Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 14 Nov 1997 03:37:28 -0800
From:      Julian Elischer <julian@whistle.com>
To:        sos@freebsd.dk
Cc:        nate@mt.sri.com, tlambert@primenet.com, hackers@FreeBSD.ORG
Subject:   Re: SUID-Directories patch
Message-ID:  <346C37F8.62319AC4@whistle.com>
References:  <199711141054.LAA01685@sos.freebsd.dk>

next in thread | previous in thread | raw e-mail | index | archive | help
S?ren Schmidt wrote:
> 
> In reply to Julian Elischer who wrote:
> > S?ren Schmidt wrote:
> > >
> > yeah I've inflicted SO MUCH on you
> > sd.c aha1542.c aha1740.c st.c /sys/i386/boot/biosboot/*(50%)
> > fdisk, ref.tfs.com etc. etc. etc. etc. etc.
> 
> Well, you wouldn't want me to comment on this now would you :)

Why not? when BSD desperatly needed a SCSI system, I had to fight
all the way up from floppies to get to it and by the time I finished
there was one that was sufficiently general to last till the present.
I'm looking forward to the CAM stuff, but that doesn't draw from the
fact that the BSD SCSI system was miles ahead of what was avaliable at
the time, and I had to fight bloody hard to get it into BSD.
(corporate lawyers don't often like employees ding that sort of thing)

If you think I had an easy time convincing purists that using the
BIOS for booting was the 'right thing', then you weren't watching.

"Greatness is achieved on the shoulders of others".
This implies that others come after us ond improve what we have done.
In the rear view mirror,it often looks like 
"It should have been done that way to start with"
but we can allsay that looking back.

> 
> Just an example: DEVFS....

And what's so bad about devfs and divert sockets(mentionned in your 
other mail)?
Divert sockets came directly from a conversation with
Kieth Sklower of the CSRG who wanted a way of bringing some 
packets out of the kernel for processing because there was too much 
in there already. It's an anti-kernel-bloat tool, and it works great.
You've never needed it obviously. (actually DIVERT is not my baby I 
just checked it in for someone else)

If I ever got the help I've asked for so many times 
devfs would be finished by now.

It's really the only way to go in the future of fully 
loadable devices, and I really feel let down by the lack of support 
I got from everyone on it. 

I've asked so many times for people to read it and help me with
parts of the VFS that I don't understand, any I got some patches 
from 1 person, and PHK has looked at it for a short time.

I've had real-life (TM) including a divorce, a 3 year time crunch
and a startup company,
to contend with and I think it's a it bloody cheak to bring up devfs
as a bad thing considerning no-one else has even TRIED to solve the
probelm it tries to solve, and no-one else has helped either.
I'd like nothing better than to have the time to do more on it.
(which incidentally I am getting a bit of now)

I've bet my entire career on freeBSD, got jobs for several
others doing the same and tried at every turn to push it forward.
It's appreciation like this that make me feel like throwing the 
whole thing in and going to take that job streamlining NT I keep 
getting told about. I won't of course because I have a vision
of what can be done here. 

I realise that others have lives too. So should you. I've done as
much as I could given the cicumstances. Your support is appreciated.

julian



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?346C37F8.62319AC4>