Paulo Gaspar wrote:
>
> Messing up with that "performance waste" is a bit delicate. It
> involves design decisions.
Couldn't you take one of the following approachs :
Since a resource is managed entirely by it's loader, why not let the
loader simply read the timestamp of the resource during
getResourceStream() and remember it. This keeps it atomic-ish. Then :
1) ResourceManager would get that timestamp and set it in the resource
in getResource() like it does now, via a different call, such as
resourceLoader.getResourceModified( resource )
OR
2) Let the resource itself get the timestamp when it calls
getResourceStream() from it's loader. This keeps the operation atomic
as far as the ResourceManager is concerned.
Either way, having the Loader maintain it seems the best thing, and
neither of the above requires any serious design issue, right?
geir
>
> Anyway, I will fix the bug, but I am afraid I don't know how
> to make patches.
> =:o/
>
> I am taking a look at cygwin to see if I can figure out how to
> do it.
>
> Have fun,
> Paulo
>
> > -----Original Message-----
> > From: Jason van Zyl [mailto:[EMAIL PROTECTED]]
> > Sent: Monday, February 19, 2001 23:23
> >
> >
> >
> > Thanks, noted. I won't get to this for a while so if
> > you like:
> >
> > 1. fix it
> > 2. test it
> > 3. send the patches please
> >
> > :-)
> >
> > jvz.
> >
> > On Mon, 19 Feb 2001, Paulo Gaspar wrote:
> >
> > > .........
--
Geir Magnusson Jr. [EMAIL PROTECTED]
Velocity : it's not just a good idea. It should be the law.
http://jakarta.apache.org/velocity
