I just vaguely thought it might be related and maybe useful in relation to
optimizing the process of 'cloning' objects that Kent talks about below. But
I'm completely out on deep water, and this could be totally unrelated, so
please disregard this. I'll stop spamming! Sorry for this noise.
Cheers,
Marc
On 2012-06-25, at 15.04, Niclas Hedhman [via Qi4j-dev] wrote:
> Good article on Memory Mapped I/O, but exactly what do you have in mind??
On 2012-06-24, at 23.00, Kent Sølvsten [via Qi4j-dev] wrote:
> eg the reference to the prototype is still valid after creating a value
> (inst1). The downside is that we have to clone the prototype always. If we
> instead require the user to fetch a new reference to the prototype:
>
> ValueBuilder<X> b = vbf.newValueBuilder(X.class);
> X p1 = b.prototype();
> p1.att1().set("value1");
> X inst1 = b.newInstance(); // p1 is now immutable
> X p2 = b.prototype();
> p2.att1().set("value2");
> X inst2 = b.newInstance();
>
> it would be possible to avoid cloning in the common cases - it should be
> pretty straightforward to let newInstance() return the prototype (and making
> it immutable), and then clone (and make mutable) on the next invocation of
> b.prototype().
> Would that be acceptable?
>
> The cloning itself could probably be optimized too - but I think that would
> be of minor importance, if the change above is implemented.
--
View this message in context:
http://qi4j-dev.23929.n6.nabble.com/Test-Failure-in-DCI-sample-tp6925p6954.html
Sent from the Qi4j-dev mailing list archive at Nabble.com.
_______________________________________________
qi4j-dev mailing list
[email protected]
http://lists.ops4j.org/mailman/listinfo/qi4j-dev
