On Tue, 1 Jun 2010, Jonathan Rochkind wrote:

> Accept-Ranges, I have no idea, I don't understand that header's purpose 
> well enough. But SRU also provides a query param for that, it seems less 
> clear to me if that's ever useful or justifiable.

Accept-Ranges is a response header, not something that the client's 
supposed to be sending.

The client sends a 'Range' header (with an optional 'If-Range' if you're 
concerned with the resource having changed), and in response, the server 
sends a 206 status with a 'Content-Range' header.

See
     http://labs.apache.org/webarch/http/draft-fielding-http/p5-range.html

...

I only know of two values for 'Accept-Ranges' -- none (ie, I don't accept 
partial downloads) and bytes, so for incomplete downloads you can start 
where you left off.  If you know the file's excessively large, I guess you 
could use it to transfer it in parallel to abuse the TCP congestion rules. 
(or if you have a way of knowing that there are multiple mirrors, to 
spread the load across servers).

-Joe