before answering to one of Emmanuels mails, just let me summarize the
discussion so far:
o Obviously, it was overseen during the standardization process that
Ingredient objects without content data could be targets of the
Preload/Unload actions.
o So far we have the common understanding that it would be useful to
target these actions to Ingredient objects without content data.
We have to be clear about whether whatever we propose for changing the
description, Provision Of Use or adding a note will break existing
applications.
Right?
On Fri, 16 Jan 1998, Emmanuel BEUQUE wrote:
> À (At) 8:53 +0200 16/01/98, Grassel Guido NRC/Hki écrivait (wrote) :
> >I think we should be careful here and avoid to wide interpreations of the
> >text.
> >Remember some time ago I asked a question about when the ContentAvailbale
> >event is generated. The answer I got, I think from Andreas, was just arguing
> >the other way round, i.e. if an object has no content data that part of
> >Preparation Behaviour of Root class would not have any effect.
Right, and we should add a note for this, too.
> >To my understanding the MHEG-5 maintence process should write down
> >clarifications for those clauses where there is clear evidence that some
> >reader might interprete the existing standard in the wrong way.
Yes.
> 1. Regarding the provision of use that embarassed Andreas, I'm just asking
> how the content attribute of an Ingredient could be Null. It seems to me
> that any concrete class that inherits from the Ingredient Class either
> defines the Content attribute as mandatory or not to be encoded at all. If
> the content is mandatory, I don't see how it could be Null. Neither the
> origContent attribute nor the setData elementary action accept Null, if I'm
> right. In case the content shall not be encoded, it just concerns my second
> question.
The description of the OriginalContent states:
"...OriginalContent is an optional attribute...Default Value: none"
and the description of Content states:
"...The Content attribute shall not be defined for classes specifying that
OriginalContent shall not be encoded.
...Optional Attribute...Initial Value: Value of OriginalContent"
(BTW, another interesting point is the descrepance between "Optional" and
"Initial Value" :-)
I think some descriptions which says something about a "Null" value (like the
one of the Preload action) makes a (IMHO wrong) assumtion about implementation
issues. But it always means "is not encoded/has no value".
Maybe this is another issue for the MTF to check the whole document for this
minor flaws (or at least add a description to chapter 3, "Terms and
Definitions").
> 2. How the Null value shall be interpreted ? Should it be considered as a
> specific value or as no value at all ? In other words, is it equivalent to
> "void" ?
"void" is a usually a data type. I like the Lisp-style "nil" better.
> In the specific case of Ingredient class content attribute, the
> default value is "none". Should we considered that the default value (no
> value) is equivalent to Null or shall an MHEG engine consider that no value
> is different from Null ? I must admit that this is not completely clear in
> my mind.
As I said above, "Null" and "none" are meant to be equivalent.
Best regards,
Andreas
-- o _ Andreas Kraft (\_|_) GMD FOKUS, kraft@fokus.gmd.de, +49 30 3463-7232 T> ] [ The sky is the limit