ietf-822
[Top] [All Lists]

Re: A spec for showing language in MIME headers

1993-11-13 11:53:33
Masataka Ohta, quoting Dana Emery (me):

  The common error with blindness and such is people tend
  to think that blind and non-blind are a lot different.
  They are not.

Interesting you should presume I (we) lack experience with 
blind people, as a matter of fact I have socialized, worked, 
and performed music with people who happened to be blind.

if the original message content was digitized voice,
then a  mapping to australian accent will be an interesting
technical challange.
  
  which has nothing to do with blindness.

but every thing to do with your original comment of:

  An Australian blind want to hear Australian English,
  even if the letter is from US.
----
  Just as an Australian non-blind can read en-us mails
  without content language information, an Australian
  blind can recognize it without content language
  information.

Ah, but the agent doing the reading for the Australian-blind 
would be able to use the content language information if it 
were present, and IMHO the agent will need it for technical reasons.

  It is a poor quality presentation system if it needs
  such a configuration.

Ohta san, I must conclude that you have missunderstood me, 
so I will restate my original premise:

I envision 2 different forms of voice mail, 

1) Mail recorded audibly, ie digitized voice transmitted 
as encoded compressed binary.
2) Mail recorded and transmitted textually, intended for 
humans to "read", but rendered audibly by a clever user 
agent "reading" it (probably for a blind person).

1) is quite posible today.
2) is demonstrable today, but impracticle due to limited 
vocabularies.

I mention "dictionarys" in association with 2) for reasons
which should be obvious, translation of written text to 
audible form requires some sort of data.  Accents become 
an issue as detailed original voice characteristics are unknowable.
--
dana s emery <de19(_at_)umail(_dot_)umd(_dot_)edu>