[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Multiple cards supported for 2.0.33?

sven@sven.cp.uconn.edu writes:
 > Hello token-ring group,
 > Wouldn't it be nice if you could deploy token-ring in informal
 > conversations at work.  Seems much of this conversation consists of
 > listening to people blurt out information (often irrelevant beliefs or
 > achievements of their life) as if only to control the ether.  If I could 
 > only programmaticly reduce the time the token can be help by any one
 > individual.... When I think of my friends, I realize they are all
 > token-ringers.  I digress.

I agree. Interpersonal conversation seems to follow the CSMA/CD
protocol. We'd get better bandwidth utilization with token-ring. I'm
not sure the mind can assimilate information that quickly,
though. Would that be 'receiver congestion'? Then we'd start
discarding frames and requesting repeat transmissions. Discarded
frames and tokens all over the floor. Not a pretty sight. Hmmm.

 > Paul Norton mentioned that a multiple token-ring adapter patch for 2.0.32
 > should exist. Anyone know of a location?  A little off-beat, but I am
 > still wondering if people think any versions of 2.1 kernel, supporting
 > multiple token adapters, are stable enough to be used in a
 > deployed environments?


I use and develop under 2.1.x almost exclusively. Obviously some
versions are more stable than others. In most, maybe even all, 2.1.x
versions there is _something_ broken. This is to be expected - it's a
development series after all. It's the bleeding edge. 

That being said, 2.1.77 works fine for me. But I'm not an average user
and I'm probably not using a common subset of kernel
functions. These later 2.1.7x kernels seem to have problems with
networking and sound. Some 2.1.4x kernels would actually trash your
filesystem, of all things. That's when the dentry stuff went
in. That's bleeding edge, Quake II style. 

If you really want to keep your finger on the stability of the
development kernels you'll need to subscribe to the linux-kernel
mailing list. That's where a lot of the problems are first reported
and a lot of fixes first appear. 

 > Sven