Re: estimated Content-Length with chunked encoding

Adrien de Croy wrote:
> 
> for this point, you want my draft ID on progress notifications :)
> 
> http://tools.ietf.org/id/draft-decroy-http-progress-04.txt

Although I agree with a desire to keep the request lightweight, so only
to pass on the information when desired, the chunked scheme is nearly
always used for progressive data.  If you amended the draft to use an
attribute extention on the chunk header, it might be very interesting.