Malke wrote:
Quote:
>...
> You must use the same passwords and the same user accounts. Of course
> it makes a difference. The ones you are entering are *not* correct if
> they are different. In a peer-to-peer network (no server),
> authentication is done on the local machine since there is no domain
> controller. This means that the machine hosting the shared resource
> must have the identical user account with identical password on it to
> authorize access from the machine requesting the resource. User
> account "John" with password 1234 is *not* the same as user account
> "John" with password 5678 and therefore authorization will be denied.
This is obviously *not* correct! How could the local machine decide
whether a user-id and the account is valid on the *other* machine? This
would widely open all doors for intrusions and other malicious
activities. A connecting system clearly has to send the credentials over
to the other system, so that they can be validated *there* and
consequential access be granted or not. Or why would Windows else
ever need to ask for credentials when accessing some share, if it could
only handle that <same workgroup plus same user-id plus same pwd>-case?
I regularly access shares on other system (using Ethernet and Wireless
LAN) where I have different user-ids and/or different passwords (and
on some I *do* have the same user-id but different passwords!). It is
always only, if I need to connect via Bluetooth (i.e. if there is no
other networking infrastrcture) that I have such headaches with
authentication!
The problem seems to be, that Windows (XP and Vista) does some
stupid business in *caching* things, i.e. if it finds credentials for a
(supposedly) already known device it tries these and if things fail
using these, it immediately reports "access denied" instead of asking
for the credentials (that might have changed since the last access)
again.
These cached ids and credentials are a constant nuissance! I have
already scripts to automatically delete them whenever I reconnect the
two systems. While these often help, alas, too often that seems to be
not enough. There must be some further mechanism, that likes to
interfere here...
And Vista even worsened things with this absolutely idiotic
classification of networks into "Local only" and "Local and Internet"
which it tends to get practically always wrong, esp. on multi-homed
machines (i.e. machines with multiple network interfaces). I always have
to endlessly fiddle with de- and re-activating network adapters,
changing "Public networks" into "Private networks" and "Merging
networks" (what a fantasticly studpid idea!!!) and what not, until - at
some point - Vista finally decides to cooperate.
Michael