news.eclipse.technology.equinox

[news.eclipse.technology.equinox] Starting a Java application from a bundle

maurizio.lilli@xxxxxxx (Maurizio ) — Sat, 04 Jul 2009 20:08:34 GMT

Dear Sir, Which are the steps to follow for activating a plain Java application from within a bundle?

Thanks in advance.
Kind regards.
Maurizio

[news.eclipse.technology.equinox] Re: p2 and Composite Features

apicard@xxxxxxx (Alain Picard) — Sat, 04 Jul 2009 15:46:51 GMT

Well I've finally figured out that it was all caused by the fact the a separate JVM was selected in the configuration.

Not sure then what it would imply to run this outside of the Workspace JVM, but I guess that is for another day.

Alain

[news.eclipse.technology.equinox] How to get resource using ServletContext ?

jsetya@xxxxxxx (Setya) — Sat, 04 Jul 2009 13:00:52 GMT

Hi all,

I've registered a servlet as follows:

In the bundle root I have folder 'config' which contains foo.xml.

Now the question is how can I access the foo.xml file from within servlet MyServlet ? I've tried ServletContext#getResource("/config/foo.xml") or ServletContext#getResourceAsStream("/config/foo.xml") but they always return null.

Any help would be greatly appreciated.

Regards,

Setya

[news.eclipse.technology.equinox] Re: P2 Installer 3.5 final fails with "An error occurred while collecting items to be installed"

c.vidal@xxxxxxx (Cédric Vidal) — Sat, 04 Jul 2009 12:55:37 GMT

I posted a bug here:
https://bugs.eclipse.org/bugs/show_bug.cgi?id=282427

Cédric Vidal a écrit :

This fails only with P2 Installer. The exact same scenario works with P2 Agent 3.5. It might be because my bundles are locally cached though.
What's more P2 Installer is freshly unzipped. First run. No History.
Regards,
Cédric
------------------------------------------------------------------------

[news.eclipse.technology.equinox] Re: P2 "Headless Updating on Startup" not working

bvitale3002@xxxxxxx (Ben Vitale) — Sat, 04 Jul 2009 01:29:41 GMT

It looks like someone reported the same issue.

https://bugs.eclipse.org/bugs/show_bug.cgi?id=282333

You may want to CC yourself on the bug.

Cheers
Ben

scott brown wrote:

Is it possible that some things changed for the final 3.5 version that broke the "Headless Updating on Startup" example? scott brown wrote:

bringing back an old topic

I've downloaded the 3.5 release and tried my best to find all the p2 documentation I could. I tried Susan's example for "Headless Updating on Startup," including her p2Util class with my application, but i'm running into an issue: I get an "Invalid Thread Access"

I'll include the stack trace from the workspace .log and from our application log below. the line from P2Util that seems to cause the error is:

manager.loadRepository(reposToSearch[i], loadMonitor.newChild(100 / reposToSearch.length));

Can anyone else reproduce this?

workspace .log stack trace:

!ENTRY org.eclipse.ecf.provider.filetransfer 4 1001 2009-06-26 14:45:19.815 !MESSAGE Transfer Exception !STACK 0 org.eclipse.swt.SWTException: Invalid thread access at org.eclipse.swt.SWT.error(SWT.java:3884) at org.eclipse.swt.SWT.error(SWT.java:3799) at org.eclipse.swt.SWT.error(SWT.java:3770) at org.eclipse.swt.widgets.Widget.error(Widget.java:463) at org.eclipse.swt.widgets.Widget.checkWidget(Widget.java:355) at org.eclipse.swt.widgets.ProgressBar.getSelection(ProgressBar.java:191) at
org.eclipse.jface.dialogs.ProgressIndicator.worked(ProgressIndicator.java:139)

at
org.eclipse.jface.dialogs.ProgressMonitorDialog$ProgressMonitor.internalWorked(ProgressMonitorDialog.java:246)

at
org.eclipse.jface.dialogs.ProgressMonitorDialog$ProgressMonitor.worked(ProgressMonitorDialog.java:241)

at org.eclipse.core.runtime.SubMonitor$RootInfo.worked(SubMonitor.java:284) at org.eclipse.core.runtime.SubMonitor.internalWorked(SubMonitor.java:570) at org.eclipse.core.runtime.SubMonitor.worked(SubMonitor.java:585) at
org.eclipse.equinox.internal.p2.repository.FileReader.handleTransferEvent(FileReader.java:161)

at
org.eclipse.ecf.provider.filetransfer.retrieve.AbstractRetrieveFileTransfer.fireTransferReceiveDataEvent(AbstractRetrieveFileTransfer.java:318)

at
org.eclipse.ecf.provider.filetransfer.httpclient.HttpClientRetrieveFileTransfer.fireTransferReceiveDataEvent(HttpClientRetrieveFileTransfer.java:1013)

at
org.eclipse.ecf.provider.filetransfer.retrieve.AbstractRetrieveFileTransfer.handleReceivedData(AbstractRetrieveFileTransfer.java:213)

at
org.eclipse.ecf.provider.filetransfer.retrieve.AbstractRetrieveFileTransfer$1.performFileTransfer(AbstractRetrieveFileTransfer.java:141)

at org.eclipse.ecf.filetransfer.FileTransferJob.run(FileTransferJob.java:73) at org.eclipse.core.internal.jobs.Worker.run(Worker.java:55)

[news.eclipse.technology.equinox] Re: Update with p2 fails in RCP app

fakemail@xxxxxxx (Christian) — Fri, 03 Jul 2009 23:44:29 GMT

ian mayo schrieb:

Christian,
did you overcome this problem?
I've been battling with it all week - but you're one of the only other people who appear to have encountered it.

As you experienced, the Eclipse updater will not install an updated version of a feature because it says the product depends on an earlier version.
I'm stumped - hopefully you had better luck.
cheers,
Ian


My current way to go is to create two update sites.

One with the repository from the export of the product. And a second one that contains all features/plugins that are not part of the product but are selectable by the user as enhancement of the program...

Though haven't had much time working on this lately.. there still seem to be problems... like stuff is shown when searching for extensions that I don't want to be shown there (some mysterious launcher that if selected creates an eclipse.exe in the directory of the program)... but I am going to solve that soon hopefully...

Christian

[news.eclipse.technology.equinox] Running OSGi-DS as an "Eclipse Application"

owong@xxxxxxx (Oliver Wong) — Fri, 03 Jul 2009 21:58:39 GMT

I'm working on a small declarative services OSGi example (for what it's worth, it's the "org.equinoxosgi.toast" example). When I try to run the example using an "OSGi framework" launch configuration, the example runs fine (it prints the intended output to System.out). But when I try to run the same code using an "Eclipse Application" launch configuration, the example doesn't print anything out at all.

After some debugging, we realized one of the big difference between an OSGi launch config and an Eclipse launch config is that OSGi has "Default autostart" set to true, and Eclipse has it set to false. When I change my plugins so that their autostart is "true" instead of default, the example finally works in the Eclipse launch config.

Now my question is: Does this imply that the "immediate='true'" attribute of the "component" element in my OSGI-INF/component.xml file doesn't actually do anything in an Eclipse launch? I'd like to build a system involving multiple components that will need to start up "immediately". Shouldn't I be able to specify this behaviour in my component.xml file, rather than in the launch configurations?

[news.eclipse.technology.equinox] Re: RCP using declarative services (occasionally) hangs for 30 secs on start-up.

zx@xxxxxxx (Chris Aniszczyk) — Fri, 03 Jul 2009 17:42:26 GMT

Mike Evans wrote:

Hi All, I am experimenting with using OSGi declarative services in a conventional RCP application. However when I launch the app within the IDE, it normally hangs for about 30 seconds, waiting for OSGi to start up. More specifically the while() loop in EclipseStarter.updateSplash() takes 30 seconds or so to acquire the semaphore. If I remove my declared service by removing the reference to the service definition xml file in manifest.mf the app starts in about 3 secs ï naturally without the service. Now hereïs the odd bit. If I restore the manifest.mf and re-run, everything runs beautifully for a few launches, opening service and all in 3 secs or so. However eventually the app will suddenly again start taking 30 seconds. Any ideas what might cause this? Any pointers much appreciated.

Service definition: http://www.osgi.org/xmlns/scr/v1.1.0"; enabled="true" immediate="true" name="Standalone Authentication">
Host bundle definition:
Manifest-Version: 1.0
Bundle-ManifestVersion: 2
Bundle-Name: EOMA framework : standalone services
Bundle-SymbolicName: uk.org.eoma.ui.application.standalone;singleton:=true
Bundle-Version: 0.1.0.qualifier
Bundle-Activator: uk.org.eoma.ui.application.standalone.internal.Activator
Require-Bundle: org.eclipse.core.runtime,
uk.org.eoma.ui.application;bundle-version="0.1.0",
org.eclipse.equinox.security;bundle-version="1.0.100"
Bundle-ActivationPolicy: lazy
Service-Component: OSGI-INF/service.xml
Bundle-RequiredExecutionEnvironment: JavaSE-1.6
Bundle-Vendor: Incremental Ltd.
Thanks for your time,
Mike E.
PS : the app is currently a proof-of-concept so if anyone wants the code (all 3 plug-inïs) you are welcome to it.


Can you file a bug against RT->Equinox with the sample projects?

DS has a timeout value if it can't get your service.

Can you post the bug here too!

Cheers,

Chris Aniszczyk | EclipseSource Austin | +1 860 839 2465
http://twitter.com/eclipsesource | http://twitter.com/caniszczyk

[news.eclipse.technology.equinox] RCP using declarative services (occasionally) hangs for 30 secs on start-up.

mike.evans@xxxxxxx (Mike Evans) — Fri, 03 Jul 2009 17:17:16 GMT

Hi All, I am experimenting with using OSGi declarative services in a conventional RCP application. However when I launch the app within the IDE, it normally hangs for about 30 seconds, waiting for OSGi to start up. More specifically the while() loop in EclipseStarter.updateSplash() takes 30 seconds or so to acquire the semaphore. If I remove my declared service by removing the reference to the service definition xml file in manifest.mf the app starts in about 3 secs ? naturally without the service. Now here?s the odd bit. If I restore the manifest.mf and re-run, everything runs beautifully for a few launches, opening service and all in 3 secs or so. However eventually the app will suddenly again start taking 30 seconds. Any ideas what might cause this? Any pointers much appreciated.

Service definition: http://www.osgi.org/xmlns/scr/v1.1.0"; enabled="true" immediate="true" name="Standalone Authentication">

Host bundle definition:
Manifest-Version: 1.0
Bundle-ManifestVersion: 2
Bundle-Name: EOMA framework : standalone services
Bundle-SymbolicName: uk.org.eoma.ui.application.standalone;singleton:=true
Bundle-Version: 0.1.0.qualifier
Bundle-Activator: uk.org.eoma.ui.application.standalone.internal.Activator
Require-Bundle: org.eclipse.core.runtime,
uk.org.eoma.ui.application;bundle-version="0.1.0",
org.eclipse.equinox.security;bundle-version="1.0.100"
Bundle-ActivationPolicy: lazy
Service-Component: OSGI-INF/service.xml
Bundle-RequiredExecutionEnvironment: JavaSE-1.6
Bundle-Vendor: Incremental Ltd.

Thanks for your time,

Mike E.

PS : the app is currently a proof-of-concept so if anyone wants the code (all 3 plug-in?s) you are welcome to it.

[news.eclipse.technology.equinox] Re: Update with p2 fails in RCP app

IanMayo@xxxxxxx (ian mayo) — Fri, 03 Jul 2009 15:36:59 GMT

Christian,
did you overcome this problem?

I've been battling with it all week - but you're one of the only other people who appear to have encountered it.

As you experienced, the Eclipse updater will not install an updated version of a feature because it says the product depends on an earlier version.

I'm stumped - hopefully you had better luck.

cheers,
Ian