January 2009 - jboss-dev-forums - Jboss List Archives

[Design of POJO Server] - Re: JBAS-6061 or weird nested jar

by jaikiran

"alesj" wrote : | Perhaps it's not getting passed from root to more exact handler ... That appears to be the case. Do you want me to open a new JIRA in JBVFS project to track this? I did not reopen the existing one since its already marked as resolved for a released version. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203233#4203233 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203233

17 years, 2 months

1
0
0 / 0

[Design of JBoss Portal] - Re: Design of Portal Deployer

by alesj

"mwringe" wrote : "mwringe" wrote : | | Things work properly is deployed separetely, it just means there will be more things to deploy on the server than just a single sar. | | | Ok, this is rather easy to get around, I can just deploy a directory that contains the sar and wars and with the proper dependency setup it will work fine and it will only be a single artifact to deploy. Exactly. ;-) Just make sure you don't have any '.' in the directory name. btw: any more info on tld? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203224#4203224 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203224

17 years, 2 months

1
0
0 / 0

[Design of OSGi Integration] - Current roadmap

by alesj

With JBoss5 out, it's time to get back to OSGi impl work. ;-) Taking it step-by-step, let's first check what we have and how to move fwd from this. OSGi = classloading + core frmwrk + services 1) Classloading Adrian already did most of the work, it's just a few details left to polish. Since new CL is metadata driven, all we need to do is provide a proper: * OSGi manifest.mf parsing * OSGi manifest metadata --> JBoss CL metadata transformation 1a) OSGi manifest parsing Currently it's based on JavaCC and I'm not very happy with it. :-( * Do we proceed with fixing/enhancing current impl? * Or do we check some existing - Felix, Equinox, ... (what license do we need to re-use it)? 1b) OSGi manifest metadata --> JBoss CL metadata transformation This step mostly needs thorough testing, OSGi's feature by feature --> mapping it to JBoss CL metadata. This way we'll also see if we cover all features in our CL. 2) Core OSGi framework as a facade First step: enable OSGi TCK env Next steps: a) BundleContext and services This needs proper impl on top of DeploymentUnit/Context. Services need proper registry - simple/smart api behind maps. b) BundleActivationPolicy This is a new feature - from 4.2 spec. I guess we'll need some sort of a hook into Classloaders? e.g. the spec says to only enable BundleActivator if some class from the bundle is accessed c) listeners and events I did some work, but I think it needs more wiring inside the MC Kernel itself. OSGi eventing should drive the listener/event introduction in the MC Kernel. d) BundleActivator This is a simple deployer - which already exists. It's the previous tasks that need to be done before to make this run. e) DeclarativeServices Perhaps we can use similar concept to what we've done with MC-Spring-int? Mapping Spring's schema to MC beans metadata, where here we would map DS schema to MC beans schema. In any case, this is optional. We should eventually support the new OSGi RFC 124. 3) Services No intention of impl it. Dunno how much the OSGi TCK tests this things. I think we should at least test the main services - logging, http, ... - how well they run in our frmwrk impl. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203198#4203198 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203198

17 years, 2 months

1
0
0 / 0

[Design the new POJO MicroContainer] - VFS bug testing

by rhusar

Hello Ales and everyone, lets establish together what we need for the testing of the VFS bug. I am running long tests for the clustering test, maybe we can add it to the test scenario, but i guess it will be easier to create one from the scrach. But depends what we need indeed. Ales, maybe you can send your "some quick/crappy test on my 2nd machine" and I can have a look at that. BTW, im on GMT+1, you can reach me on IRC. The latest snapshosts: - VFS 2.2.0-SNAPSHOT - Deployers 2.2.0-SNAPSHOT - the vfs.xml from the JBoss5 forum (with combined cache) Cheers, Rado View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203197#4203197 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203197

17 years, 2 months

1
0
0 / 0

[Design of Messaging on JBoss (Messaging/JBoss)] - Re: Failures on testPagedMessageDeliveredMultipleConsumersAf

by timfox

My mistake. I just need to unset the header before routing to the next one. I'll fix it. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203185#4203185 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203185

17 years, 2 months

1
0
0 / 0

[Design of Messaging on JBoss (Messaging/JBoss)] - Re: Failures on testPagedMessageDeliveredMultipleConsumersAf

by ataylor

Ok, so this is failing because of changes in the way we now route. If 2 messages are routed to 2 queues 'A' and 'B' and then both subsequently canceled we check the redelivery delay and if its not 0 we reschedule the message with that delay. If queue 'A' has no delivery delay but queue 'B' has a 5 second delay, we update the journal to say that the reference for queue 'B' has a scheduled delivery time. Now If the server is restarted we check the record for a scheduled delivery time and if it is not 0 we set it as a property on the message before rerouting via the queue. The problem here is that once it is set then every time it is rerouted to any other queue it uses the same redelivery value. Previously we didn't set the property but after routing it would be set on the reference before adding to the queue. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203184#4203184 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203184

17 years, 2 months

1
0
0 / 0

[Design of POJO Server] - Re: A taster - jboss reflect docs

by richard.opalka＠jboss.com

"alesj" wrote : I think we can just remove this? | As I see CL project uses Translator from Commons. If you're 100% sure it's not referenced from else where remove it ASAP to don't confuse other users of the API ;) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203173#4203173 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203173

17 years, 2 months

1
0
0 / 0

[Design of POJO Server] - Re: A taster - jboss reflect docs

by richard.opalka＠jboss.com

"alesj" wrote : Some better (more than just saying 'Nice') feedback: | * Initial diagram should show TypeInfo/ClassInfo as well | * TypeInfo's / ClassInfo's weak cache could be mentioned (TypeInfoFactory as a singleton) | * Joinpoint spi/api is deprecated? Adrian? | * Is Classadapter still good usage? It should be either enhanced or removed? | * Explain BeanAccessMode OK, I'll do it soon (in my next spare time). View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203172#4203172 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203172

17 years, 2 months

1
0
0 / 0

[Design of POJO Server] - Re: A taster - jboss reflect docs

by alesj

"richard.opalka(a)jboss.com" wrote : "alesj" wrote : What's "ClassLoading SPI" doing there at the end? | It's in the jboss-reflect/trunk SVN repo. The TODO at the end of document needs to be evaluated. I think we can just remove this? As I see CL project uses Translator from Commons. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203168#4203168 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203168

17 years, 2 months

1
0
0 / 0

[Design of Messaging on JBoss (Messaging/JBoss)] - Re: XML configuration

by jmesnil

i've committed XML schema for all our configuration files. Validation occurs in XmlDeployer subclasses. Some tidbits about the XML schema: == order matters == I've not found how to provide a "bag" of elements with a XML schema. Elements must now be ordered for the configuration to be valid == ID/IDREF == I took advantage of ID & IDREF in the schema to validate our conf. E.g. if the backup-connector-ref element references a "blah" connector, there MUST be a connector element with the "blah" attribute. I did not change the code but we can now remove some validation done in Java and now handled by XSD. However this does not apply to cross-configuration files (e.g. jbm-jms.xml connection-factory can not IDREF an element from jbm-configuration.xml). This still must be done in Java == Minimal configuration == The minimal valid configuration file is | <deployment> | <configuration> | <acceptor>...</acceptor> | </configuration> | </deployment> | there MUST be at least 1 declared acceptor. All other elements are optional. == Default values == The XML schemas DOES NOT define any default values. The defaults are taken from Configuration interface if the corresponding XML element is not present in the configuration View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203165#4203165 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4203165

17 years, 2 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 January 2009