[mpeg-OTspec] New "Composite Font Standard" document

Tue Dec 15 23:36:31 CET 2009

Leonardo,

Again, many thanks for the valuable and timely feedback, which has been instrumental in moving this standard forward.

Allow me to explain the use, meaning, and intention of "optional" for this standard.

First, some background is necessary. All current composite font (and fallback font) formats are semi- or entirely proprietary, and the existence of formal documentation cannot always be assumed. In other words, we're in somewhat new or unexplored territory.

The standard specifies several tags, and each tag has one or more attributes. All of the data associated with each tag is specified through the use of attributes, so any tag without at least one attribute is meaningless. And, not all tags are required. Still, a CFS object without any tags or tag attributes serves no purpose. The intention of "optional" is thus meant to mean the opposite of "required." For example, it is possible to define a CFS object using only the <cfs.CompositeFont> and <cfs.ComponentFont> tags, in which case the Encoding and Language information is implicit by the <cfs.ComponentFont> tag's "Target" attribute. A CFS object probably needs a name, thus the requirement to use the <cfs.CompositeFont> tag. And, because CFS objects need to reference at least one component fonts, requiring at least one instance of the <cfs.ComponentFont> tag seemed like a good idea. This is a summary of the optional/required situation as it applies to the tags. Similar arguments could be made for each tag's attributes.

Then again, I am sure that someone more creative than me could conjure up realistic usage scenarios through which it can be demonstrated that all tags, along with their attributes, are optional, thus rendering ineffective the use of "optional" anywhere within the standard.

If you still feel that removing any reference to the word "optional" is the right thing to do, then I shall effect this change, and the next version of the document will reflect this.

Regards...

-- Ken

On 2009/12/15, at 14:08, Leonardo Chiariglione wrote:

> Ken,
> To me the document is fine the way it is, save for the "optional" part.
> The text 
> 
> NOTES: Although not yet reflected in this document, the notion of "optional"
> refers to an attribute whose inclusion is deemed to be optional according to
> this specification. Likewise, some attributes are required. A future version
> of this document will make the "optional" versus "required" distinction
> clear, for both tags and their attributes.
> 
> As a matter of philosophy I object to the use of the term "optional" in a
> standard, because the purpose of a standard is to tell users what they
> should do to achieve interoperability.
> Of course I know that not everybody may have the same notion or requirement
> of interoperability. As you may remember from an earlier post, since its
> early days MPEG has struggled with this reality and has come up with the
> "profile and level" approach. 
> Instead of having a placeholder for "optional" with the intention to give it
> a meaning later, I would like to give a meaning to "optional" and then find
> a place for it.
> As I said interoperability requirements are not the same for all, so there
> is nothing new for MPEG in your request, but I would like to get a
> formalisation of the thing before admitting it in
> Leonardo 
> 
> -----Original Message-----
> From: Ken Lunde [mailto:lunde at adobe.com] 
> Sent: 15 December 2009 19:41
> To: leonardo at chiariglione.org
> Cc: mpeg-OTspec at yahoogroups.com
> Subject: Re: [mpeg-OTspec] New "Composite Font Standard" document
> 
> Leonardo and others,
> 
> I spent a significant part of my morning continuing the editing and refining
> of this document, and have issued a 2009/12/15 version. See:
> 
> 
> http://groups.yahoo.com/group/mpeg-OTspec/files/CompositeFontStandard-200912
> 15_kl.doc
> 
> I also hosted the same file on my website, along with a PDF version:
> 
>  http://lundestudio.com/PDF/CompositeFontStandard-20091215_kl.doc
>  http://lundestudio.com/PDF/CompositeFontStandard-20091215_kl.pdf
> 
> Regards...
> 
> -- Ken
> 
> On 2009/12/15, at 4:12, Leonardo Chiariglione wrote:
> 
>> Ken,
>> Excellent.
>> I have done the following
>> 1.      Converted your pdf file to a doc file (MS word is the format
> adopted by MPEG because it allows reuse of text, a basic feature in a
> collaborative environment like this)
>> 2.      Edited the file in track change mode
>> 3.      Changed the file name by using yyyymmdd_lc
>> 4.      Attached the file to this email
>> You may think that some changes are not essential and there may be
> elements of truth in this. However, if you agree to the changes we get
> agreement on issues that I believe are of substance.
>> I will try and explain my point
>> 1.      The title of the first section has been changed to "Design
> features". OS independence is a consequence of the fact that CFS is an
> information representation standard, so it has been taken out of the
> bulleted list
>> 2.      At the abstract level there are 4 actors: creator, encoder,
> decoder, consumer. Don't let's talk about the names, yet. If you don't like
> the names I am fine with calling the actors A,B, C and D, Creator and
> consumer are humans (maybe there are more general cases when they are not,
> but let's stick to humans for the moment).
>> 3.      Creator has intentions that he captures in his encoder - an
> application
>> 4.      Consumer has intentions that he expresses by setting his decoder
>> 5.      The standard only addresses the decoder. The document is silent on
> exactly "what" of the decoder is specified. I guess that the interface
> through which the consumer sets the decoder is not part of the standard, but
> then it would be good to say this explicitly (note that all this is work
> saved for the moment in which the standard will be written)
>> 6.      In MPEG we use the name conformance, but you can read it as
> compliance
>> Please feel free to react on anything I did.
>> For me this is a very pleasant experience.
>> Leonardo
>> 
>> From: mpeg-OTspec at yahoogroups.com [mailto:mpeg-OTspec at yahoogroups.com] On
> Behalf Of Ken Lunde
>> Sent: 14 December 2009 19:07
>> To: mpeg-OTspec at yahoogroups.com
>> Subject: [mpeg-OTspec] New "Composite Font Standard" document
>> 
>> 
>> All,
>> 
>> In an effort to move this along, I spent the morning incorporated
> Leonardo's comments and suggestions, and produced a new version of the
> document. You should have received a "new document notification" email with
> the following URL:
>> 
>> 
> http://groups.yahoo.com/group/mpeg-OTspec/files/CompositeFontStandard-121420
> 09.pdf
>> 
>> I also hosted the same file on my website:
>> 
>> http://lundestudio.com/PDF/CompositeFontStandard-12142009.pdf
>> 
>> When you have a chance, please read this document and provide feedback.
>> 
>> Regards...
>> 
>> -- Ken
>> 
>> 
>> No virus found in this incoming message.
>> Checked by AVG - www.avg.com
>> Version: 8.5.427 / Virus Database: 270.14.106/2563 - Release Date:
> 12/13/09 19:47:00
>> 
>> <CompositeFontStandard-20091214_lc.doc>
> 
> No virus found in this incoming message.
> Checked by AVG - www.avg.com 
> Version: 8.5.427 / Virus Database: 270.14.106/2563 - Release Date: 12/15/09
> 07:52:00
> 