[mpeg-OTspec] cmap format4 structure question

Opstad, Dave dave.opstad at monotype.com
Mon Jan 27 19:27:24 CET 2014

Hi Bob,

Sorry I wasn't clearer! The "code" I referred to was the original Apple implementation from the late 1980s, and is the way all copies of that original TrueType 1.0 behave. Basically, anyone who licensed that code from Apple, or from one of its licensees in turn (like Microsoft) would have the format 4 'cmap' subtable implementation behave as I described. So there is a body of existing code that expects TrueType fonts to behave that way.

It is a separate and very valid question to ask, as to whether this existing dependency is appropriate. I was merely providing some of the history, as one of the authors of that original implementation. If you expect fonts produced in accordance with this new spec to work on existing TrueType installations then it's an important issue to resolve.


From: Bob Hallissy <bobh528 at yahoo.com<mailto:bobh528 at yahoo.com>>
Date: Monday, January 27, 2014 9:52 AM
To: OTspec <mpeg-OTspec at yahoogroups.com<mailto:mpeg-OTspec at yahoogroups.com>>
Subject: Re: [mpeg-OTspec] cmap format4 structure question

Thanks for your reply, Dave. Some follow-up questions (for you or anyone):

When you say "the TrueType code" to what product(s) are you referring?

Is there something in the spec that I'm missing that would suggest
depending on sentinel of (0xFFFF, 0xFFFF) is appropriate?

If not, do we need to adjust the spec?


On 2014-01-24 at 8:41 Opstad, Dave wrote:
> As I recall, the TrueType code for format 4 'cmap' subtables is
> looking for a sentinel of (0xFFFF, 0xFFFF) specifically, so I don't
> think the (0xFFFC, 0xFFFF) you posit would work.
> Cheers,
> Dave Opstad

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.aau.at/pipermail/mpeg-otspec/attachments/20140127/94721e08/attachment.html>

More information about the mpeg-otspec mailing list