Re: RFC: EOTL font file specification version 1

Thomas Phinney wrote:
>> If compatibility with legacy user agents (MSIE) is not required,
> 
> The entire reason for considering this format is compatibility with
> legacy user agents; it is a requirement of the spec.

I agree. I was trying to portrait the minimum overhead for the future,
where even MSIE has implemented conforming support for EOTL. If support
for EOT classic is not required in the future, then the EOTL as I
specified has only 8 byte overhead over plain OTF files.

>> In the end, any EOT classic font without XOR or MTX scrambling
> 
> I am of the impression that no pre-existing EOT classic fonts would
> meet that requirement. Somebody could create new or revised EOT fonts
> specifically with these restrictions in mind, however.

OK. In that case, this practically the same as below.

>> Cons:
>>
>> - Undermines the "protection" provided with rootstrings in current EOT
>> files (though, I'd assume that most EOT files with rootstrings also
>> include XOR or MTX scrambling and would not be successfully loaded).
> 
> I'm under the impression that it's "all" rather than "most"; it would
> be good to get somebody knowledgeable about ancient EOT history to
> comment on this.

I'd be interested about this, too.

Then I cannot see any cons with the EOTL format I suggested :-)

-- 
Mikko