Re: [Quantlib-dev] long and possibly inifinite observer/observablenotifying loops

[Luigi B. <lui...@gm...>]

On Thu, 2007-03-22 at 16:53 +0100, DU VIGNAUD DE VILLEFORT FRANCOIS
GASAPRD PHI wrote:
> Sorry for this wrong analysis I thought that once a LazyObject has
> been updated it doesn't propagate any notification until it has been
> recalculated. 
> Don't you think that this should break any circular chain and solve
> the pb of reconvergency ?

No, it wouldn't work. If it didn't forward the notification, it might
not be asked for recalculation at all (see my previous post.)

Luigi



---------------------------------------- 

The first rule of intelligent tinkering is to save all the parts. 
-- Paul Erlich