Brad,
Content-Length should be preserved if the client asks for it. There isn't
much point in sending content-length if you are going to shut the connection
down afterwards anyway. It is always preserved between the two backhand
servers. There was a case that wasn't handled correctly. If the client asks
for 1.1 but doesn't explicitly ask for kepp-alives, it would shut down the
connection.
This has been fixed in CVS too. But, I found that it doesn't trigger with any
of the clients I tried. Which client is triggerring this behaviour for you?
Brad Fitzpatrick wrote:
> Similar to problem I just posted about earlier, it seems that
> content-length headers aren't being transferred from the backhandee and
> the backhander.
>
> Are only certain headers passed along, or is there a reason that
> content-length isn't preserved?
--
Theo Schlossnagle
1024D/A8EBCF8F/13BD 8C08 6BE2 629A 527E 2DC2 72C2 AD05 A8EB CF8F
2047R/33131B65/71 F7 95 64 49 76 5D BA 3D 90 B9 9F BE 27 24 E7