Re: PUBLIC CHALLENGE: (was RE: devfs again, (was RE: USB device a

danielt@digi.com
Fri, 8 Oct 1999 09:56:14 -0500 (CDT)


On Fri, 8 Oct 1999, Alexander Viro wrote:

>
>
> On Fri, 8 Oct 1999, Nathan Hand wrote:
>
> > - UNIX-like /dev without UNIX-like rw fs (good for embedding on romfs)
> Point taken, but without rw filesystem you are going to have a
> nasty time with _many_ other things, including the persistency in devfs
> itself.
>
If you do not have a rw filesystem you hard-code your permissions
anyway. This sort of system does not resemble your workstation.

> > > Also: It is extra code, has to be maintained and updated, and has to be
> > > accounted for in new driver developments. It _will_ add new bugs, even new
> > > classes of bugs. This doesn't come for free.
> >
> > Well, perhaps all kernel developers should stop coding right now: you have
> > equally well argued against all new features and drivers.
>
> So all coding goes into new features and to hell with cleaning the mess
> up, right? Damn, I understand Theo. Features are nice _iff_ you have a
> half-decent implementation. Featuritis for its own sake leads to mosters
> a-la NT and EMACS. Sheesh...
>
And without devfs, we end up with a dozen different implementations
of it under /proc. Or we end up with SCO Openserver-style device
management. Or each device driver writer has to invent their own
device management system for dynamic devices.

It doesn't _have_ to be Richard Gooch's Devfs but I have yet
to hear a better idea, or see alternative running code for the
general case.

-- 
Daniel Taylor      Senior Test Engineer     Digi International
danielt@digi.com                             Open systems win.

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/