W3C home > Mailing lists > Public > www-voice@w3.org > October to December 2003

RE: Reply: The mark tag My comment on SSML working draft 5 April 2002.

From: Daniel Burnett <burnett@nuance.com>
Date: Wed, 19 Nov 2003 14:05:18 -0800
Message-ID: <ED834EE1FDD6C3468AB0F5569206E6E90273212D@MPB1EXCH02.nuance.com>
To: "Stijn \"Adhemar\" Vandamme" <Stijn.Vandamme@UGent.be>
Cc: <www-voice@w3.org>

Dear Adhemar,

Thank you for your comments.

We maintain our official Rejection of your request.  Here is our response:

>>> Upon consideration it has become even clearer to us that mark labels
>>> should not even be xsd:id's.  As an example, the uniqueness constraints
>>> of ID's are a hindrance rather than a benefit, e.g. it may well be
>>> desirable to repeatedly use the same mark label (equivalent to repeatedly
>>> sending back the same event).  We also wish to permit integer labels, for
>>> example.
>>> Because of this desire to have fewer restrictions than those introduced
>>> by ID's, we have decided to change the name attribute to be of type
>>> xsd:token.

If you believe we have not adequately addressed your request,
please let us know as soon as possible.  If we do not hear
from you within 7 days, we will take this as tacit acceptance.

Again, thank you for your input.

-- Dan Burnett
Synthesis Team Leader, VBWG

-----Original Message-----
From: Stijn "Adhemar" Vandamme [mailto:Stijn.Vandamme@UGent.be]
Sent: Monday, August 18, 2003 1:41 PM
To: Daniel Burnett
Subject: Reply: The mark tag My comment on SSML working draft 5 April


Thanks for your official response.

On Fri, 8 Aug 2003, Daniel Burnett wrote:
> We have two responses to this. First, we use a special element
> rather than allowing every "id" attribute to act as a mark
> because it could be expensive to stream back an event for every
> "id" you encounter during production, especially for a small-
> footprint device using a network-based synthesis processor.
> Second, we use a "name" attribute on the mark element because
> that's a better description of what it is. If you'll notice,
> the schema actually uses ID as the type. In short, XML id's
> are generally used as anchors for external references *into*
> a document -- you can never have too many -- while
> <mark name= ...> is primarily used to convey information back
> while processing the document.

If I understand this correctly, you use the "name" attribute forconveying
information back, because ideally one wants a lot of id's (in various
tags) for external references, but one doesnt want to stream back an
event for every id one encounters, because it isn't necessary. I can
understand that.

But I don't understand why one can't stream back an event for every id
(instead of every name) one encounters in a <mark> tag. This way, every
internal reference can automatically be used as an external reference,
and one doesn't stream back unnecessary events.

(As for "name" being a better description: in the old HTML, external
references were made via <a name="ref"> because name was a better
description of what the attribute was. This is changed in the newer XHTML,
for (informally speaking) complying with XML custom.)

Received on Wednesday, 19 November 2003 17:05:23 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:07:37 UTC