July 2008 - jboss-dev-forums - Jboss List Archives

[Design of JBossCache] - Re: Service threads

by manik.surtani＠jboss.com

I agree that this should be a separate discussion. :-) Why don't you start one based on your comments above - which are very valid? For the purposes of this discussion though, how much flexibility are we looking for in terms of configuring/sharing/providing an executor service? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165125#4165125 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165125

17 years, 8 months

1
0
0 / 0

[Design of POJO Server] - Re: Making Creators declare supported annotation types

by emuckenhuber

"skajotde" wrote : | Technical remark ;) Preferred way (by me and described in Effective Java) is to return always collections, empty or not emty instead of null ;) | Hehe, actually i meant 'nothing' by saying null. Though i preferred EMPTY_SET for whatever reason ;-) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165122#4165122 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165122

17 years, 8 months

1
0
0 / 0

[Design of JBoss Portal] - Re: portal clustering problem when using optimistic locking

by prabhat.jha＠jboss.com

Sure. I will give it a shot. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165114#4165114 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165114

17 years, 8 months

1
0
0 / 0

[Design of POJO Server] - Re: Making Creators declare supported annotation types

by skajotde

sigh, handling links is broken in this forum on firefox (2.0, 3.0). It's correct link: http://java.sun.com/j2se/1.5.0/docs/api/java/util/Collections.html#EMPTY_... View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165113#4165113 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165113

17 years, 8 months

1
0
0 / 0

[Design of JBoss Portal] - Re: portal clustering problem when using optimistic locking

by bstansberry＠jboss.com

Prabhat, Can you try your tests with your cache configured for READ_COMMITTED instead of REPEATABLE_READ? Once a tx reads an item from the 2nd Level Cache, Hibernate caches it in the Session object and doesn't need to go back to the 2LC for that same object. So there should be no benefit to REPEATABLE_READ. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165111#4165111 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165111

17 years, 8 months

1
0
0 / 0

[Design of POJO Server] - Re: Making Creators declare supported annotation types

by skajotde

"emuckenhuber" wrote : and null for Method and Field annotations. | | Any further ideas / comments ? Technical remark ;) Preferred way (by me and described in Effective Java) is to return always collections, empty or not emty instead of null ;) http://java.sun.com/j2se/1.5.0/docs/api/java/util/Collections.html#EMPTY_... 5 Collections#EMPTY_LIST View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165109#4165109 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165109

17 years, 8 months

1
0
0 / 0

[Design of Messaging on JBoss (Messaging/JBoss)] - Re: Please don't use IDE auto format on commit

by skajotde

"clebert.suconic(a)jboss.com" wrote : "skajotde" wrote : Yes, I know I don't work on project ;) | | Your mistake! By participating on this forum *you are* working on the project :-P Don't be shy and you're welcome to contribute more ;-) Ha ;) Personally I don't like when somebody stays on the sidelines and give advice on my project ;) Now I'm working on my masters and when finish I will think on some work on jboss and later maybe full time job ;) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165103#4165103 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165103

17 years, 8 months

1
0
0 / 0

[Design the new POJO MicroContainer] - BootstrapDeployersTest enhancements and AssembledDirectory

by adrian＠jboss.org

I've added some support to the BootstrapDeployersTest so you can use the vfs's AssembledDirectory. i.e. you can create deployments out of disparate parts of the testsuite structure For some simple examples see: http://viewvc.jboss.org/cgi-bin/viewvc.cgi/jbossas/projects/jboss-deploye... But this was a lot harder work than I expected. The AssembledDirectory has a number of problems (bugs?) and is not very easy to use. There's some FIXMEs here: http://viewvc.jboss.org/cgi-bin/viewvc.cgi/jbossas/projects/jboss-deploye... which show where the problems are, but I'll give some more explanation on this thread First the only factory for AssembledDirectory is the AssembledContext which is not a part of the public api. Also the api should return an AssembledDirectory so you don't have to cast from the super class. | import org.jboss.virtual.plugins.context.vfs.AssembledContext; | ... | | // FIXME Missing factor method in the public api | protected AssembledDirectory createAssembledDirectory(String name) throws Exception | { | AssembledContext ctx = new AssembledContext(name, ""); | return (AssembledDirectory) ctx.getRoot().getVirtualFile(); | } | There is no support for doing something as simple as adding a META-INF directory and all its contents to the assembled directory, so I wrote it myself, but this obviously should be in the AssembledDirectory class. | // FIXME why doesn't AssembledDirectory support this simple use case? | protected void addPath(final AssembledDirectory dir, String path, String name) throws Exception | { | URL url = getResource(path); | if (url == null) | fail(path + " not found"); | VirtualFile file = VFS.getVirtualFile(url, name); | | final VisitorAttributes va = new VisitorAttributes(); | va.setLeavesOnly(true); | SuffixesExcludeFilter noJars = new SuffixesExcludeFilter(JarUtils.getSuffixes()); | va.setRecurseFilter(noJars); | | VirtualFileVisitor visitor = new VirtualFileVisitor() | { | public VisitorAttributes getAttributes() | { | return va; | } | | public void visit(VirtualFile virtualFile) | { | dir.mkdirs(virtualFile.getPathName()).addChild(virtualFile); | } | }; | file.visit(visitor); | } | The urls for assembled directories are broken. e.g. if you have vfs://a/ and do mkdir("b") on it, you get back a url that says vfs://ab/ I think this is because it passes null to the "path" when it constructs the URI in AssembledContext, but I didn't confirm it. For now I've hacked the url handling in the test. | protected DeploymentUnit assertChild(DeploymentUnit parent, String name) | { | // FIXME AssembledContext URLs are broken | String parentName = parent.getName(); | if (parentName.endsWith("/")) | parentName = parentName.substring(0, parentName.length()-1); | name = name + "/"; | | name = parentName + name ; | List<DeploymentUnit> children = parent.getChildren(); | for (DeploymentUnit child : children) | { | if (name.equals(child.getName())) | return child; | } | throw new AssertionFailedError("Child " + name + " not found in " + children); | } | addResources() is a not very easy to use api and it is slow (takes about 1 second for deployers-vfs). The example here looks easy, but it actually scans the whole class and resource structure, even though I've already told it which package I want it to add. Having filters is nice, but there's no need to be dumb or slow about simple cases. | protected void addPackage(AssembledDirectory dir, Class<?> reference) throws Exception | { | String packagePath = ClassLoaderUtils.packageNameToPath(reference.getName()); | dir.addResources(reference, new String[] { packagePath + "/*.class" } , new String[0]); | } | Part of the reason it is slow is that it is trying to open jars (some of which in deployers-vfs are deliberately broken to test error handling). I hacked the test to reduce the noise from this, as the FIXME says, there should really be a method in AbstractTestCase to change the logging level, rather than having to assume log4j | protected void setUp() throws Exception | { | super.setUp(); | // This is a hack for a hack. ;-) | AbstractJDKChecker.getExcluded().add(BootstrapDeployersTest.class); | | // Reduce the noise from the VFS | // FIXME add method change logging levels to AbstractTestCase | Logger.getLogger(FileSystemContext.class).setLevel(Level.INFO); | } | View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165066#4165066 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165066

17 years, 8 months

1
0
0 / 0

[Design of POJO Server] - Re: MainDeployer MBean methods in JBoss 5

by Wolfgang Knauf

Filed a bug for the failing test (though I haven't seen it myself): https://jira.jboss.org/jira/browse/JBAS-5776 View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165061#4165061 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165061

17 years, 8 months

1
0
0 / 0

[Design of Messaging on JBoss (Messaging/JBoss)] - Re: Prototype for Management API for JBM 2

by jmesnil

forgot to mention: all the Server configuration is listed as read-only attributes in "Core > Server" MBean View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4165041#4165041 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4165041

17 years, 8 months

1
0
0 / 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

jboss-dev-forums July 2008