FW: device upload update

Lucas Vogel lvogel@exponent.com
Wed, 15 Mar 2000 13:48:37 -0800


-----Original Message-----
From: James Salsman [mailto:jps-devup@bovik.org]
Sent: Wednesday, March 15, 2000 2:32 PM
To: devup-supporters@bovik.org
Cc: timbl@w3.org
Subject: device upload update


Dear Friends:

Thank you for all of your messages in support and endorsements, 
which have made a great difference.

Today Tim Berners-Lee informed me of the HTML working group chair's 
Addendum to the device upload submission comment, dated 3rd March:
  http://www.w3.org/Submission/1999/09/Comment

The addendum claims that input devices "shouldn't be visible in 
the markup" -- a position that would require the user of a web-based 
OCR application to select a scanner over a camera for each page of 
text, while a user of a teleconferencing application would need to 
select a camera over a scanner for each photo.

The questions asked in the Addendum indicate that the submission was
not read very well:

> What should happen if the user doesn't have a camera, but does have 
> a photograph and a scanner? What if there is a camera, but not 
> connected to the computer, so that the user has to take a picture, 
> download it, and then upload it from a file?

Those questions are answered even in the earliest submission draft:

"Under most conditions the operator should be allowed to select the 
device from ambiguous sources of input....  ...the browser operator 
should still be offered to select from multiple devices, with the 
only difference being the default selection...."
  -- http://www.w3.org/TR/device-upload

The W3C reply Addendum correctly states that device upload is 
possible with a proper implementation of HTML 4, ignoring the 
reality that most current browsers treat the ACCEPT attribute
as a filename pattern instead of a list of media types.  Nor 
is there any acknowledgement that basic microphone upload is 
only implemented on wintel-platform browsers, with inherently 
insecure binary plug-ins.

The media type alternate provisions and unsupported/unavailable 
device work within multipart/form-data headers is also ignored, 
as is the MAXTIME limitation for compressed media.  Those problems 
would not be solved by the Addendum's suggested solution of, "a 
note on recommendations to browser manufacturers on how to 
implement the user interface to file upload."

If you want the W3C to take a position on device upload which will
address these problems, requiring browser vendors who aspire to 
conform to W3C standards to implement device upload in a way that 
is both platform-independent and compatible with their own legacy 
implementations of simple file upload, then please reply to this 
message saying so.  Please don't change the subject line; the 
reply-to header on this message should send your reply to Tim Berners-Lee, 
the W3C Director, who has final say in all W3C process appeals.

Cheers,
James Salsman