[jcifs] maxMpxCount

[Dan Dumont]

I'm pretty darn sure something is wrong here.  If you recall, in the past, I
posted that I found where to stop the decrementing of maxMpxCount.  Mike
said something along the lines of "that would very quickly waste a vast
amount of resources... ".

But consider this.   I want to talk to three different computers via this
jcifs package.  So I fire up 3 threads, and they each make an smbFile, one
for each host I want to talk to.

Now, suppose one of these hosts was a slow host.  Your untouched code would
throttle the maxMpxCount down to 1, and all queries over the existing
transport would end up executing line-fashion. 

1 slow host destroys the performance of the transport.   Now, I'm not sure
why you did this, all of your threading examples work fine on 1 host where
this type of behavior makes sense.  But when you try to apply it to extra
hosts, bad things happen.

p.s. sorry, I meant to send this to the list the first time.