Re: Dynamic IP hack (PR#294)

Erik Corry (erik@arbat.com)
Mon, 19 Oct 1998 10:25:44 +0200

On Mon, Oct 19, 1998 at 09:18:45PM +1300, Chris Wedgwood wrote:
> On Mon, Oct 19, 1998 at 10:12:56AM +0200, Erik Corry wrote:
>
> > Because most of these solutions will cause more dialling. We
> > don't want a situation where people complain that 2.2 is costing
> > them a lot of money because it dials when 2.0 didn't.
>
> When I was running with modular ipv4, I saw relatively few problems
> that would make me think existing applications would break.
>
> Sure - its not the same as `things that would cause it to dial' but
> perhaps an indication that this might be an option.

Loading an ipv4 module doesn't cost you 7 cents per time, so
users would be unlikely to notice if it happened often. Demand
dialling does, many places.

> > Also, the original dyn_addr hack is already in 2.1, it works
> > the problems are theoretical, and you can switch it off.
>
> It's a hack - and an ugly hack IMO. I don't want the code in my
> kernel (although, it's not all that much really).

It's not much, and if you don't like it, you don't switch
it on. Many people like it a lot.

> Anyhow - for 2.2.x is must remain,

There are two parts. The dynaddr part must remain, and the
RST-provoking part should go in, unless a better solution is
forthcoming.

> but I think that for 2.3.x maybe
> we should `solve' the problem by redefining it as a userspace and
> application level problem - not a kernel problem.

Be my guest.

-- 
Erik Corry erik@arbat.com           Ceterum censeo, Microsoftem esse delendam!


-
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/