This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
The problem appears in the following scenario: 1. Load Extension Library to Default Device. 2. Stop Default Device. 3. Invoke "Resume(Debug)" for Default Device. 4. Invoke Debug on Extended Applet project that depends on the Extension Library. Build script stops after the following message: "OffCard Installer [v3.0.2] Copyright (c) 2009 Sun Microsystems, Inc. All rights reserved. Use is subject to license terms. [Unloading the bundle ExtendedApplet6 from http://localhost:8039/cardmanager/] [ INFO: ] "Off Card Installer validating unload information" [ INFO: ] "Off Card Installer sending unload request" Connection refused: connect load-bundle: " No output is displayed in Default Device output tab. Apparently Debug Proxy isn't started after Resume(Debug).
This is rather important issue. It prevents from debugging of Java Card projects that depend on some libraries that should already be present on device. Also it prevents from debugging scenarios that check persistence features.
Build script stops or exits? If you invoke $JDK_HOME/bin/jps.exe do you see the debugger proxy running? If it is not there, then that is the problem. "Connection refused" suggests that it is not started.
> It prevents from debugging of Java Card projects that depend on some libraries that should already be present on device. You can still simply invoke Debug on the project, and the libraries and the project will be deployed and the debugger will be started, correct? If so, then this issue does not prevent you from debugging such projects. It just prevents you from simulating the exact scenario of connecting to a card which already has the library on it. I will agree that is not good, but I do not think it is P2 not-good.
(In reply to comment #2) > Build script stops or exits? Debug script stops after "Connection refused: connect". cjcre.exe process is terminated. > If you invoke $JDK_HOME/bin/jps.exe > do you see the debugger proxy running? If it is not there, then that is the > problem. "Connection refused" suggests that it is not started. No there's no debugger proxy process.
(In reply to comment #3) > > It prevents from debugging of Java Card projects that depend on some libraries that should already be present on device. > > You can still simply invoke Debug on the project, and the libraries and the > project will be deployed and the debugger will be started, correct? Only if "Deploy To Card" strategy is used for all dependencies. > If so, then this issue does not prevent you from debugging such projects. It > just prevents you from simulating the exact scenario of connecting to a card > which already has the library on it. I will agree that is not good, but I do > not think it is P2 not-good. It's up to you to decide the priority. But dependency is not the only problem. We can't test in debug mode anything that requires card reset.
this seems to require more evaluation
Sounds like the problem may be as simple as the debug proxy process is not started. Bear in mind there is some extremely fragile timing code - if the debug proxy is started too soon after the emulator, it will crash the emulator when it tries to connect. So it also may be that it is started, but there is a timing issue that causes it to exit. I put pretty granular logging in the javacard modules, which might help to figure out what's really happening.
Setting to P3 as the primary debugger functionality is not affected, just some particular use case.
changing owner