebCGM WG approved the
>substance  >  (and Class) of all of the errata solutions at the 8/30
>telecon [2] (and the
>  >
>  >  WebCGM TC approved the same substance earlier at its f2f).
>
>  >  So your job is to check that the completed text in the document [1]
> >  correctly implements our approved resolutions, and to check for
>editorial  >  errors.
>
>  >  Chris -- E01 (URI/IRI)
>  >  Dieter -- E02, E03
>  >  Benoit -- E04, E05
>  >  Dave -- E06, E07
>  >  Stuart -- E08, E09
>  >  Don -- E10, E11
>
>  >  It would be good to email your review ("okay" or corrections),
>especially  >  if you will not be present Thursday.  Feel free to review
>other ones in  >  addition to your assigment.
>
>  >  Thanks in advance,
>  >  -Lofton.
>
>  >  At 04:09 PM 9/6/2007 -0600, Lofton Henderson wrote:
>
>  >  >WebCGM WG --
>  >  >
>  >  >This should implement all of the decisions that we made at the 8/30
> >  >telecon (and before):
>  >
> >http://www.w3.org/Graphics/WebCGM/WG/2007/errata-10/WebCGM10-errata-200
>706
>  >  21.html
>  >  >
>  >  >It is ready for final editorial approval, which we will do at the
>9/13  >  >telecon.  Please have a careful look at it before then, and
>send any  >  >comments to this list.  Are there any substantial errors
>in implementing  >  >our decisions?  Any editorial corrections?
>  >  >
>  >  >After approval, it will have a 4-week (public) review, as it
>contains a  >  >couple of Class 3 corrections.
>  >  >
>  >  >Regards,
>  >  >-Lofton.
>  >  >
>  >  >
>  >  >
>  >  >
>  >  >