[Oberon] ETHO 'holds' TCP connection.

[eas lab]

ETHO's Mail & USEnet free-form-interfaces are so much better than
the inflexible pre-determined normal alternatives.
But for unknown reasons my NNTP is often slower and even "hangs"
sometimes.

Now I see:
-> lsof | grep obe |grep TCP | awk '{print $1 ":" $2 ":" $8 $9}' ==
linux.obe:9288:100.84.235.103:59049->213.239.209.88:nntp(ESTABLISHED)
linux.obe:9288:100.87.22.6:59022->213.239.209.88:nntp(ESTABLISHED)
linux.obe:9288:100.87.22.6:59023->213.239.209.88:nntp(ESTABLISHED)
linux.obe:9288:100.84.235.103:59046->213.239.209.88:nntp(ESTABLISHED)
linux.obe:9288:100.84.235.103:59047->213.239.209.88:nntp(ESTABLISHED)
linux.obe:9288:100.78.139.55:39596->213.239.209.88:nntp(ESTABLISHED)
linux.obe:10193:10.170.96.162:58466->213.239.209.88:nntp(ESTABLISHED)
linux.obe:10193:10.203.48.170:41384->213.239.209.88:nntp(ESTABLISHED)
linux.obe:10193:10.203.33.66:35201->213.239.209.88:nntp(ESTABLISHED)
linux.obe:10193:10.170.96.162:58465->213.239.209.88:nntp(ESTABLISHED)

This means that 2 of LEO, which were recently on-line are still "ESTABLISHED"

Other inet facilities, of linux, which were used and done/disconnected
don't show, as I'd expect, and the "still-connected" from PC to rPi shows:
sshd:7519:192.168.0.1:ssh->192.168.0.2:37236(ESTABLISHED)

I expect the NetSystem.* to be rather complex.
Is there a simple way to debug/fix this?

How did old N-O handle the critical time-sharing?
With LEO the display is frozen while NetSystem is connected, but mouse still
is active. I thought display and mouse had the same rank/priority; ie. each
Frame has its own display and mouse handler?

== Chris Glur.