Stas Bekman wrote:
> Catching up with httpd-dev: So should we adjust our docs/tests to strip
> the input Range headers too, in addition to C-L in output filters? I'm
> not quite sure I understand what are the circumstances when this needs
> to be done.

no, it is not required to strip Range at the moment - the short answer is to
just forget about that thread :)

the long answer is that the issue in the thread is that currently content
handlers need to serve up all their content, even when byteserving, so that
the byterange filter has all the data and can work.  the suggestion made in
the thread is essentially to let content handlers (more specifically
mod_proxy, which isn't a content handler) do the byteserving themselves, in
which case they would need to let future filters know that only the
specified Range is actually coming down the brigade, not the entire content
for the request.

to really get a feel for what is going on, see graham's response to one of
my posts early in the thread

  http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=108972633012892&w=2

this post seems to sum up where the thread was headed (and the dominant
httpd-dev mindset) before it was restarted

  http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=108973604000918&w=2

so basically, I wouldn't worry about Range at all, at least not at the
moment - we may need to completely redesign our filter API if any of these
proposed changes become real, but I really don't see that happening in the
near future, certainly not with 2.0 and probably not even in 2.1.

--Geoff

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to