it is supected of being the cause of the race in #1675 amongst other
evils.

With backend waiters it makes no sense to bring it back.

Fixes: #1726

In future versions of HTTP requests can be terminated abnormally
without ditching the entire session, and this change starts the
transition to that logic.

Fixes: #1732

and plunk it down in req_fsm for now.

Only allow VDP_ungzip to set a length if it is the top VDP.

consuming ESI string as we go, rather than the other way around.

While passing by, update the text and formatting to
our regular style.

With PROXY protocol support, the behaviour of varnishd startup argument
-a changed.

Reflect this in the man page and adjacent documentation.

This is obviously different than in the original patch, a change
which I seem to have missed.

Spotted by Dridi.

Time to start writing down what will be part of 4.1.

a unstable beast :-/

platforms, but I'm not quite sure I know why...

This variable will be true when the response object's body will be
streamed from the backend.

Fixes: #1729

The real fix should be in configure

If the client sends [LO]-HI range, we trust it knows something we don't (yet),
and do the Range request.  If we later ran out of data, we close the session.

Other range requests on streaming objects are ignored.

Fixes: #1506
Fixes: #1618

so the SessClose VSL can reporte it correctly.

Probably broken by: Martin ?

on streaming objects.

Fixes #1730