cleaned up in trunk.
-igor
On Feb 13, 2008 2:57 PM, Ned Collyer <[EMAIL PROTECTED]> wrote:
>
> You are correct :).
>
> I would generally discount any overhead associated with a cache map lookup
> (because thats the cache, which is meant to be the quick bit, and its only
> going to happen the first time)
>
> I brought it up specifically because of
>
> PropertiesFactory.load
> which has 3 places where it can return.
>
> 1. returns null or the actual object
> 2. returns the actual object
> 3. returns null
>
> Rgds
>
> Ned
>
>
> igor.vaynberg wrote:
> >
> > a contract of the cache is that
> > if cache.put(k,v); then v.equals(cache.get(k)) or v is null
> >
> > so if you just created v and put it into cache, why also introduce the
> > overhead of the lookup?
> >
> > this style of code also has a single return point:
> >
> > V v=cache.get(k);
> > if (v==null) {
> > v=new V();
> > cache.put(k,v);
> > }
> > return v;
> >
> > its just coding preference, we seem to like the example above more.
> >
> > -igor
> >
>
> --
> View this message in context:
> http://www.nabble.com/Ideas-on-cache-tp15452545p15469853.html
>
> Sent from the Wicket - Dev mailing list archive at Nabble.com.
>
>
