Re: WOFF2 spec changes

Hello Vlad,

Tuesday, March 10, 2015, 10:55:45 PM, you wrote:

> The following changes have been implemented in the spec
> (http://dev.w3.org/webfonts/WOFF2/spec/#woff20Header) :
>  
> -          Action 165: revised the second paragraph of subclause
> 5.4 to optionally allow alphabetically sorted tables inserted
> between ‘glyf’ and ‘loca’ only when individual font files are
> encoded in WOFF2. For font collection files – ‘loca’ MUST 
> immediately follow the ‘glyf’ table.

"however when WOFF2 contains a font collection file both glyf and loca
tables MUST immediately follow each other"

That doesn't seem optimally clear, to me. How about

"however when WOFF2 contains a font collection file, for each font in
the collection, the glyf and loca table pairs MUST immediately follow
each other"

perhaps add an example:

not a collection
cmap glyf hhea hmtx loca maxp ...

collection
cmap glyf loca glyf loca hhea ...


>  
> -          Action 163: revised the language of subclause 5.4 to
> remove any notion of the nominal size. Removed UA conformance case
> “mayRejectWrongGlyfTableLength” – I don’t think this neither necessary nor useful.

Not sure why it says

"Typically, the origLength field value would specify the actual glyf
table size; "

rather than, more simply,

"The origLength field value specifies the actual glyf table size of
the original font; "

Other than that, seems good to me

> The text needs review, there are potential changes still needed to
> firm up the description of the origLength value and there may be
> changes to the conformance requirements for file format.
>  
> -          Added “conform-magicNumber” and
> “conform-noMagicNumber-reject” conformance requirements  in the description of WOFF2 header.

OK

> -          Clarified the meaning of the “uncompressed” font data
> size in subclause 5 (para #5) and removed the editorial note.
>  
> -          Extended metadata and private data blocks – changed the
> 4-byte alignment language to accommodate the differences between
> WOFF1 (table-based) and WOFF2 (stream-based) structure. The 4-byte
> alignment for metadata block is needed but no  longer assured since
> it follows a variable-length compressed data stream. Added
> “conform-metadata-padalign” conformance requirement.

OK



-- 
Best regards,
 Chris Lilley, Technical Director, W3C Interaction Domain

Received on Wednesday, 18 March 2015 13:40:58 UTC