<AWT Dev> [RFC] Tray icons for applications are not displayed in the GNOME notification bar.

Anthony Petrov anthony.petrov at oracle.com
Wed Oct 26 05:56:38 PDT 2011

Hi Danesh,

On 10/25/2011 10:00 PM, Danesh Dadachanji wrote:
> On 24/10/11 08:49 AM, Anthony Petrov wrote:
>> While you're at it, there's just been filed a bug report:
>> http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=7103610
> WM_CLASS is already set to something relative to what is being run. E.g. 
> when using xprop on a Frame
> $ xprop
> WM_CLASS(STRING) = "sun-awt-X11-XFramePeer", "test-TestFrame"
> Reading the specs of WM_CLASS[1], I think the above is incorrect. I 
> believe "test-TestFrame" makes more sense as the first st ring, the 
> instance name.
> As for the second one, from my understanding of the spec, you would want 
> it to be set to something more generic that. Do you think "Java" is a 
> better option?

I agree that the strings might be in the inverse order currently, and it 
would be nice to have them in the correct order. There's some sense in 
using the "sun-awt-X11-XFramePeer" however, since it may be used to 
distinguish between apps run using the XToolkit, and legacy MToolkit 
(removed in JDK7, but still available for apps run with JDK6 and 
earlier). If in the future we introduce some alternative AWT toolkit, 
then having the toolkit name in the string might be useful.

Having said that, I don't think we should change the value of the 
property. There's two reasons:

1. The WM_CLASS property is used to look up X resources for the app. 
Java apps (and AWT itself) have never used X resources, and as such it 
actually doesn't matter at all what is specified in these property.

2. Some developers may rely on the current order of the strings if they 
want to look up Java windows from their native apps, for example. If we 
change the order now, we may break these use cases. While the order has 
never been specified, I don't see a good reason for this change at this 

>> which suggests that the mentioned window properties need to be set on
>> all application's windows, not just the tray icon window. What do you
>> think about extending this logic to all windows? We could add a method
>> in the XWindowPeer class for this, and install the properties on all
>> windows (including the tray icon - since it's an embedded frame that
>> uses the XEmbeddedFramePeer as its peer).
> If all app windows need to be set, wouldn't this mean all X components 
> as well? In which case, how does adding the code to XBaseWidnow (looking 
> specifically at init) sound?

I don't think so. The window properties are used by the window manager 
(mostly). The window manager will never see our child windows, and as 
such setting the property for them seems useless.

>>> There's no any guarantees regarding the format of the string returned.
>>> Which means that if the XToolkit code is going to be run with a VM other
>>> than, say, Hotspot, the method may return a differently formatted
>>> string, and hence this code may fail.
>>> Is there a more robust way to obtain the PID and the client FQDN?
> For the PID, I think using JNI to access getpid() would be the next best 
> option. I'll look for another way to find the FQDN.

Yes, using getpid() via JNI seems reasonable. I guess gethostname() from 
unistd.h might be used to obtain the FQDN.

best regards,

More information about the awt-dev mailing list