USB memory stick hotswap problems

Jeremy Chadwick freebsd at
Mon Sep 27 12:51:40 PDT 2004

Whoa, easy there.  I think Mike was just providing some (what may
be considered superfluous/obvious) feedback to the issue at hand; he
honestly stated up front he had no solution, which implies he's not sure
of how to go about fixing the problem at hand.

I viewed his "criticism" as more positive than negative; I took it to
read "this is something we definitely need to look at and resolve, not
just put on the back-burner."  I think the number of people who manage
and maintain the USB stack is, as you ptu it, limited.  I don't know
what happened to Nick Hibma (the original maintainer/porter of the USB
stack), but it seems things have changed hands.

Finally, albeit a bit off-topic: financial contributions don't always
make things happen quicker in the open-source world.  Although we have
less "managerial" red-tape to deal with, there are still only 24 hours
in a day, and many of us work full-time jobs.  Time is limited,
regardless if there's a donation involved or otherwise.  :-)

Just my $0.02.

| Jeremy Chadwick                                 jdc at |
| Parodius Networking               |
| UNIX Systems Administrator                   Mountain View, CA, USA |
| Making life hard for others since 1977.                             |

On Mon, Sep 27, 2004 at 12:39:04PM -0700, Brooks Davis wrote:
> On Sat, Sep 25, 2004 at 03:50:30PM -0500, Mike wrote:
> > Although I fully understand your reasoning, there is still a major
> > problem here that really should be resolved.  I don't claim that I
> > have a solution - but I think I have an example that really requires
> > for this to be solved in some way other than "you must unmount the
> > USB device before unplugging".
> You can make statements like this all you want, but at the end of the
> day all you do is delay the solution.  The few developers who are
> capable of fixing these sorts of problems understand the issues and
> would like to fix them, but the code is question is large, complex, and
> fairly crufty.  It also wasn't designed with this sort of scenario in
> mind.  The more time those developers have to spend reading messages
> from people demanding a fix and pontificating on the importance of
> the fix, the less time they can spend of the necessary architectural
> cleanups to actually fix the problem.
> If you want to do something to hasten the arrival of a fix, consider
> donating to the Foundation for work in this area or perhaps contributing
> to a developer like PHK when they solicit donations to enable them to do
> related work.
> -- Brooks
> -- 
> Any statement of the form "X is the one, true Y" is FALSE.
> PGP fingerprint 655D 519C 26A7 82E7 2529  9BF0 5D8E 8BE9 F238 1AD4

More information about the freebsd-current mailing list