[JBoss Microcontainer Development] New message: "Re: logging 2.2.0.CR1 and OnDemandDependencyUninstallTestCase"
by Ales Justin
JBoss development,
A new message was posted in the thread "logging 2.2.0.CR1 and OnDemandDependencyUninstallTestCase":
http://community.jboss.org/message/529911#529911
Author : Ales Justin
Profile : http://community.jboss.org/people/alesj
Message:
--------------------------------------------------------------
> I upgraded to logging 2.2.0.CR1 since I was getting some strange dependency/classpath issues
> due to a bizarre circular dependency in logging 2.1.0.GA.
>
> logging-log4j -> jboss-cl -> jboss-dependency -> logging-log4j
This is my "fault". :-)
The reason is the actual implementation of the TCLMCFilter:
* http://anonsvn.jboss.org/repos/common/common-logging-log4j/branches/2.1/s...
Which does deployment matching based on proper API, instead of wild string guessing (previous impl).
But afaik, main code (not tests) should only depend on logging-spi,
where for tests you exclude the "jboss-cl" part -- then it should work with np.
> Its a bit of a bizarre test to me (playing about a private logger to see if something gets logged), why not just attach an appender?
> I've removed the test for now. If you think it is an important test, I'll let you figure out how to recreate it using public api. ;-)
This is explained in more details in some other thread.
The issue is that we suppress the exception, only posting a warning -- the only way to know if the wanted behavior occured.
Kabir, can you please have a look at this?
ps: what's the issue with the new test, I didn't understand that?
--------------------------------------------------------------
To reply to this message visit the message page: http://community.jboss.org/message/529911#529911
16 years, 1 month
[JBoss Microcontainer Development] New message: "Structure deployer cannot access manifest"
by Thomas Diesler
JBoss development,
A new message was posted in the thread "Structure deployer cannot access manifest":
http://community.jboss.org/message/529906#529906
Author : Thomas Diesler
Profile : http://community.jboss.org/people/thomas.diesler@jboss.com
Message:
--------------------------------------------------------------
This code used to determine whether a bundle gets deployed
public boolean determineStructure(StructureContext structureContext) throws DeploymentException
{
VirtualFile root = structureContext.getRoot();
// This file is not for me, because I'm only interested
// in root deployments that contain a MANIFEST.MF
Manifest manifest = VFSUtils.getManifest(root);
if (root != structureContext.getFile() || manifest == null)
return false;
// This file is also not for me, because I need to see Bundle-SymbolicName
Attributes attribs = manifest.getMainAttributes();
String symbolicName = attribs.getValue(Constants.BUNDLE_SYMBOLICNAME);
if (symbolicName == null)
return false;
....
With deployers-2.2.0.Alpha3 (i.e. jboss-vfs-3.0.0.CR2) the manifest returned from VFSUtils is null.
Is this a bug, perhaps related to the root not mounted?
--------------------------------------------------------------
To reply to this message visit the message page: http://community.jboss.org/message/529906#529906
16 years, 1 month
[JBoss Microcontainer Development] New message: "Re: ClassLoadingAdmin"
by Thomas Diesler
JBoss development,
A new message was posted in the thread "ClassLoadingAdmin":
http://community.jboss.org/message/529904#529904
Author : Thomas Diesler
Profile : http://community.jboss.org/people/thomas.diesler@jboss.com
Message:
--------------------------------------------------------------
Eventually we will need a notion of try run for module resolution. The concept of application also needs to be supported - not sure if this can aleady be done by the Module abstraction.
The use case is:
>From a gui select the application located at some repository to download/install into the framework.
An application consists of potentially very many bundles.
Try run the resolver using the application metadata.
Only when the application can be resolved in the running framework, actually do the install.
The worst case, which must be avoided is a partial install that fails and leaves the framework in a state where random bits are installed/resolved that cannot easily been uninstalled. Essentially, the resolve step for a set of bundles must be atomic in nature.
--------------------------------------------------------------
To reply to this message visit the message page: http://community.jboss.org/message/529904#529904
16 years, 1 month