Re: [apps-discuss] font/*

[John C Klensin <[e-mail]>]

--On Tuesday, November 08, 2011 09:20 -0500 Eric Burger
<[e-mail]> wrote:

> Is the idea we would see something like
> 
> 	font/Times
> 
> or one of 
> 
> 	font/PostScript/Times
> 	font/TrueType/Times
> 	font/OpenType/Times
> 	font/METATYPE/Times
> 
> or one of
> 
> 	font/Times/PostScript
> 	font/Times/TrueType
> 	font/Times/OpenType
> 	font/Times/METATYPE
> 
> One cannot just say it's font/* and assume it is an opaque
> container, as one could see battles over the latter examples.

Yes, that is part of what I was concerned about.

If the intent is merely to pass the descriptions/ definitions
around, then things like
   font/TrueType name="Times" 
might work, where "TrueType" is what I called a description
language in my earlier note.

But...


>...

--On Tuesday, November 08, 2011 18:39 +0100 "t.petch"
<[e-mail]> wrote:

>...
> BTW, before naming this thing, please have a discussion with a
> typographer about the difference between a "font" and a
> "typeface."
> 
> <tp>
> and typeface, versus typeface family (which, I suspect, is
> what many mean when they say 'font')
> 
> Tom Petch
> 
> </tp>

Indeed.  And that is why I made the comment about size ranges
(for description types that facilitate scaling) and about
styles. Does, e.g., 
   font/TrueType name="Times"
require that the entire typeface family be sent?  Think about
what happens if one wants to send the Bold style only.  Is 

   font/TrueType name="Times" style="Bold" 

sufficient?  Can TrueType handle some or all of that internally?
Is external labeling (as part of the Content Type) needed to
provide a referencing handle?  Can all of the other description
formats do the same thing.

Note that it is common to have mail messages in which a body
part (possibly text/html) references images that are additional
body parts of the same message.   Could one contemplate sending
a font and the content that uses it in the same message?  Does
that create any special issues?

I think this is probably worth doing, but I think there are a
lot of questions to be answered... and that the result is
unlikely to be a one-page RFC.

For example, I can see a really interesting IANA registry for
subtypes and keywords coming along -- probably not with a
one-page spec.

    john