Good news about CFR (aka ISO/IEC 14496-28:2012)

[Ken Lunde]

Vladimir,

Good catch. Although someone at Apple, such as Tony Tseung, should confirm, I believe that the text in Section 5.5, along with the DTD, needs to be changed via a corrigendum to allow both elements, 'ComponentDef' and 'LanguagePreferredList', as children of the 'Components' element.

For the text in Section 5.5, simply removing the word "either" should suffice:

> Required. The 'Components' element contains one or more of the 'ComponentDef' or 'LanguagePreferredList' elements.

For the DTD, the 'Components' element should be changed to the following, which should allow what is intended, meaning that it must have children, and that both elements are allowed:

> <!ELEMENT Components (ComponentDef*,LanguagePreferredList*)+>

Someone with more experience than me with XML should confirm that the above reflects the intention.

Perhaps Dave Pawson can regenerate the RelaxNG schema based on this DTD change.

Thank you for finding this, and for bringing it to our attention.

Regards...

-- Ken

On Aug 6, 2012, at 10:50 AM, "Levantovsky, Vladimir" <Vladimir.Levantovsky at MonotypeImaging.com> wrote:

> Hi Ken, all,
> 
> I was looking at the text of the CFR (ISO/IEC 14496-28) standard and I noticed something that I believe would require a corrigendum.
> 
> Both the verbal description of the 'Components' element and the DTD of it in Annex A define that the element "contains one or more of either the 'ComponentDef' or 'LanguagePreferredList' elements", which imply an "exclusive OR" relationship between two types of children - i.e., either one or more of 'ComponentDef', or one or more of 'LanguagePreferredList' elements must be present, but both types cannot be present at the same time. 
> 
> However, in Annex B, the examples 3 and 4  show 'Components' elements that contain both 'ComponentDef' and 'LanguagePreferredList' elements as children of 'Components'. Since the examples and the spec proper are clearly at odds with each other, something must be wrong and needs to be corrected. If both 'ComponentDef' and 'LanfguagePreferredList' elements *can* be present as children of 'Components' at the same time, then we need to update the spec proper and the DTD. The main question here is what's the intended element structure and, most importantly, how it is currently implemented.
> 
> Thank you,
> Vlad
> 
> 
> 
>> -----Original Message-----
>> From: mpeg-OTspec at yahoogroups.com [mailto:mpeg-OTspec at yahoogroups.com]
>> On Behalf Of Ken Lunde
>> Sent: Friday, July 27, 2012 4:17 PM
>> To: OTspec
>> Subject: [mpeg-OTspec] Good news about CFR (aka ISO/IEC 14496-28:2012)
>> 
>> All,
>> 
>> Mac OS X Version 10.8 (Mountain Lion) is the first CFR implementation,
>> and thus the first implementation to break the 64K glyph barrier. Many
>> thanks to the folks at Apple for pulling this off. I wrote about it
>> this morning on our CJK Type Blog. See:
>> 
>>  http://blogs.adobe.com/CCJKType/2012/07/cfr-support-in-mountain-
>> lion.html
>> 
>> Regards...
>> 
>> -- Ken
>> 
>> 
>> 
>> ------------------------------------
>> 
>> Yahoo! Groups Links
>> 
>> 
>>