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

[Design of JBoss Identity] - Bringing in thirdparty dependencies

by anil.saldhana＠jboss.com

Since JBoss Identity is going to be used by a number of projects, we need to ensure that each external dependency we bring in is notified to the stakeholders...... Historically during the JBAS development, we have had issues with dependencies - so nothing new here. :) There are some dependencies (JAXB, JPA API) etc that are needed for JDK5 which go away when JDK6 is used. I guess it is wise to just check if a particular dependency is going to be a problem. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226036#4226036 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226036

16 years, 11 months

1
0
0 / 0

[Design of POJO Server] - Re: VFS thread safety issues

by jason.greene＠jboss.com

"alesj" wrote : You're free to fix any thread or opt issues. ;-) | | There is no concrete logic behind it | apart from how classes are used. | e.g. single VFSContext per vfs root | | The idea was to solve all FS problems, | opt kicks in later. Sure, I just wanted to make sure I got the rules right before I started changing everything. Are all classes and all plugins supposed to support access from different threads? Is anything supposed to be immutable or tied to a single thread? View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226035#4226035 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226035

16 years, 11 months

1
0
0 / 0

[Design of POJO Server] - Re: VFS thread safety issues

by alesj

You're free to fix any thread or opt issues. ;-) There is no concrete logic behind it apart from how classes are used. e.g. single VFSContext per vfs root The idea was to solve all FS problems, opt kicks in later. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226034#4226034 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226034

16 years, 11 months

1
0
0 / 0

[Design of JBoss jBPM] - bad news for branchers

by tom.baeyens＠jboss.com

i'm gonna go ahead and commit the api package renaming we need to do. https://jira.jboss.org/jira/browse/JBPM-2173 andries, alejandro, sorry to give you a hard time. i hope that when you merge this is doable with a couple of global change/replaces and some quick-fix organise imports. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226029#4226029 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226029

16 years, 11 months

1
0
0 / 0

[Design of POJO Server] - VFS thread safety issues

by jason.greene＠jboss.com

It is not very clear what the thread access semantics are supposed to be for the VFS. Can someone explain what the intended thread access patterns are supposed to be (i.e. which classes are supposed to be multi-thread accessable, and which ones are supposed to be allocated and used by a single thread)? Assuming the cases where there is some form of locking are correct, it is not consistently applied, or there are other races. As an example AbstractVFSContext uses a synchronizedMap (previously a CHM) to store temporary information. However, all other fields on this class are mutable, yet not guarded or volatile (vfs, rootPeer, and options). Also, now that this class is using a synchronized map, the iterator usage by getTempInfos and cleanupTemp is not safe, since synchronized maps do not protect iterators. It looks like switching to a ConcurrentSkipListMap would solve this particular safety issue, and provide faster search performance (it's an O(log n) sorted map). It is JDK 6 but we can pull in the source to common-core (public domain). However CHM could be restored if putIfAbsent() was used. A number of the classes which extend AbstractVFSContext have guard issues on their fields as well (FileSystemContext, TempContext, ZipEntryContext, AbstractVirtualFileHandler (and its subclases), etc). There are also cases of non-volatile double checked-locking (VirtualFile, DefaultOptions, VFSCacheFactory) Somewhat related but not a thread safety issue (yet), is the odd interaction between the VFS and the cache policy code in common-core. The default thread safety flag for IterableTimed is true, which is logged, however the policy is constructed with it being false. It turns out that the policy does synchronization on some methods regardless of the setting, and it looks like the VFSCache stuff uses locking in the calling methods, so it seems to work out ok. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226025#4226025 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226025

16 years, 11 months

1
0
0 / 0

[Design of JBoss jBPM] - Re: GWT Console login problem

by heiko.braun＠jboss.com

Yes, the current error handling is bad. It sometimes leaves the console in a weird state. More of problem is that the error messages are not very self explaining. But from what you describe, you seem to be running into a timeout in situation b) View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226018#4226018 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226018

16 years, 11 months

1
0
0 / 0

[Design of JBoss jBPM] - Re: GWT Console login problem

by kukeltje

Oh, and after this, I do see the console, but nothing realy responds if requests have to be made to the server. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226017#4226017 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226017

16 years, 11 months

1
0
0 / 0

[Design of JBoss jBPM] - Re: GWT Console login problem

by kukeltje

I was just going to provide some additional info, mainly regarding b), but to answer in order: a) FF 3.0.8 with Firebug 1.3.3 (will try without firebug tonight) b) Initially none, but patient as I am, I left it open and after 5 minutes of 100% cpu (one cpu lucky I have two :-) so I did not notice it) this error occured: anonymous wrote : | [ERROR] [ERROR] Failed to load remote data | com.gwtext.client.data.StoreLoadException: [object Object] | at com.gwtext.client.data.Store.createStoreLoadException(Store.java:907) | at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) | at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57) | at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) | at java.lang.reflect.Method.invoke(Method.java:616) | at com.google.gwt.dev.shell.MethodAdaptor.invoke(MethodAdaptor.java:103) | at com.google.gwt.dev.shell.moz.MethodDispatch.invoke(MethodDispatch.java:80) | at org.eclipse.swt.internal.gtk.OS._g_main_context_iteration(Native Method) | at org.eclipse.swt.internal.gtk.OS.g_main_context_iteration(OS.java:1428) | at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:2840) | at com.google.gwt.dev.GWTShell.pumpEventLoop(GWTShell.java:720) | at com.google.gwt.dev.GWTShell.run(GWTShell.java:593) | at com.google.gwt.dev.GWTShell.main(GWTShell.java:357) View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226016#4226016 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226016

16 years, 11 months

1
0
0 / 0

[Design of JBoss jBPM] - Re: schema for mail template and activity

by bradsdavis

Many JBoss projects use Apache Commons projects; all commons projects are just simplified interfaces or utility classes. Some of those include: Commons-collections Commons-logging Commons-lang Commons-codec Commons-io etc. And these were pulled from successful projects: Seam, Messaging, Cache. Users are going to use our mail stuff; they are going to extend on it. Might as well make it easier on them rather than having to use some module we wrote from scratch that provides the same functionality as Commons. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226015#4226015 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226015

16 years, 11 months

1
0
0 / 0

[Design of JBoss jBPM] - Re: schema for mail template and activity

by tom.baeyens＠jboss.com

i've looked at the branch and it looks like i thought. the library dependency is a showstopper for me. we need to depend only on JavaMail the other two points (2 and 3) are basically taste and hence less important to me. the configuration of the templates and the jpdl language are important. that needs to be discussed as we build out the examples. in the test with the mail template configurations, all looks good except for the actors attribute. that should now be changed into attributes users or groups specifically. i'm surprised to see so much code without a single end-to-end example that we can discuss in detail. we should go in smaller iterations and discuss sooner. otherwise the risk throwing away a lot of work in case the discussions lead to different outcome View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4226014#4226014 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4226014

16 years, 11 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 April 2009