geronimo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From David Jencks <>
Subject Re: [jira] Updated: (TUSCANY-527) First cut of the work scheduler implementation
Date Mon, 10 Jul 2006 07:34:01 GMT

On Jul 9, 2006, at 8:57 PM, Kevan Miller wrote:

> On Jul 9, 2006, at 3:00 PM, Jeremy Boynes wrote:
>> On Jul 9, 2006, at 11:08 AM, Jeremy Boynes wrote:
>>> On Jul 9, 2006, at 9:16 AM, David Jencks wrote:
>>>> There's a geronimo version of the commonj spec at  
>>>> org.apache.geronimo.specs:geronimo-commonj_1.1_spec:1.0:jar  at  
>>>> ibiblio.  I'm not entirely sure what auspices this was published  
>>>> under since geronimo doesn't have a tck for commonj.  If there  
>>>> are any problems with this jar please speak up on the geronimo  
>>>> dev list.
> David, I'm not sure what you mean by "what auspices"...

I mean that this is the only spec we have released without having a  
tck to test against.  It's not entirely clear that the corba jar is a  
"spec" since IIUC it is entirely generated from IDL, but we didn't  
release it until the tck passed.  On the other hand IIUC we did later  
release a corba spec jar that doesn't work at all.

david jencks

>>> Thanks David, I'll use that for now.
>>> There is also a version in the Apache WebServices WSRF project  
>>> but they don't seem to have published it.
>> I found a couple of problems in Geronimo's version with the  
>> signatures (e.g. exceptions missing). I modified the Geronimo  
>> source to match the spec javadoc I downloaded and it seems to  
>> compile now. Attached is a patch for geronimo-spec (from https:// 
>> <geronimo-commonj.patch>
>> This bumps the commonj spec jar version to 1.0.1-SNAPSHOT - hope  
>> that's the right version.
>> Dave, if this works for you can you get a snapshot version  
>> uploaded so that Tuscany folk won't have to build from source.
> Jeremy,
> Thanks for pointing these problems out. I'll start working  on  
> getting a SNAPSHOT...
> A few comments:
> TimerManager.isSuspended(),  isSuspending(), waitForStop(), and  
> waitForSuspend() look correct to me. The javadoc for  
> TimerManager.suspend() gives mixed messages. I'm going to keep the  
> "suspend() throws IllegalStateException" for now... Let me know if  
> you have evidence as to what it should be.
> Your WorkCompletedException used a few Java 5 methods, I backed it  
> down so it will build under 1.4.
> --kevan

View raw message