[mpeg-OTspec] RE: font media types

Tue Apr 5 23:00:50 CEST 2011

Hi Sairus,

a)      My understanding is that TTC could use either font-ttf or font-off type;

b)      The inability to render a particular type of outlines would prevent an application from displaying the text, the inability to process a layout table may cause text to not look good/right but would still allow producing something readable or, at least, visible. I don't think that intended use of MIME types is to communicate every possible detail about a resource, but it does offer a chance to skip the downloading a resource that cannot be processed.

c)       I am not sure I would agree with this interpretation. A compliant rasterizer will simply ignore any table that it cannot use/understand, but I don't think that font containing a set of tables not specified by the standard should be considered compliant. While OT/OFF spec doesn't forbid custom extensions of any kind but it doesn't mean that fonts containing an unspecified and unknown set of tables can be considered a compliant resource (and it definitely will not pass a conformance test having certain tables flagged as unknown).

Best regards,
Vlad

From: Sairus Patel [mailto:sppatel at adobe.com]
Sent: Tuesday, April 05, 2011 4:41 PM
To: Levantovsky, Vladimir; Thomas Phinney
Cc: John Hudson; David Lemon; OTspec; Chris Lilley; Tab Atkins; Karsten Luecke
Subject: RE: [mpeg-OTspec] RE: font media types

a. What about TTCs?

b. I'm unclear about the actual problem the MIME types are supposed to solve. Why is it useful to know the glyph outline technology but not the layout table technology (OTL vs AAT) in a MIME type? If an HTML UA doesn't support AAT, for example, it could skip trying to download a font.

c.
> SFNT structures that are not compliant with the ISO standard

I do not think that OT/OFF forbids AAT tables (or any other tables unknown to OT/OFF) to be present in an sfnt. An sfnt containing TT outlines, cmap/hmtx and other required tables, and AAT layout tables but no GSUB/GPOS/ OTL tables should be considered to be a valid OT/OFF font. The AAT tables certainly aren't required to be supported by an OT/OFF engine.

Best,
Sairus

From: mpeg-OTspec at yahoogroups.com [mailto:mpeg-OTspec at yahoogroups.com] On Behalf Of Levantovsky, Vladimir
Sent: Tuesday, April 05, 2011 1:13 PM
To: Thomas Phinney
Cc: John Hudson; David Lemon; OTspec; Chris Lilley; Tab Atkins; Karsten Luecke
Subject: RE: [mpeg-OTspec] RE: font media types

Okay, thanks.

From: thomas.phinney at gmail.com [mailto:thomas.phinney at gmail.com] On Behalf Of Thomas Phinney
Sent: Tuesday, April 05, 2011 4:13 PM
To: Levantovsky, Vladimir
Cc: John Hudson; David Lemon; OTspec; Chris Lilley; Tab Atkins; Karsten Luecke
Subject: Re: [mpeg-OTspec] RE: font media types

If you dropped explicit mention of having both outlines a! t once, you could just sy that OFF was a generic type that could have either kind of outlines. I think that is fine (and potentially useful under some circumstances).

Regards,

T
On Tue, Apr 5, 2011 at 1:01 PM, Levantovsky, Vladimir <Vladimir.Levantovsky at monotypeimaging.com<mailto:Vladimir.Levantovsky at monotypeimaging.com>> wrote:
I agree with you, but this is only a recommendation - there is no guarantee that this [mixing both types of outlines. i.e.] would never happen.
What would be your recommendation - should we account for it? Would having a generic type that indicates compliance with th! e OpenType / OFF spec be useful?

Regards,
Vlad

From: thomas.phinney at gmail.com<mailto:thomas.phinney at gmail.com> [mailto:thomas.phinney at gmail.com<mailto:thomas.phinney at gmail.com>] On Behalf Of Thomas Phinney
Sent: Tuesday, April 05, 2011 3:54 PM
To: Levantovsky, Vladimir
Cc: John Hudson; David Lemon; OTspec; Chris ! Lilley; Tab Atkins; Karsten Luecke

Subject: Re: [mpeg-OTspec] RE: font media types

That's a couple of times you've mentioned a font having both kinds of outlines. I don't think that's a possibility that should be encouraged, and I for one would prefer NOT to explicitly designate how to handle it. It is specifically recommended against in the OT spec: "Both Microsoft and Adobe recommend against mixing outline formats within a single font. Choose the format that meets your feature requirements." (http://www.microsoft.com/typography/otspec/recom.htm)

Cheers,

T

On Tue, Apr 5, 2011 at 12:26 PM, Levantovsky, Vladimir <vladimir.levantovsky at monotypeimaging.com<mailto:vladimir.levantovsky at monotypeimaging.com>> wrote:

Thank you John,

I believe making a distinction based on the types of glyph outline data has always been the primary goal, and what you proposed seem to make perfect sense. However, I'd like to keep an "umbrella" media type to separate OpenType/OFF-based SFNT objects vs. arbitrary SFNT-based format.

Strictly speaking, defining a media type for an arbitrary SFNT-based format would go beyond the scope of ISO OFF specification, but I think it would be acceptable to attempt to define this media type as part of the ISO document anyw! ay, to offer a way to distinguish SFNT font structures that are not compliant with the ISO standard (and we know there are many of them out there). With this in mind, would the following be acceptable to everyone:

- application/font-ttf - for traditional TrueType fonts and OpenType/OFF fonts with TrueType outlines;
- application/font-cff - for OpenType/OFF fonts with CFF outlines;
- application/font-off - for OpenType/OFF fonts with both TrueType and CFF outlines, or as a generic type for a font that is compliant to ISO OFF spec, and
- application/font-sfnt - for generic SFNT-based font formats.

Please voice your objections / concerns / comments / agreement - I need to finalize the text of the annex and submit it to ISO SC29 by the end of the week.

Thank you,
Vlad

> -----Original Message-----
> From:! John Hudson [mailto:j ohn at tiro.ca<mailto:john%40tiro.ca>]
> Sent: Tuesday, April 05, 2011 2:08 PM
> To: Levantovsky, Vladimir
> Cc: David Lemon; OTspec; Chris Lilley; Tab Atkins; Karsten Luecke
> Subject: Re: [mpeg-OTspec] RE: font media types
>
> Vladimir wrote:
>
> > We can also consider changing the proposed
> > media types, e.g. to be more specific and use font-ttf, font-cff and
> > font-otf for the purposes outlined above.
>
> If the distinction is to be based on outline type, then I think
> 'font-ttf' and 'font-cff' is much preferable to using 'font-otf' for
> the
> latter. The only reason .otf is not typically used as a file extension
> for TrueType fonts is so that the latter are backwards compatible with
> older Windows systems that do not recognise .otf; this is obviously
> less
> of a practical concern than it w! as in 1998, and now persists mostly as
> convention. I have, however, seen some TrueType fonts with an .otf
> extension.
>
> OTF stands for OpenType Font, not PostScript OpenType Font, and the
> first item in the OpenType catechism is that an OpenType font may
> contain either TrueType or PostScript outlines. Hence, I think it is
> misleading or at least confusing to identify 'font-otf' with PostScript
> outlines.
>
> I also wonder about using either 'font-off' or 'font-otf' to indicate
> an
> unknown outline type: the former because OFF is not a widely used or
> recognised abbreviation -- the Open Font Format being reasonably seen
> as
> a particular publication of the OpenType Font specification --, and the
> latter because of potential confusion in the minds of those who, due to
> convention, associate .otf with PostScript outlines.
>
> I wonder if there m! ight be grounds for a generic 'font-sfnt' MIME type,
> w! hich wou ld not only express the desired agnosticism regarding outline
> type, but would also be useful for fonts containing non-OTL layout
> tables.
>
> So we might have:
>
> application/font-ttf
> application/font-cff
> application/font-sfnt
>
>
> JH
>
> --
>
> Tiro Typeworks www.tiro.com<http://www.tiro.com>
> Gulf Islands, BC tiro at tiro.com<mailto:tiro%40tiro.com>
>
> A pilgrimage is a journey undertaken in the
> light of a story. -- Paul Elie

--
"Puritanism: The haunting fear that someone,
 somewhere, may be happy."
&! nbsp;-H.L. Mencken

--
"Puritanism: The haunting fear that someone,
 somewhere, may be happy."
 -H.L. Mencken

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.aau.at/pipermail/mpeg-otspec/attachments/20110405/6270c335/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/x-ygp-stripped
Size: 322 bytes
Desc: image001.jpg
URL: <https://lists.aau.at/pipermail/mpeg-otspec/attachments/20110405/6270c335/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/x-ygp-stripped
Size: 322 bytes
Desc: image002.jpg
URL: <https://lists.aau.at/pipermail/mpeg-otspec/attachments/20110405/6270c335/attachment-0001.bin>