It could probably be done (was done to the manager). Not that it's used in *that* many places, and it's only one step away from the manager. Same could be said of the Resolver (used even less, though) On Sat, Dec 6, 2008 at 12:04 AM, Gavin King <gavin(a)hibernate.org&gt; wrote: > Is there any reason why we can't just make the MetaDataCache object a > static singleton? > > -- > Gavin King > gavin.king(a)gmail.com > http://in.relation.to/Bloggers/Gavin > http://hibernate.org > http://seamframework.org > _______________________________________________ > webbeans-dev mailing list > webbeans-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/webbeans-dev > -- --- Nik _______________________________________________ webbeans-dev mailing list webbeans-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/webbeans-dev

Is there any reason why we can't just make the MetaDataCache object a static singleton? -- Gavin King gavin.king(a)gmail.com http://in.relation.to/Bloggers/Gavin http://hibernate.org http://seamframework.org _______________________________________________ webbeans-dev mailing list webbeans-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/webbeans-dev

I think MetaDataCache is probably a different case, but I agree on ManagerImpl.

I think we got our wires slightly crossed earlier - when I said that Manager could be made a static thing,

I was meaning for *integration* purposes only! i.e. the Manager used for EL resolution and injection into Servlets would be a static thing. There are definitely multiple managers per application, now that we have child managers. And I agree that the internal usage of the static reference makes shit more difficult to test. On Fri, Dec 5, 2008 at 8:49 PM, Matt Drees <matt.drees(a)gmail.com&gt; wrote: > I'll speak up about one concern I have. If MetaDataCache and > ManagerImpl > are stored in static references, I imagine it will be difficult to > perform > simultaneous/multi-threaded integration testing for webbeans apps. > For > example, with Seam, you can't have two SeamTests running at the > same time in > the same JVM, because Lifecycle.java has a static reference to the > application map, and the two tests would interfere with each > other. I'd be > sad if the same thing occurs for the webbeans ri. > > Now, maybe you could could use some ThreadLocal magic or something > to avoid > having to pass the ManagerImpl/metadatacache around everywhere in > method > calls. But my preference would be that the webbeans RI doesn't use > static > singletons at all. > > What do you guys think? > > -Matt > > > On Fri, Dec 5, 2008 at 5:04 PM, Gavin King <gavin(a)hibernate.org&gt; > wrote: >> >> Is there any reason why we can't just make the MetaDataCache >> object a >> static singleton? >> >> -- >> Gavin King >> gavin.king(a)gmail.com >> http://in.relation.to/Bloggers/Gavin >> http://hibernate.org >> http://seamframework.org >> _______________________________________________ >> webbeans-dev mailing list >> webbeans-dev(a)lists.jboss.org >> https://lists.jboss.org/mailman/listinfo/webbeans-dev > > > _______________________________________________ > webbeans-dev mailing list > webbeans-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/webbeans-dev > > -- Gavin King gavin.king(a)gmail.com http://in.relation.to/Bloggers/Gavin http://hibernate.org http://seamframework.org _______________________________________________ webbeans-dev mailing list webbeans-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/webbeans-dev

Is there any reason why we can't just make the MetaDataCache object a static singleton? -- Gavin King gavin.king(a)gmail.com http://in.relation.to/Bloggers/Gavin http://hibernate.org http://seamframework.org