[jcifs] Why would auth work on one server and not on another?

[Dave Mobley]

Thanks guys, I tried Allens suggestion (even though his was string to
byte[], and I needed the other direction...made the
adjustment....and.....DRUMROLL......  I got Mozilla 1.4 to log in and give
me back my username in my servlet (it succeeded in doing the smblogon....no
exceptions)

Now I just need to figured out why IE is hiccuping...this is supposed to be
IE's natural ability.  Any ideas on how I could go about figuring this out?
(I don't know enough about the http protocol to know where to look at the
moment.)

As always, you guys are awesome with helping out, and if someone ever needs
a jdk 1.1 version of this package, please let me know.

thanks,

Dave Mobley
scprotz at hotmail.com

----- Original Message ----- 
From: <eglass1 at comcast.net>
To: "Allen, Michael B (RSCH)" <Michael_B_Allen at ml.com>
Cc: "'Dave Mobley'" <scprotz at hotmail.com>; <jcifs at lists.samba.org>
Sent: Friday, July 11, 2003 7:30 AM
Subject: RE: [jcifs] Why would auth work on one server and not on another?


>
>
> > > 3.  (THIS IS A BIG ONE) Mozilla REALLY REALLY wants to try to
authenticate
> > > with my servlet (even though it isn't the nice automatic NTLM - it
requires
> > > me to enter my credentials)
>
> I'd have to look at the Mozilla source, but it sounds like they are
supplying
> gathered credentials to SSPI, rather than using the default credential
handle
> (which authenticates using the current logged-in user).
>
> > > Any suggestions on a 1.1 substitute for "UnicodeLittleUnmarked"? (Or
will I
> > > have to write some sort of byte converter, and then convert it to a
String
> > > with another encoder that is much more jdk1.1. friendly)?
> > >
>
> You could see if "UTF16-LE" works.  According to this:
>
> http://java.sun.com/j2se/1.3/docs/guide/intl/encoding.doc.html
>
> it is mapped directly to UnicodeLittleUnmarked, and UTF16-LE is one of the
> encodings required to be supported in JDK 1.3+ (I think there was some
question
> as to whether the byte-order mark might be optional in UTF16-LE, rather
than
> definitely unmarked, which I'm guessing is why jCIFS uses
UnicodeLittleUnmarked
> instead of UTF16-LE).  Of course, that's no guarantee that it's supported
in
> 1.1.
>
> Eric
>