Hi.
        This seems an interesting network. I would try to test the network
performance, say with "ping", "ftp", etc. I wonder why you have not a Linux
machine, which may act as an ftp server, and could provide a lot of
diagnostic tools.
        I would also try to detect when are collisions shown starting with
one/two machines, and turning on the others one at the time, but I suppose
you have done this already. If the W98 machine is causing something...
        Or may be the hub firmware gets totally confused with its dead
port. Try to swap it.
        BTW, what is the BeOS system? Is it made out of berillium? I
thought all OS's used today were silicon or gallium-arsenide based.

>
> I have a five computer network: two NT workstations, one Win95, one Win 98,
> and one BeOS system.  The systems all use 3Com 10Mb/s cards, except for the
> Be machine which uses a generic NE2000 compatible card.  One of the systems
> (running 98) is a notebook with a 3Com Etherlink III PCMCIA card.  One of
> the NT boxes has two 3c509's because I'm about to use it as a gateway to an
> ADSL router.  The whole mishmash of hardware uses a Linksys eight port hub
> which I already know has one dead port.
>
> I appear to be suffering from a tremendous number of packet collisions, if I
> can believe what the indicator on the hub is saying.  The network is an
> amalgamation of NetBUEI and TCP/IP, if that makes any difference.  I'm
> honestly not sure if the collisions are having an effect on the network's
> performance because it's been happening for quite a while.  Also, both NT
> machines have service pack 3 freshly installed, the 95 system is using OSR2,
> the 98 system uses RC0 and the Be computer is running release 3.
>
> There you have it.  Any ideas?  And, for that matter, do I actually have a
> problem?
>
> Thanks!
>
> Drew Dunn
> [log in to unmask]
>



************************************
Javier Vizcaino. Ability Electronics. [log in to unmask]
  Starting point:        (-1)^(-1) = -1
  Applying logarithms: (-1)*ln(-1) = ln(-1)
  Since ln(-1) <> 0, dividing:  -1 = 1     (ln(-1) is complex, but exists)