5 Apr
2021
5 Apr
'21
10:39 p.m.
On Mon, Apr 05, 2021 at 03:30:25PM -0400, Clem Cole wrote:
On Mon, Apr 5, 2021 at 2:08 PM John Gilmore
<gnu(a)toad.com> wrote:
In 1999 I was program committee chair for Linux Expo which was trying to
sort of be Usenix for Linux.
I had been a paper reviewer for a number of Usenix conferences (I still
have my review notes for those papers, I was a cocky pain in the ass
that was full of myself, dunno why anyone put up with my nonsense but
they did).
I had the pleasure of reviewing this:
http://mcvoy.com/lm/papers/rtlmanifesto.pdf
The summary is Victor slipped a real time kernel underneath Linux and
ran all of Linux, kernel and userspace, as the real time idle process.
It worked fantastically well and was really clever. If you understand
time sharing and you understand hard real time, trying to shovel hard
real time into a time sharing system is like trying to prove 1 + 1 != 2.
They are trying to do two completely different things and if you are
good at one, you'll be bad at the other. Victor showed a way to have
your cake and eat it too, it was brilliant.
The paper was rejected and I believe it was partially because Victor
wasn't well known in Unix circles, he wasn't "in the club". If Bill
Joy
had written the same paper I'm 100% sure it would have been published.
Which lead to me campaigning for blind reviews.
Back to Linux expo, it was a success, some of the papers weren't up to
Usenix standards, but many of them were (I can dig up the proceedings
if anyone cares).
Whoever was running Usenix contacted me after seeing the success of
Linux Expo and begged me to bring those people to Usenix. I was offered
a board seat, I could be reviewer for life, anything I wanted.
All I asked for was blind reviews. Didn't ask anything for myself, just
blind reviews so you could be a nobody and get judged on the quaility of
your work rather than your name.
Apparently, I could get anything I wanted except that. I've had nothing
to do with Usenix ever since.
Clem came after whoever it was and has told me he cleaned things up quite
a bit. If I had known him back then maybe I would have come back.
The paper was rejected by the program committee,
on the objection that
"ports aren't research". So
the pro-academic, anti-engineering mindset was already in place back then.
IMHO,
USENIX walked away from a group that they should have been
trying to cultivate.