dev@grizzly.java.net

Re: [HEADS-UP] OSGi problem with LTQ

From: Hubert Iwaniuk <neotyk_at_kungfoo.pl>
Date: Wed, 21 Jan 2009 17:03:21 +0100

Did so, and guess what :) It did work :)
-> ps
START LEVEL 6
   ID State Level Name
[ 0] [Active ] [ 0] System Bundle (1.4.1)
[ 1] [Active ] [ 5] Grizzly OSGi HttpService (1.9.5.SNAPSHOT)
[ 2] [Active ] [ 5] osgi.core (4.1.0.build-200702212030)
[ 3] [Active ] [ 1] osgi.compendium (4.1.0.build-200702212030)
[ 4] [Active ] [ 5] grizzly-http-servlet (1.9.5.SNAPSHOT)
[ 5] [Active ] [ 5] grizzly-utils (1.9.5.SNAPSHOT)
[ 6] [Active ] [ 5] grizzly-http (1.9.5.SNAPSHOT)
[ 7] [Active ] [ 5] grizzly-framework (1.9.5.SNAPSHOT)
[ 8] [Active ] [ 5] grizzly-rcm (1.9.5.SNAPSHOT)
[ 9] [Active ] [ 5] grizzly-portunif (1.9.5.SNAPSHOT)
[ 10] [Active ] [ 1] OPS4J Pax Logging - API (1.3.0)
[ 11] [Active ] [ 1] OPS4J Pax Logging - Service (1.3.0)
[ 12] [Active ] [ 1] Apache Felix Shell Service (1.0.2)
[ 13] [Active ] [ 1] Apache Felix Shell TUI (1.0.2)
-> services

System Bundle (0) provides:
---------------------------
org.osgi.service.startlevel.StartLevel
org.osgi.service.packageadmin.PackageAdmin

Grizzly OSGi HttpService (1) provides:
--------------------------------------
org.osgi.service.http.HttpService

OPS4J Pax Logging - Service (11) provides:
------------------------------------------
org.osgi.service.log.LogReaderService
org.osgi.service.log.LogService, org.knopflerfish.service.log.LogService,
org.ops4j.pax.logging.PaxLoggingService, org.osgi.service.cm.ManagedService

Apache Felix Shell Service (12) provides:
-----------------------------------------
org.apache.felix.shell.ShellService,
org.ungoverned.osgi.service.shell.ShellService

Good work :)
Hubert.

On Wed, Jan 21, 2009 at 4:58 PM, gustav trede <gustav.trede_at_gmail.com>wrote:

>
> i fixed that in trunk 30 sec later.
> check out the code again
>
>
> 2009/1/21 Hubert Iwaniuk <neotyk_at_kungfoo.pl>
>
>> Bad news {write,read}Object methods of LTQ both throw
>> NotImplementedException from sun.reflect.generics.reflectiveObjects.But
>> as we don't use them it should be easy to remove them, and have it fixed.
>>
>> On Wed, Jan 21, 2009 at 4:44 PM, gustav trede <gustav.trede_at_gmail.com>wrote:
>>
>>>
>>> I commited a fixed LTQ now. sacrificing serialization does not affect
>>> us.
>>>
>>> 2009/1/21 Hubert Iwaniuk <neotyk_at_kungfoo.pl>
>>>
>>> Gustav in order to test it go to contrib/grizzly-httpservice and execute
>>>> mvn clean install pax:provision.You shouldn't see any exceptions like:
>>>> org.osgi.framework.BundleException: Unresolved constraint in bundle 5:
>>>> package; (package=sun.misc)
>>>> at
>>>> org.apache.felix.framework.Felix._resolveBundle(Felix.java:1792)
>>>> at
>>>> org.apache.felix.framework.Felix._startBundle(Felix.java:1652)
>>>> at org.apache.felix.framework.Felix.startBundle(Felix.java:1588)
>>>> at
>>>> org.apache.felix.framework.Felix.setFrameworkStartLevel(Felix.java:1180)
>>>> at
>>>> org.apache.felix.framework.StartLevelImpl.run(StartLevelImpl.java:265)
>>>> at java.lang.Thread.run(Thread.java:595)
>>>>
>>>> HTH,
>>>> Hubert.
>>>>
>>>> On Wed, Jan 21, 2009 at 3:57 PM, gustav trede <gustav.trede_at_gmail.com>wrote:
>>>>
>>>>>
>>>>> ok i get onto it right now.
>>>>>
>>>>>
>>>>> 2009/1/21 Oleksiy Stashok <Oleksiy.Stashok_at_sun.com>
>>>>>
>>>>>> Hi Gustav,
>>>>>>
>>>>>>
>>>>>> how urgent is this , is tomorrow ok ?.
>>>>>>
>>>>>> Well, if you don't have time today - it's fine. But, IMHO, it's
>>>>>> priority 1 issue, as we need to have that fixed before 1.9.5 release.
>>>>>>
>>>>>> WBR,
>>>>>> Alexey.
>>>>>>
>>>>>>
>>>>>>
>>>>>> 2009/1/21 gustav trede <gustav.trede_at_gmail.com>
>>>>>>
>>>>>>>
>>>>>>> I will see what other projects do about this situation.
>>>>>>>
>>>>>>> 2009/1/21 Oleksiy Stashok <Oleksiy.Stashok_at_sun.com>
>>>>>>>
>>>>>>> Think it's not OSGi problem, but we can not use sun internal packages
>>>>>>>> at all.
>>>>>>>> Gustav, can you pls. take a look at that?
>>>>>>>> I'm not sure what Unsafe is used for, looks like it is used to
>>>>>>>> change "final" Class members. We can try to remove the dependancy by
>>>>>>>> removing "final" declaration and the check if it still performs good with
>>>>>>>> our testcases.
>>>>>>>>
>>>>>>>> Thanks.
>>>>>>>>
>>>>>>>> WBR,
>>>>>>>> Alexey.
>>>>>>>>
>>>>>>>>
>>>>>>>> On Jan 21, 2009, at 10:24 , Hubert Iwaniuk wrote:
>>>>>>>>
>>>>>>>> Hi,
>>>>>>>>>
>>>>>>>>> Since introduction of LinkedTransferQueue we have problems with
>>>>>>>>> OSGi packaging.
>>>>>>>>> LTQ depends on sun.misc.Unsafe and sun.misc is not provided by any
>>>>>>>>> bundle, so trying to provision OSGi bundle will fail.
>>>>>>>>> Trying to figure out how to get this package visible in OSGi
>>>>>>>>> environment.
>>>>>>>>>
>>>>>>>>> Any help is appreciated,
>>>>>>>>> Hubert.
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> ---------------------------------------------------------------------
>>>>>>>> To unsubscribe, e-mail: dev-unsubscribe_at_grizzly.dev.java.net
>>>>>>>> For additional commands, e-mail: dev-help_at_grizzly.dev.java.net
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> your servant
>>>>>>> gustav trede
>>>>>>>
>>>>>>> coding is art - not only something that bring food on the table,
>>>>>>> everybody should be able to feel proud about their code,
>>>>>>> that they have performed their best considering the given conditions.
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> your servant
>>>>>> gustav trede
>>>>>>
>>>>>> coding is art - not only something that bring food on the table,
>>>>>> everybody should be able to feel proud about their code,
>>>>>> that they have performed their best considering the given conditions.
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> your servant
>>>>> gustav trede
>>>>>
>>>>> coding is art - not only something that bring food on the table,
>>>>> everybody should be able to feel proud about their code,
>>>>> that they have performed their best considering the given conditions.
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> Hubert Iwaniuk
>>>>
>>>
>>>
>>>
>>> --
>>> your servant
>>> gustav trede
>>>
>>> coding is art - not only something that bring food on the table,
>>> everybody should be able to feel proud about their code,
>>> that they have performed their best considering the given conditions.
>>>
>>>
>>
>>
>> --
>> Hubert Iwaniuk
>>
>
>
>
> --
> your servant
> gustav trede
>
> coding is art - not only something that bring food on the table,
> everybody should be able to feel proud about their code,
> that they have performed their best considering the given conditions.
>
>


-- 
Hubert Iwaniuk