Filing naming weirdnesses in NT4SP3, 95/98

[Dan Kaminsky]

Here's a snippet of human readable logs from my DoxPrint system(I'll release
the source RSN, lemme update my documentation first!):

Dunne: 09-Mar-1999, 20:45:51, starting to print CWINNTProfilesAdministrator
for root at ciagon on IP 129.210.96.49, served by , Local = 1, Error Code 0
Dunne: 12-Mar-1999, 00:26:49, starting to print ROOT.b07785 for root at
ciagon on IP 129.210.96.49, served by , Local = 1, Error Code 0
Dunne: 12-Mar-1999, 00:26:57, starting to print MicrosoftWord-lab6-2.rpt for
root at ciagon on IP 129.210.96.49, served by , Local = 1, Error Code 0

("served by" is a bug I'll fix later, but essentially it would always say
DoxPrint.  The system generates both CSV and human readable logs, here are
results from the latter.)

Anyway, as should be distressingly obvious, Samba is picking *THREE*
different systems for creating a filename for NT...I have logs showing
behavior like this stretching back months.  I'd say 80% of prints use the
"ROOT.b07785" method, though.

I've tried to get 2 NT machines to connect.  Neither expand the %S in the
comment line, and only one is able to actually "connect" with the Samba
printers.  However, the other hasn't been tested since I implemented printer
definitions, so we'll see if this changes things.

There's another issue:

Dunne: 11-Mar-1999, 22:19:50, starting to print ROOT.a07785 for root at
ciagon on IP 129.210.96.49, served by , Local = 1, Error Code 0
Dunne: 11-Mar-1999, 22:20:03, starting to print ROOT.a07785 for root at
ciagon on IP 129.210.96.49, served by , Local = 1, Error Code 0
Dunne: 11-Mar-1999, 22:20:18, starting to print ROOT.a07785 for root at
ciagon on IP 129.210.96.49, served by , Local = 1, Error Code 0
Dunne: 11-Mar-1999, 22:20:30, starting to print ROOT.a07785 for root at
ciagon on IP 129.210.96.49, served by , Local = 1, Error Code 0

Note the file print name *NEVER CHANGES*.  My script sends this file out
through nwprint--if the jobs come in too fast, there's a possibility of
writeover.

I'll *try* to get some packet traces and instrumentation done this weekend.
I also plan to release DoxPrint, since my school has forced my hand in
releasing it here.

Anyway, I better sleep.

Yours Truly,

    Dan Kaminsky
    DoxPara Research
    http://doxpara.netpedia.net


--
If a vacuum cleaner doesn't suck, does it suck?