Re: [ptp-dev] Problems after update to latest head

[Clement Kam Man Chu <clement.chu@xxxxxxxxxxxxxxxxxxxxxx>]

I got exactly problems as you got.  For remote service id, I just hard 
code for fixing it temporary on my linux machine.  Regarding to 
connection problem, the problem I think orte_proxy cannot connect to 
eclipse probably by some strange arguments.  For fixing this problem, go 
to source org.eclipse.ptp.remote.AbstractRemoteProxyRuntimeClient method 
startupProxyServer() and comments two lines of codes like as belows.  I 
think both problems will be fixed soon.

                   if (portForwarding) {
                       //args = "--host=localhost " + args;
                   } else {
                       //args = "--host=" + getSessionHost() + " " + args;
                   }


Clement

Ricardo Marin Matinata wrote:
> On Mon, 2007-07-30 at 10:08 -0400, Dave Wootton wrote:
>   
> > I'm not sure if this is the right way to fix the problem with the null 
> > pointer exception in handleNewRemoteConnectionSelected, but if I check 
> > remoteServices for null before trying to use it (lines 285 & 286), then 
> > when I click New, the remote services combo box gets updated with 'Local', 
> > and I complete the wizard. Then I can start my proxy and run my PE 
> > application.
> > Dave
> >
> >
> >     
> Although i could successfully make everything work on Mac OS, I'm
> experiencing the exact same problem in Linux (tested both x86 and
> x86_64), with today's CVS HEAD. The way i've tried to work around the
> Wizard problem was to *force* an updateConnectionPulldown() somewhere in
> the createContents() (see the pacth below). It seems that in Linux, we
> are not getting the initial *Modify* event on the remoteCombo right
> after it is filled.
>
> With that, i was able to proceed with starting the orte_proxy, and
> actually observe another problem in Linux: the proxy client (in Eclipse)
> never gets the server connect, even though there is a server process
> (orte_proxy) running. If I manually try to kill the running server, and
> start another with the exact port parameter, everything proceeds
> normally.
> Is it possible that the server is coming up faster than the client is up
> to listening (even makes sense?) ?
>
> Index:
> src/org/eclipse/ptp/remote/ui/wizards/AbstractRemoteResourceManagerConfigurationWizardPage.java
> ===================================================================
> RCS
> file: /cvsroot/tools/org.eclipse.ptp/core/org.eclipse.ptp.remote/src/org/eclipse/ptp/remote/ui/wizards/AbstractRemoteResourceManagerConfigurationWizardPage.java,v
> retrieving revision 1.5
> diff -u -r1.5 AbstractRemoteResourceManagerConfigurationWizardPage.java
> ---
> src/org/eclipse/ptp/remote/ui/wizards/AbstractRemoteResourceManagerConfigurationWizardPage.java	13 Aug 2007 19:34:07 -0000	1.5
> +++
> src/org/eclipse/ptp/remote/ui/wizards/AbstractRemoteResourceManagerConfigurationWizardPage.java	14 Aug 2007 18:48:06 -0000
> @@ -314,6 +314,7 @@
>  		});	
>  
>  		updateRemotePulldown();
> +		updateConnectionPulldown();
>  		
>  		Composite proxyComp = new Composite(parent, SWT.NONE);
>  		GridLayout proxyLayout = new GridLayout();
> @@ -433,6 +434,7 @@
>  		IRemoteServices remoteServices =
> PTPRemotePlugin.getDefault().getRemoteServices(remoteServicesId);
>  		IRemoteConnectionManager connMgr =
> remoteServices.getConnectionManager();
>  		connMgr.newConnection(getControl().getShell());
> +		
>  	}
>  
>  	/**
>
>
>   


--
Clement Kam Man Chu
Research Assistant
Faculty of Information Technology
Monash University, Caulfield Campus
Ph: 61 3 9903 2355