Re: jde-2.3.3 error by jde-bug-evaluate-expression
Petter Måhlén [EMAIL PROTECTED] writes: The exception is throw by the Java side of JDEbug. It keeps a map of object id (269 in this case) and object reference in the process being debugged. The exception indicates that it cannot find the object with reference 269. I haven't looked at the 2.3.3 code, but if I remember right, in previous versions, JDEbug never cleaned out the object store. I would say that the most likely sources of the problem are: - the elisp side of JDEbug is using an invalid object ID (due to a bug), or - the Java side of JDEbug is nowadays cleaning out the object store, and the object in question has been cleaned out from the Java side but not the elisp side (meaning there is some bug in the cleaning out of objects). More questions in addition to what Andrew asked: does this happen immediately and always for the same object? Or do you have to run the debugger for some time first? Does it matter which program you are debugging? It happens for all objects in every small test contexts. The debug window 'JDEbug 1' appears on the left side - unfortunatly only under Linux with a width of zero. In the JDEbug 1 window the values look fine. I did this test in a test scenario with 'emacs -q'. Ok, I am now seeing that this feature is maybe now not so important because looking to the JDEbug 1 window is more convenient. How can I recompile the java files because I want to fix the width problem? I do not found any build or make file. Harald
Re: jde-2.3.3 error by jde-bug-evaluate-expression
Does it ask for an expression to evaluate first? Harald Maier [EMAIL PROTECTED] writes: Hello all, I installed jde-2.3.3 with eieio-0.17 elib-1.0 semantic-1.4.4 speedbar-0.14beta4 When I evaluate 'jde-bug-evaluate-expression' in a JDEbug session I get the following error: , | Error: cannot get object 269. | Reason: Exception during command execution: jde.debugger.JDEException: No such object exists. | eieio-oref: Wrong type argument: (or object-p class-p), nil ` jde-2.3.2 works fine on this operation. Is this a known bug or it it related to a configuration problem? Harald
Re: jde-2.3.3 error by jde-bug-evaluate-expression
Andrew Hyatt [EMAIL PROTECTED] writes: Does it ask for an expression to evaluate first? Yes, it asks for an expression and afterwards the above error is signaled. When I evaluate 'jde-bug-evaluate-expression' in a JDEbug session I get the following error: , | Error: cannot get object 269. | Reason: Exception during command execution: jde.debugger.JDEException: No such object exists. | eieio-oref: Wrong type argument: (or object-p class-p), nil ` Harald
Re: jde-2.3.3 error by jde-bug-evaluate-expression
What sort of thing are you trying to evaluate? Perhaps the bug is that the evaluation returns null (legitimately, perhaps) and throws an error incorrectly? Or perhaps a variable has gone out of scope and you are trying to evaluate it? Harald Maier [EMAIL PROTECTED] writes: Andrew Hyatt [EMAIL PROTECTED] writes: Does it ask for an expression to evaluate first? Yes, it asks for an expression and afterwards the above error is signaled. When I evaluate 'jde-bug-evaluate-expression' in a JDEbug session I get the following error: , | Error: cannot get object 269. | Reason: Exception during command execution: jde.debugger.JDEException: No such object exists. | eieio-oref: Wrong type argument: (or object-p class-p), nil ` Harald
RE: jde-2.3.3 error by jde-bug-evaluate-expression
The exception is throw by the Java side of JDEbug. It keeps a map of object id (269 in this case) and object reference in the process being debugged. The exception indicates that it cannot find the object with reference 269. I haven't looked at the 2.3.3 code, but if I remember right, in previous versions, JDEbug never cleaned out the object store. I would say that the most likely sources of the problem are: - the elisp side of JDEbug is using an invalid object ID (due to a bug), or - the Java side of JDEbug is nowadays cleaning out the object store, and the object in question has been cleaned out from the Java side but not the elisp side (meaning there is some bug in the cleaning out of objects). More questions in addition to what Andrew asked: does this happen immediately and always for the same object? Or do you have to run the debugger for some time first? Does it matter which program you are debugging? / Petter -Original Message- From: Andrew Hyatt [mailto:[EMAIL PROTECTED] Sent: den 26 januari 2004 19:32 To: Harald Maier Cc: Andrew Hyatt; [EMAIL PROTECTED] Subject: Re: jde-2.3.3 error by jde-bug-evaluate-expression What sort of thing are you trying to evaluate? Perhaps the bug is that the evaluation returns null (legitimately, perhaps) and throws an error incorrectly? Or perhaps a variable has gone out of scope and you are trying to evaluate it? Harald Maier [EMAIL PROTECTED] writes: Andrew Hyatt [EMAIL PROTECTED] writes: Does it ask for an expression to evaluate first? Yes, it asks for an expression and afterwards the above error is signaled. When I evaluate 'jde-bug-evaluate-expression' in a JDEbug session I get the following error: , | Error: cannot get object 269. | Reason: Exception during command execution: jde.debugger.JDEException: No such object exists. | eieio-oref: Wrong type argument: (or object-p class-p), nil ` Harald
JDE Bug : Error while compiling
olivier Fresse writes: Please enter the details of your bug report here When I compile, I have an error : let: Wrong type argument: listp, jikes I used jde-2.2.8, no problem. The data type of jde-compiler has changed in the latest release. Please open a customization buffer for jde-compiler, select reset to standard setting to get the new data type, close the buffer, reopen, and select jikes. Save the new setting in your project file. - Paul
Problems calling jde-bug-display-variable
Hi, I am wondering if the function' jde-bug-display-variable' normally should display the value of a (local) variable? In my installation it causes the following lisp error (with cursor at a variable named 'character'): Debugger entered--Lisp error: (wrong-type-argument integerp character) mapconcat(identity ((character variable int nil nil nil (...) [4031 4045])) .) (jde-bug-evaluate-expression (mapconcat (quote identity) qualified-expr .)) (if qualified-expr (jde-bug-evaluate-expression (mapconcat ... qualified-expr .))) (let ((qualified-expr ...)) (if qualified-expr (jde-bug-evaluate-expression ...))) jde-bug-display-variable() My installation: NTEmacs 21.2.1 jde-2.2.9 semantic-1.4 speedbar-0.14beta4 elib-1.0 eieio-0.17 Any hint appreciated. Thanks, juergen
jde-bug-jdk-directory
Does this var still exist? Regards, Daniel _ MSN Photos is the easiest way to share and print your photos: http://photos.msn.com/support/worldwide.aspx
jde-bug-jdk-directory
Daniel Hegyi writes: Does this var still exist? It was removed in a recent beta as it is no longer necessary with the advent of the jde-jdk-registry/jde-jdk system for registering the presence of JDKs on your system. - Paul
Re: jde-bug-jdk-directory
hasn't it been superseded by jde-jdk-registry (in 2.2.9)? Daniel Hegyi writes: Daniel Does this var still exist? Regards, Daniel Daniel _ Daniel MSN Photos is the easiest way to share and print your photos: Daniel http://photos.msn.com/support/worldwide.aspx -- Danny Siu
jde-bug-thread-show-thread-info ??
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 (defun jde-bug-thread-show-thread-info () (interactive) (message not implemented)) What would it do IF it were implemented??? - -- Kevin A. Burton ( [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] ) Location - San Francisco, CA, Cell - 415.595.9965 Jabber - [EMAIL PROTECTED], Web - http://relativity.yi.org/ Learn from other people's mistakes, you don't have time to make your own. -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.6 (GNU/Linux) Comment: Get my public key at: http://relativity.yi.org/pgpkey.txt iD8DBQE8WlonAwM6xb2dfE0RAtZHAJoDHazuXSeX48DCcqw1URZ3pdc3gACffalG lRUpLdIekjAtu/zqB7Vl1gg= =TWCq -END PGP SIGNATURE-
Better interactive definition for jde-bug-attach-local-host
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Doesn't this make sense. By default is uses the port for input. I got sick of typing '2112' every time :) (defun jde-bug-attach-local-host (process-port) Attaches the debugger to a process running on local host. This command connects to the process via a socket. (interactive (list (read-string Process Port: (cdr jde-bug-server-socket - -- Kevin A. Burton ( [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] ) Location - San Francisco, CA, Cell - 415.595.9965 Jabber - [EMAIL PROTECTED], Web - http://relativity.yi.org/ And the beast shall come forth surrounded by a roiling cloud of vengeance. The house of the unbelievers shall be razed and they shall be scorched to the earth. Their tags shall blink until the end of days. - from The Book of Mozilla, 12:10 -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.6 (GNU/Linux) Comment: Get my public key at: http://relativity.yi.org/pgpkey.txt iD8DBQE8WwB1AwM6xb2dfE0RApZwAKCjulCSWftxWLBbY2I3gieAwz1D4QCeINGT uGXCWRHRoGhCPj3x6HhEsb0= =qBX9 -END PGP SIGNATURE-
jde-bug-continue should set overlay-arrow-position to nil.
Hm. jde-bug-continue leaves the overlay-arrow around after it runs. Obviously this is incorrect as the VM will be running on its own at this point. Kevin -- Kevin A. Burton ( [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] ) Location - San Francisco, CA, Cell - 415.595.9965 Jabber - [EMAIL PROTECTED], Web - http://relativity.yi.org/ Linux is kind of like UNIX but it isn't as expensive, isn't fragmented across multiple vendor, doesn't restrict your Freedom, and is just infinitely more pleasant.
RE: Does jde-bug-attach-local-host depend on jde-run???
No, it does not depend on jde-run. I just tried attaching to a process started from the command line, by shared memory and through a socket(On local host) and both work. Did you start the server with the appropriate parameters? -Xdebug -Xnoagent -Djava.compiler=NONE -Xrunjdwp:transport=dt_socket,server=y,address=serverport,suspend=n Javier
Does jde-bug-attach-local-host depend on jde-run???
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 OK. I want to connect to an external VM *outside* of Emacs. Specifically because this way I can exit Emacs without shutting down my server. Does JDEBug depend on me calling jde-run? If I try to do a jde-bug-attach-local-host I get the error: eieio-default-superclass: Unbound slot: #class jde-dbs-debugger, #jde-dbs-debugger JDEbug, process, oref The Attach Process | On local host command is grayed out... (this was from 2.2.8 and from CVS) Kevin - -- Kevin A. Burton ( [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] ) Location - San Francisco, CA, Cell - 415.595.9965 Jabber - [EMAIL PROTECTED], Web - http://relativity.yi.org/ Any programming language is at its best before it is implemented and used. -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.6 (GNU/Linux) Comment: Get my public key at: http://relativity.yi.org/pgpkey.txt iD8DBQE8CHulAwM6xb2dfE0RAkM9AKC94fdEvalzgtbH89Y8KrlgQiJWwwCg0grH q4z5NjzFA2P7LN3x+piMxFE= =eMXd -END PGP SIGNATURE-
Re: Thoughts on jde-bug
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Paul Kinnucan [EMAIL PROTECTED] writes: Kevin A. Burton writes: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Paul Kinnucan [EMAIL PROTECTED] writes: snip After you've made a few successful submissions via me. Note that your submissions should be thoroughly tested before submission and work on all platforms supported by Emacs/XEmacs/JDK and all modern versions of Emacs and I try to write code as portable as possible obviously. The biggest screw up I have made in the past is the use of font-lock-add-keywords as this is not portable on XEmacs. XEmacs and handle all cases, e.g., no feature that works only for special cases. Hm. Later Emacsen support features such as images, etc that would be nice to take advantage of. Of course the code can be written to work on all platforms as we could enable the uplevel features when running in downlevel clients. I would like the JDE to take advantage of these features but not require them. My thoughts as well. Thus, it would be great to provide toolbar buttons for the most common debug commands (set/clear breakpoint, step in, run/continue, etc). The code for this should check to be sure that a toolbar is available (it is not on the Windows version of Emacs 21). Really? No toolbar on Windoze? hm Kevin - -- Kevin A. Burton ( [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] ) Location - San Francisco, CA, Cell - 415.595.9965 Jabber - [EMAIL PROTECTED], Web - http://relativity.yi.org/ Windows XP. Now with more evil in every box! -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.6 (GNU/Linux) Comment: Get my public key at: http://relativity.yi.org/pgpkey.txt iD8DBQE7+M59AwM6xb2dfE0RApZRAKDPb1njV7OZIw8HT4B6Hou3XNhsoACgqb9S zM/PE9R+rw54KzOty3FYRUQ= =gJ5c -END PGP SIGNATURE-
RE: jde-bug should use * special buffers
I like your idea of using the *. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] Sent: Friday, November 16, 2001 1:20 PM To: JDE Mailing List Subject: jde-bug should use * special buffers -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I can't help but think that jde-bug should use special buffers. AKA intead of buffer names like Process classman.Classman(4) I think we should use: *Process classman.Classman(4)* Any thoughts here? If anyone likes the current behavior then could make this an option one could customize? Kevin - -- Kevin A. Burton ( [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] ) Location - San Francisco, CA, Cell - 415.595.9965 Jabber - [EMAIL PROTECTED], Web - http://relativity.yi.org/ In the end, only kindness matters. -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.6 (GNU/Linux) Comment: Get my public key at: http://relativity.yi.org/pgpkey.txt iD8DBQE79WbEAwM6xb2dfE0RAsd7AKC/BBjfDvh/jgZK500E8JVLAjxBCwCfae9L djz5xp9VgMkWqrVCPrr1H0k= =oENF -END PGP SIGNATURE-
[jde-bug] before setting a breakpoint, make sure the current file is in jde-db-source-directories
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 OK. This doesn't make sense. Right now it is possible to toggle a breakpoint in a file that is not in jde-db-source-directories. Isn't this bad? Shouldn't we signal an error if the user does this? Kevin - -- Kevin A. Burton ( [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] ) Location - San Francisco, CA, Cell - 415.595.9965 Jabber - [EMAIL PROTECTED], Web - http://relativity.yi.org/ Boycott Amazon.com http://www.gnu.org/philosophy/amazon.html -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.6 (GNU/Linux) Comment: Get my public key at: http://relativity.yi.org/pgpkey.txt iD8DBQE79ZeyAwM6xb2dfE0RAgcSAJ9uJpIzUTTUYI89R5POmMNf3ffAPACgzhwP 4c6o+hoQCAlmS5PUSwUwhEg= =FGuk -END PGP SIGNATURE-
Thoughts on jde-bug
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 OK. I just wanted to serialize my thoughts here. In the past I have not used debuggers. Primarily because I knew all of my source code and understood what was going on. Recently I have been working on a lot of projects where I don't know the code thus I find myself wanting to play with jde-bug. It seems like jde-bug is a little rough around the edges for new users. Specifically there are some things it should catch when thing fail. I mean when things work, they work great but when things fail, they fail silently and without any debug information. This is one area of improvement. The other main area isn't really our fault but the VMs. It seems like the JDK 1.3 VM has problems in a lot of places. (I can't put a breakpoint in main() ). Does anyone know how the 1.4 VM behaves? I can't run it for a number of reasons. Has anyone thought about integrating this with the ECB? Could we add the jde-bug-menu to the other buffers? (CLI, Process, Local Variables, etc). Also. It seems like we should be able to handle things a little better when things go wrong. If the JDEBug process terminates with an error. We should probably display this to the user. Anyway... I will probably grab the latest beta and play with this. This should allow me to work on patches as I break it. Other than that everything seems great! After all this was just constructive criticism. :) Kevin - -- Kevin A. Burton ( [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] ) Location - San Francisco, CA, Cell - 415.595.9965 Jabber - [EMAIL PROTECTED], Web - http://relativity.yi.org/ Did you hear why they're using Windows 2000 as a prison guard? Because it always locks up! -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.6 (GNU/Linux) Comment: Get my public key at: http://relativity.yi.org/pgpkey.txt iD8DBQE79aDPAwM6xb2dfE0RAnqQAKCHa2Pv8RqLIUCae9wtxNIuFvLWbACgyPwy T7PSciLzaJ4tJT+/qrKC9vM= =O67r -END PGP SIGNATURE-
jde-bug-key-bindings
Hello, I took Paul's advice of customizing jde-bug-key-bindings to setup my functions keys for stepping in JDEbug. Previously they were set using global-set-keys. Unfortunately, I must be specifying the function keys incorrectly. If I do a C-h k F6, for example, I get: f6 is undefined Here is the definition from my .emacs: '(jde-bug-key-bindings (quote (([F6] . jde-bug-step-over) ([F7] . jde-bug-step-into) ([F8] . jde-bug-step-into-all) ([F9] . jde-bug-step-out) ([? ? ?] . jde-bug-continue) ([? ? ?] . jde-bug-toggle-breakpoint Can someone help me with the correct definition of these keys in customize. Thanks. Jeff
Re: key bindings of the jde-bug
- Original Message - From: Stefan Holst To: [EMAIL PROTECTED] Sent: Friday, August 03, 2001 2:03 PM Subject: key bindings of the jde-bug Dear Paul, I'm using the jde for quit a while after installing the latest version 2.2.7.1 I've tried the JDEbug the first time. I think it is really great, but in my xemacs 21.4 for solaris I don't have any key-bindings for the entries of the JDEbug menu. They get displayed with jde-bug-keys but are not bound to the keymap ony buffer as C-h b should display. At least when I start the debuger with C-c C-v C-d. Any help would be apreciated very much. Thanks in advance I implemented JDEbug as a minor mode. I don't understand how minor mode keybindings work on XEmacs. Hence the lack of keybindings for JDEbug. My intention is to research this issue and provide keybindings for XEmacs as soon as I have some free time. - Paul
Re: JDE Bug
Paul Kinnucan wrote: At 10:33 PM 6/6/2001 -0700, Ben Wing wrote: Paul Kinnucan wrote: At 08:37 AM 6/6/2001 -0600, Brad Porter wrote: Hello. I am finding that I have to modify 'jde-compile.el' in order to get the JDK -d option to work correctly under cygwin. I set directory-path-sep to ?/ in my .emacs file for the rare times I use XEmacs (for testing JDE compatibility).This eliminates the need for quoting. Note that the default separator under NT/Emacs is forward slash. is this change required for jde to work in xemacs? Presently yes, but not in the next release. i certainly hope not. Why? a package that requires a non-standard setting of an application is not a well-behaved package. i'm glad you've fixed jde to work with either directory-path-sep, and hopefully we can work together to get the new version in xemacs. I systematically use forward slashes as the path separator on Windows and I have never experienced the slightest problem in doing so. On the other hand, I regularly get email from new users who think they have to use backslashes in paths on Windows and who are unaware that back slash is the Emacs string escape character. This result is that they innocently put things like (setq jde-compile-option-directory c:\foo\bar) in their .emacs file and are puzzled by the result (c:foobar). I applaud your determination to make XEmacs a native Windows application. However, I think in this case you are mistaken on two counts. First, forward slash is a legitimate Windows path separator so making it the default sep character in no way violates your goal. On the other hand,. making back slash the default sep character conflicts with the established Emacs meaning of backslash, which is escape the next character. This creates endless headaches for new users and for people who have to support them. - Paul -- ben I'm sometimes slow in getting around to reading my mail, so if you want to reach me faster, call 520-661-6661. See http://www.666.com/ben/chronic-pain/ for the hell I've been through.
JDE Bug speed question
Hi, I find that JDE bug can take like 10 seconds or so when I step-over sometimes (quite often). Does anyone else find this ? :)
RE: JDE Bug
Ah... The problem seems to be that I have 'jde-cygwin-path-converter' set to 'jde-cygwin-path-converter-cygpath'. This calls the cygwin command 'cygpath' which is the culprit of putting the backslashes in. Since this command is shelled out, the directory-sep-char is a moot point. I prefer to use the 'jde-cygwin-path-converter-cygpath' setting because it allows more UNIX like paths in my settings - and when I can use relative paths, UNIX to NT porting is easier. Are there any other suggestions? Is there anything wrong with quoting the directory in the -d option/ Thanks. -brad -Original Message- From: Paul Kinnucan [mailto:[EMAIL PROTECTED]] Sent: Wednesday, June 06, 2001 8:53 AM To: Brad Porter Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: Re: JDE Bug At 08:37 AM 6/6/2001 -0600, Brad Porter wrote: Hello. I am finding that I have to modify 'jde-compile.el' in order to get the JDK -d option to work correctly under cygwin. I set directory-path-sep to ?/ in my .emacs file for the rare times I use XEmacs (for testing JDE compatibility).This eliminates the need for quoting. Note that the default separator under NT/Emacs is forward slash. I think it was a mistake on the part of the NT/XEmacs developers to make the default separator a back slash. First, there was no need to do it as Windows applications don't care whether the slash is forward or backward. Secondly, it makes life difficult for users who uses applications, such as bash, that treat the forward slash as an escape character. - Paul smime.p7s
RE: JDE Bug
At 09:23 AM 6/6/2001 -0600, Brad Porter wrote: Ah... The problem seems to be that I have 'jde-cygwin-path-converter' set to 'jde-cygwin-path-converter-cygpath'. This calls the cygwin command 'cygpath' which is the culprit of putting the backslashes in. Since this command is shelled out, the directory-sep-char is a moot point. I prefer to use the 'jde-cygwin-path-converter-cygpath' setting because it allows more UNIX like paths in my settings - and when I can use relative paths, UNIX to NT porting is easier. Are there any other suggestions? Use the JDE's relative path feature? - Paul Is there anything wrong with quoting the directory in the -d option/ Not if it is done in a way that does not break compile for non-cygwin users and if someone is willing to provide correct, tested code that meets this condition. Otherwise it might not get done for a while as it is not a high priority item for me. - Paul
Re: JDE Bug
Paul Kinnucan wrote: At 08:37 AM 6/6/2001 -0600, Brad Porter wrote: Hello. I am finding that I have to modify 'jde-compile.el' in order to get the JDK -d option to work correctly under cygwin. I set directory-path-sep to ?/ in my .emacs file for the rare times I use XEmacs (for testing JDE compatibility).This eliminates the need for quoting. Note that the default separator under NT/Emacs is forward slash. is this change required for jde to work in xemacs? i certainly hope not. I think it was a mistake on the part of the NT/XEmacs developers to make the default separator a back slash. First, there was no need to do it as Windows applications don't care whether the slash is forward or backward. Secondly, it makes life difficult for users who uses applications, such as bash, that treat the forward slash as an escape character. obviously we disagree. the explicit goal of xemacs was to make it a *native* windows app, not a windows app pretending to be a unix one. - Paul -- ben I'm sometimes slow in getting around to reading my mail, so if you want to reach me faster, call 520-661-6661. See http://www.666.com/ben/chronic-pain/ for the hell I've been through.
Re: Why can't I attach to weblogic using JDE-bug?
IAt 01:53 PM 2/12/01 -0800, you wrote: Hi, I've installed the lastest version of JDE packages on my machine. Text editing and speedbar work very well on my machine. However, I encounter some problem when I try to attach to a running weblogic server. I started the server using the following options from a script(not from emacs, I'm use bea's jdk130 included in weblogic 6.0): -Xdebug -Xnoagent -Djava.compiler=NONE -Xrunjdwp:transport=dt_shmem,address=weblogic,server=y,suspend=n I can attach to it easily using jdb outside emacs: jdb -attach weblogic According to the documentation, I set the following variables: '(jde-db-debugger (quote ("JDEbug" "" . "Executable"))) '(jde-bug-vm-includes-jpda-p t)) '(jde-bug-jdk-directory "d:/bea/jdk130/") But when I try to use jde-bug-attach-via-shared-memory weblogic, it gives the error with the following stack trace: [snip] Could anyone help me out? Thanks! Please submit a full problem report. - Paul HOW TO ASK FOR TECH SUPPORT When requesting technical support, please generate a problem report, using the JDE-Help-Submit Problem Report command, and submit this report along with your help request. Please do this even if you think most of the report is irrelevant. If you knew what were relevant, you wouldn't need my help. To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] JDE website: http://sunsite.auc.dk/jde/ JDE mailing list archive: http://www.mail-archive.com/jde@sunsite.auc.dk/maillist.html