Subject: RE: vague proposal for new scheduler primitive: asynchronous "sleep"
To: None <firstname.lastname@example.org>
From: Martin Husemann <email@example.com>
Date: 05/10/1999 09:43:20
I long hesitated to ask, but this discussion triggered a related topic
First, let me state there is absolutely no pun intended, I don't blame
anyone, I'm not criticising anything - I'm just for pragmatic solutions!
There is IMHO massive interest in multi-threaded kernels and SMP, and later
on userland threads with propper kernel support. This interest is growing,
as more people get their hands on SMP machines (dual PII boards are dirt
Any attempts from volunteers to do work in this direction has always been
stoped by the "work in progress that noone is allowed to talk about".
Volunteers generally don't like to reinvent the wheel (and for the NetBSD
ones this is an even stronger argument).
I'd like to ask if its not time to set up an open SMP/multi-thread mailing
list and start collecting volunteers to work on this. IF there has been
something useable produced by that famous shadow project, let those who know
it speak up and show something - if not, ignore it. (Please, there is realy
no pun intended!)
I don't think anyone likes the second alternative, but a project with no
output at all (at least from the outsiders point of view) multiple years
after the scheduled deadline (which may have been a soft one, how should we
know?) should not be allowed to prevent all further progress in the
direction it was supposed to push NetBSD.
Just MHO, of course, as an ordinary user.
P.S.: I'm aware of the flame-war potential this has, so I deliberatley
restricted the distribution to tech-kern