Re: Needing state in NTLMSSP

Henrik Nordstrom <hno@squid-cache.org> writes:

> tor 2003-01-16 klockan 09.35 skrev Kinkie:
>
>> We didn't really make it stateless, we just thought it best to leave as
>> much status as possible in the client. One problem we have is that it's
>> guarranteed that the connection to the client will be shut down between the
>> negotiate and challenge phases, because we shut it down.
>
> ???
>
> To my knowledge we only shut down the connection in the proposal packet,
> before NTLMSSP starts. Once the negotiate packet has been received from
> the client the connection is kept persistent.

Whoops, you're right.

> The termination after the authentication proposal is quite likely
> unnecessary and causes a bit of problems with request entities
> currently.. (but the real cause to these problems needs to be fixed
> anyway, making it persistent only hides the real problem there).

Possible.[1]

Footnotes:
[1] Neverending years of damnation to whoever perverted HTTP into
being stateful!

-- 
	kinkie (kinkie-squid [at] kinkie [dot] it)
	Random fortune, unrelated to the message:
Why are these athletic shoe salesmen following me??