JDK-6173972 : backspace, arrow keys, and return key not working in java applications
  • Type: Bug
  • Component: client-libs
  • Sub-Component: java.awt
  • Affected Version:
    other,solaris,solaris_10,solaris_10u1,solaris_11,1.0,2.0,2.10.0,2.4.0,5.0,5.0u1,5.0u2,5.0u7 other,solaris,solaris_10,solaris_10u1,solaris_11,1.0,2.0,2.10.0,2.4.0,5.0,5.0u1,5.0u2,5.0u7
  • Priority: P2
  • Status: Resolved
  • Resolution: Fixed
  • OS: solaris,solaris_9,solaris_nevada,solaris_10
  • CPU: generic,x86,sparc
  • Submitted: 2004-10-04
  • Updated: 2012-10-10
  • Resolved: 2009-06-11
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availabitlity Release.

To download the current JDK release, click here.
Other JDK 6
5.0u3 b05Fixed 6Fixed
Related Reports
Duplicate :  
Duplicate :  
Duplicate :  
Duplicate :  
Duplicate :  
Duplicate :  
Duplicate :  
Duplicate :  
Relates :  
Relates :  
Relates :  
Relates :  
Relates :  
Relates :  
Relates :  
Neither my backspace key nor my delete key have any effect in any field when using Bugster.  The control-H key, however, does work as I would expect backspace to work.
###@###.### 10/4/04 21:47 GMT

Thanks to Alan Harder for the test program texttest.java (see attachments).  This is NOT a bugster problem, but a problem with java, possible related to keyboard identification/keymappings.  I've got a Sun type 6 usb keyboard attached to an Opteron desktop running java 1.5.0 b64, S10_69+amd64 bits and Cinnabar_19.  When I have the numlock turned on, many keys cease functioning, including but not necessarily limited to: backspace, delete, return, tab, and the zero on the number pad.
###@###.### 10/6/04 20:36 GMT

I can confirm this as well.  This also explains why I only see this problem
on my desktop and not on my laptop which runs identical Solaris and Java bits.
###@###.### 10/6/04 21:08 GMT
I would like to request to have this bug re-opened.  We are seeing reports of this issue again and it seems like it has been re-introduced.  We are seeing it in 1.5.0_17  and 1.5.0_18 we did not see this in 1.5.0_14.

We have had several reports of this on the sunray servers within Swan. 

Mike Miller (###@###.### ) has taken several calls from users on this.
why it's marked as 'Fix failed' on 2009-05-06 ?

EVALUATION This bug's Notes suggest that the original Fix Delivered/Available status was incorrect. The problem was never fixed in Java 5.0, and it was never intended to be fixed. Instead, it was fixed in a later version, Java 6.0 (covered under sub-CR 2121519). There were code changes put back in Java 5.0, but they did not fix the problem. Instead, they enabled a workaround. This bug probably should be closed as "Will Not Fix" in Java 5.0, with a note that workarounds exist.

WORK AROUND There are two different workarounds: 1. Set the _AWT_USE_TYPE4_PATCH environment variable to false before starting the Java application. For example: export _AWT_USE_TYPE4_PATCH=false 2. Toggle the "Num Lock" key on the keyboard to disable Num Lock mode.

EVALUATION /**************************************************** **************************************************** ****************************************************/ <B>To enable fix in 1.5.0, please set _AWT_USE_TYPE4_PATCH=false environment variable.</B> /************************************************ ************************************************ ************************************************/ (now it is visible, isn't it:-)

SUGGESTED FIX A quick fix is of workaround variety: allow processing _AWT_IGNORE_XKB environment variable not only on Linux but on every Unix platform. ------- src/solaris/native/sun/awt/awt_GraphicsEnv.c ------- *** /tmp/sccs.54jMhE Wed Nov 3 15:07:10 2004 --- awt_GraphicsEnv.c Wed Nov 3 15:07:00 2004 *************** *** 694,707 **** #endif awt_lock = (*env)->NewGlobalRef(env, this); - #ifdef __linux__ if (getenv("_AWT_IGNORE_XKB") != NULL && strlen(getenv("_AWT_IGNORE_XKB")) > 0) { if (XkbIgnoreExtension(True)) { printf("Ignoring XKB.\n"); } } - #endif dpy = awt_display = XOpenDisplay(NULL); #ifdef NETSCAPE --- 694,705 ---- ###@###.### 11/3/04 12:08 GMT

EVALUATION I think the problem is with the window system or java's ability to set the correct keybindings for the window system. I logged into the user's system and did a remote display of bugster to both gnome and CDE on my sparc/solaris8 system.. both worked fine. ###@###.### 10/5/04 17:35 GMT In existing native key-handling code for both Motif and XAWT, we should carefully process keysyms if NumLock is on, distinguishing subcases "XKB on/off" && "keysym selection convention Xsun/Xorg". ###@###.### 2005-2-18 09:31:39 GMT