[starksm64]

"alesj" wrote:

How come this didn't pop-up before?

As PS tests should see this, when using DeploymentManager to do remote deployment.
Or is the thread that takes the stream, that does the suspend and resume?

The PS tests were disabling hot deployment at the profile level, so that the getModifiedInfo() always was returning an empty list. The hdscanner therefore never would interfere with changes to the repository.

[alesj]

But that flag might be set too late,
once getModified is past that check, our suspend is invoked.
Hence our partial deployment could be picked as one of the appURI's children.

[starksm64]

Right, and it also required exposing locking detail on the repository, so its a bad solution.

[alesj]

Right, and it also required exposing locking detail on the repository, so its a bad solution.

What about this Mutex example from:
- http://java.sun.com/j2se/1.5.0/docs/api/java/util/concurrent/locks/AbstractQueuedSynchronizer.html

I think this would do,
as we don't really need exclusive read vs. write locks.