Stuart_McRae/STA/Lotus (Stuart_McRae/STA/Lotus@uk.ibm.com)
Fri, 24 Aug 2001 08:04:47 +0100
Eric, Thanks for your comments...
>Although it might not be obvious, a "voice message context" without audio
>does make sense. For example, a bunch of people are working on
>transcription services, where they record a voice message and store or
send
>a text e-mail. The semantics are still voice (e.g. a person cared enough
to
>call), but there's no audio part to the message.
I agree that one can image a message context of voice-message that does
not contain an audio part, if that voice message is encoded in some other
way (dictated text, a url pointing to the voice message, whatever).
I was less clear before what a user agent is supposed to do with such a
message. I understand that the Message Context draft specifically does not
try to prescribe such behaviour, but what does it do in practice (what
hint is it giving)? A telephone icon would make sense (at least in your
example), but launching a special viewer may or may not (depending on
whether it is able to cope with the content type provided), and whether or
not responding with voice makes sense is unclear (it probably depends
whether the e-mail originator is the real originator or a transcription
service).
Ultimately I decided that allowing voice-message without audio parts this
was (on balance) a good thing - the thing that tipped the balance was
using the icon (and everything that goes with that), since this is a voice
message and should be presented as such.
However, right now I think this is orthogonal to the Internet Voice
Message question. I know that Message Context started out as a way to
identify vocie messages for IVM, but it has now taken on an existence of
its own. And, in a way, relieved IVM of the need to worry about this.
I think this allows IVM to be more concerned with what it has become
primarily about - ensuring that you can deliver an audio voice message to
user agents in a way that is usable, particularly on the installed
(legacy) base of user agents. It is a question of simplification, rather
than overloading the specification with things it no longer needs to worry
about. After all, the reality is that is you do not include an audio
content using one of the defined codecs in the message, then IVM is adding
almost now value to the transaction at all.
Therefore... revised proposal...
An Internet Voice Message MUST contain at least
one audio part, which may be at any location within the message and
MUST
be contained in either an audio/wav or audio/basic content-type, unless
the originator is aware that the recipient can handle
other audio content. Specifically, Audio/32kadpcm
MAY be used when the recipient is known to support VPIM v2 [VPIMV2].
It should be noted that voice messages which are presented in a form
other than audio (for example a transcribed text message) MAY still be
marked with a Message Context of voice-message (per [HINT]) so that the
recipient is aware that they were originally left as a voice message.
However these are not Internet Voice Messages in the context of this
document.
What do you think?
Regards,
Stuart
This archive was generated by hypermail 2.0b3 on Fri Aug 24 2001 - 10:08:54 IDT