new draft available

[Markus Gylling]

Hi all,
As decided on our last concall, we have now pushed a new draft of EPUB 3.0 as Proposed Specification to the trunk and to idpf.org [1]. We hereby ask working group members to voice any remaining concerns or issues within 24 hours from now (before 22:00 UTC September 7), else we will consider this draft endorsed by the working group, and proceed with board submission.

/the chairs

[1] http://idpf.org/epub/30/spec/

[Peter Sorotokin]

A minor typo: change from profile to prefix attribute name was not done in
section "2.2.4.2 The nav Element: Types".

Peter

[matt.g...@bell.net]

Thanks, Peter! I've removed that parenthetical[1] and linked the word default to the section defining the default vocabulary. Changes should be on the site shortly.
 
[1] http://epub-revision.googlecode.com/svn/trunk/build/30/spec/epub30-contentdocs.html#sec-xhtml-nav-def-types
 
Matt

[Markus Gylling]

> However, in the section 3.3.3 CSS 3.0 Speech,
> http://idpf.org/epub/30/spec/epub30-contentdocs.html#sec-css-voice-characteristics,
> the '-epub-voice-rate' and '-epub-voice-pitch' are omitted.
> Since the real world voice book products are already able to adjust the
> voice rate by hand,

Hi Chen,
the features you refer to are marked "at risk" in CSS3-Speech, and have therefore not been included in the EPUB CSS Profile. This decision stands. However, note that Reading Systems may still implement support for these features, and stylesheets may still include them, something that would contribute to removing [1] the "at-risk" branding from the CSS WD.

The informative statement in the Overview document that you refer to ("such as pronunciation, volume, pitch and rate"), will be changed to "such as pronunciation, prosody and voice characteristics". Thanks for noting this inconsistency.

/markus

[1] http://lists.w3.org/Archives/Public/www-style/2011Sep/0061.html

On Sep 7, 2011, at 7:48 AM, jtchen0901 wrote:

> hi Markus,
> According to the chapter,
> http://idpf.org/epub/30/spec/epub30-overview.html#sec-tts ,
> the EPUB 3 should provide the TTS facilities such as pitch and rate.
> However, in the section 3.3.3 CSS 3.0 Speech,
> http://idpf.org/epub/30/spec/epub30-contentdocs.html#sec-css-voice-characteristics,
> the '-epub-voice-rate' and '-epub-voice-pitch' are omitted.
> Since the real world voice book products are already able to adjust the
> voice rate by hand,
> I suggest you to consider putting both '-epub-voice-rate' and
> '-epub-voice-pitch' attributes back to the section 3.3.3.
>
> Sincerely
>
> Jiang-Chun Chen
>
> --------------------------------------------------
> 寄件者: "Markus Gylling" <markus....@gmail.com>
> 日期: 2011年9月7日 06:37
> 收件者: "Google Groups EPUB" <epub-work...@googlegroups.com>
> 主旨: new draft available

>> [1] http://idpf.org/epub/30/spec/=
>
> ====================================================================
> 本信件可能包含工研院機密資訊，非指定之收件者，請勿使用或揭露本信件內容，並請銷毀此信件。 This email may contain confidential information. Please do not use or disclose it in any way and delete it if you are not the intended recipient.
>

[Leonard Rosenthol]

Since discussions about "fixed format" are still ongoing, I'd like to
suggest a rewording of the NOTE in 3.3.1.

I recommend that we use simply state the following:

The ability of Reading Systems to paginate absolutely positioned layouts
is not guaranteed.

Leonard

[Leonard Rosenthol]

[Apologies in advance if this has been previously discussed]

In 3.3.4, we specifically require an RS to support OpenType and correctly
refer to the ISO version of that specification.

However, there are many people who believe that OTF fonts only include
SFNT-based (aka TrueType) content and do not also implement the CFF-based
version. Do we think there is a need for a NOTE about the need for
supporting both?

Leonard

On 9/6/11 6:37 PM, "Markus Gylling" <markus....@gmail.com> wrote:

[Daniel Weck]

Hi Chen,
I don't think your email was sent to the EPUB mailing list, so I am discovering it only now (thanks Markus for responding).
Note that I responded separately to your message on the CSS mailing list, but your question wasn't explicitly related to EPUB:

http://lists.w3.org/Archives/Public/www-style/2011Sep/0061.html

Please note that "pitch" and "rate" CSS properties were *not* included in the EPUB3 CSS-Speech subset, these features were/are only (wrongly) mentioned in the "overview" section, see my recent email about rectifying this issue (Markus, thank you for fixing the first sentence of [2.10 Text-to-speech], which I did not catch):

http://groups.google.com/group/epub-working-group/browse_thread/thread/cca9aacb98b4a84f

References:

http://epub-revision.googlecode.com/svn/trunk/build/30/spec/epub30-overview.html#sec-tts

http://epub-revision.googlecode.com/svn/trunk/build/30/spec/epub30-contentdocs.html#sec-css-voice-characteristics

Regards, Daniel

[Cramer, Dave]

Agreed. Better to at least tolerate fixed formats as part of EPUB 3 than
for them to exist outside of the spec entirely. Publishers will be making
these anyway.

Dave Cramer

[Bill McCoy]

I don't disagree with Leonard's recommendation, which could be commended even just on grounds of brevity since the text he suggests eliding ("....so reliance on absolute positioning is discouraged. Reading Systems might not support these property values") is pretty  much just a corollary/restatement of the non-guarantee he suggests retaining, which itself is arguably just a corollary of the normative statement above about the fixed value of position property not being part of the profile.

However, my $.02 is that at this point WG members should be suggesting to Markus & Matt only extremely important changes, i.e. fixes for bugs/typos, with focus on normative statements. We could be fine tuning and debating non-normative prose forever and in the big picture that would not be a good thing in order that we get EPUB 3 out the door and broadly adopted. And, we know we will soon need an EPUB 3.01 etc. if for no other reason than expected changes to unfinished dependent W3C specs. So please take schedule considerations into account in making further last-minute suggestions.

Thanks!

--Bill

--

Bill McCoy

Executive Director

International Digital Publishing Forum (IDPF)
bmc...@idpf.org

+1 (206) 353-0233

[Markus Gylling]

Leonard,

> Since discussions about "fixed format" are still ongoing, I'd like to
> suggest a rewording of the NOTE in 3.3.1.

I see the point you (and Dave C) are making. However, as I recall (and as unlikely as it perhaps may seem) we spent quite some time fiddling with this passage early this year to get it to a state that everyone could live with. Therefore I am really hesitant to change it now.

Future minor updates can revise this passage to adapt it to changes in the environment.

Hope this works for you.

/markus

[Markus Gylling]

> Do we think there is a need for a NOTE about the need for supporting both?

Thanks for this suggestion. As there is a potentially infinite amount of such clarifying notes we could add, I have added this as an issue to the tracker with status deferred (a label we to use facilitate retrieval in the future when doing revision work).

This means that I am suggesting we do not make any edits to the spec now with regards to this suggestion; let me know if this does not work for you.

http://code.google.com/p/epub-revision/issues/detail?id=173

Regards, /markus

[Leonard Rosenthol]

Perfectly fine - thanks!

[Leonard Rosenthol]

I would not be my preference, but I'll live with it :).

[Eric Muller]

On 9/7/2011 6:46 AM, Leonard Rosenthol wrote:

[Apologies in advance if this has been previously discussed]

In 3.3.4, we specifically require an RS to support OpenType and correctly
refer to the ISO version of that specification.

However, there are many people who believe that OTF fonts only include
SFNT-based (aka TrueType) 

sfnt is not synonymous with TrueType.

sfnt is the packaging: organization in tables, with one or more (TTC) list of tables that form a font. As generally used, the term does not imply any meaning on the tables.

What the tables mean comes from 1) the Apple True Type specification, 2) the Microsoft True Type specification, 3) the Microsoft True Type Open specification, 4) the Microsoft/Adobe OpenType specification, 5) the ISO Open Font Format (OFF) specification, 6) the Graphite specification, 7) the Apple AAT specification (arguably an extension of their TT spec). [And yes, 2, 3 and 4 are successors and more or less compatible]

Taking the aggregate of all the specifications, there are two main axis of differences:

- the format in which the outlines are represented: TrueType (loca and glyf tables) vs. CFF (CFF table, descendant of Postscript)

- the format in which the layout information is represented: OpenType (GSUB, GPOS, etc), AAT (mort, morx, etc), or Graphite (don't remember the name of their tables).

The two axis are largely orthogonal.

What Leonard is trying to say is that EPUB should be explicit that both TrueType and CFF outlines should be supported. The reference to OFF implies OpenType layout (as it's the only supported).

Eric.

[Bill McCoy]

Eric,

This came up only a few months back  in WG discussion (see thread "OpenType formats" started May 11 2011). The resolution was that we were OK referencing ISO spec since it defines both TrueType and CFF outlines and we didn't want to go down the slipperly slope of redundantly defining conformance requirements for the specs we depend on. I.e. an EPUB Reading System that implements only a subset of ANY spec we normatively reference is by definition not fully conformant.

--Bill

[Markus Gylling]

Hi all,
below is a list of the edits/bugfixes done to the draft since yesterdays 24 hr deadline announcement. With these edits in place, I intend to submit the spec to the Board at 18 UTC tomorrow the 8th, with a recommendation to proceed with membership vote to approve as a final Recommended Specification.

* fixed profile reference in ContentDocs 2.2.4.2 (r3171)
* fixed a typo (missing word) in ContentDocs 2.1.3.1.3 (r3172)
* fixed profile reference and corrected example indentation in MediaOverlays (r3173, r3174)
* fixed CSS Writing Modes dated reference in ContentDocs 3.3.6 (r3175)
* fixed informative prose re CSS Speech in Overview 2.10 (r3176)
* added restriction on -epub-hyphens 'all' value in ContentDocs 3.3.5 (r3176)

(Note, in the list above, the parentheses reference the revision number. Refer to http://code.google.com/p/epub-revision/updates/list for diffs.)

/markus