Re: A way forward

2009/7/24 John Daggett <jdaggett@mozilla.com>:
> John Hudson wrote:
>
>>> But, if that is the case - if the existing cross site restriction is
>>> good enough for foundries who support EOT, and their aim is to get
>>> profiting from web fonts ASAP, why isn't supplying TTFs with corrupt
>>> NAME tables and a changed file extension good enough?
>>
>> Because its a hack, because it exposes the font to unknown dependencies
>> in which it might not function correctly, and because we've spent the
>> past ten years getting good at producing fonts to spec rather than
>> putting in hacks to solve short-term software issues. We want a nice
>> clean web font spec, against which we can test our products. Further,
>> some of us have customers whose procurement requirements would prevent
>> us from delivering fonts with corrupt data.
>
> Name tables can be obfuscated without being "corrupt", they can be
> entirely copasetic but constructed in such a way that they would
> effectively be unusable in desktop apps.

Typotheque's and one other influential foundry that I have seen
licensing TTFs for "raw" usage but with referrer checking did corrupt
the NAME table. That seems to be what they prefer, despite when you
publicaly suggested this approach you (as you do here) suggested
spec-compliant but 'copasetic' NAMEs.

Received on Friday, 24 July 2009 22:50:33 UTC