[Olsr-dev] Seeking comments: OLSR+ETX v/s DSR+ETX

[sebastian]

Wed 23 Jan 2008 14:13, Henning Rogge wrote:
> On Jan 23, 2008 2:07 PM, sebastian sauer <(spam-protected)> wrote:
> > Fri 21 Dec 2007 21:42, Hannes Gredler wrote:
> > > Henning Rogge wrote:
> > > > What's about this ? It's not as "fast" as the old formula, but I think it works:
> > > >
> > > > lq = (total - lost) * / (  (total + olsr_cnf->lq_wsize) / 2 );
> >
> > also here, lq is not necessarily btwn. 0 and 1.
> > the same voodoo so to speak ;)
> No, it is between 0 and 1...
NAK. just do e.g. the algebraic proof to see that this is just wrong.

> total is always smaller as lq_wsize... so it's always smaller as the
> average of total and lq_wsize.
ok, now you are no longer refering to the formula, but to the current code.
infact the lq_wsize is a completely redundant variable.


also every node in the network can have a different lq_wsize, therefore a
different bias in the lq meassurement. just the next ugly thing due to
voodoo.

s.