> -----Original Message----- > From: Paul Kinnucan [mailto:[EMAIL PROTECTED]] > Sent: Tuesday, June 27, 2000 11:39 AM > To: Kellner, Sean > Cc: [EMAIL PROTECTED] > Subject: RE: FAQs > > > At 10:07 AM 6/27/00 -0400, you wrote: > > As someone who asked one of those FAQs on here, maybe I can > help a little > >too. Maybe even this time I won't be too redundant. > > > > First off, I've been an (X|Lucid|FSF)Emacs user for 10+ > years, but never > >really tried the Lisp stuff. I'm really just a user. > > > > It's really difficult to really find what I am looking for > when it comes to > >JDE. One the web site, it appears to want to be a > navigation/toc on the left > >frame, contents on the right. So then why does seemingly > every link pop up a > >new window? It makes navigating the site looking for something more > >difficult than it has to be, at least for me. > > > > The main TOC for the JDE website has 14 entries. Of these, > othree open new > frames. Of these, one is the JDE user's guide, which has its > own TOC, so to > open this in the contents pane of the main frameset would be ugly. The > other two are the contributions page and the troubleshooting guide. I > suppose that when I created these a couple of year's ago I > was planning on > them also having TOCs. Anyway, when I get a chance I'll change them to > display in the main pane of the main frame. > > Sorry. > My main intent was never to say that anyone did a bad job, or had a reason to apologize. Quite the opposite, I am very grateful for the job the JDE team is doing. Each work day is a better day because of you guys. > > > > Searching the mailing list is extremely frustrating. For a > while, every > >time I hit the ';' key, it would insert a return and indent > me on the next > >line. I really did not want this behavior. I was able to > turn it off in > >C-mode, but it would come back whenever I entered JDE. Just > try searching > >the mailing list archive for that one. There were a couple > of variables I > >figured were involved, searching for them revealed every > person who included > >their .emacs in the message or other helpful debugging > content. So I was > >afraid to post on here again and lived with it for a while > until I finally > >had some time to track it down. > > > > As for additional docs on JDEbug, the existing ones were > nearly adequate to > >get me going. Never using JPDA before, I didn't realize that > the Classic JIT > >in 1.2.2 didn't like the debug options and would give an > illegal access > >error in windows. Apparently HotSpot is OK with it, and > looking at the Sun > >JPDA site I found that -Xnoagent and -Djava.compiler=NONE > were necessary. > >Had these two flags been mentioned on the JDEbug page, I > would have been up > >and running in minutes of curiosity. > > > > These were inadvertently omitted from the doc. When you are > writing doc at > 3 am and anxious to get features into the hands of users, > errors tend to > occur. Sorry. > I completely understand. Maybe that's one way I can contribute. > > As an aside, JDEBug works much better than the Cafe > debugger for debugging > >EJBs in WebLogic 5.1. > > > > As another aside, while on the topic of useful JDEbug tidbits ... > > > > Setting a breakpoint before attaching to a process is > pointless, it just > >gets ignored. > > > > You are dealing with JDEbug 1.0; there are lot of areas that need > polishing, one of them is breakpoint handling. These issues will be > addressed in upcoming releases. Coolness. > > > I have no idea why I can't get the CLI window up, and in > some situations > >would give anything to get that window. > > > > JDEbug currents supports the CLI window only for processes that it > launches. For these processes, you can display the CLI buffer > by selecting > JDEbug->Show Buffer->CLI. I'm not sure whether it is possible > to connect to > the standard I/O of attached processes. I plan to investigate > this issue > and will provide CLI support if JPDA supports it. Ah, so the problem is really my misunderstanding of what the CLI is. I saw it mentioned in the menu, and just assumed it was a way for me to enter commands to the debugger by hand. Sometimes it's easier for me to type "stop in mypackage.MyClass.foo(java.lang.String, int)" than it is to find the method and click. As for attaching to standard i/o of attached processes, my vote would be for "Dont worry about it", however much weight that might carry. > > A word of explanation. The Java side of JDEbug was written > last summer by > Amit Kumar, a summer intern hired by Sun's JPDA team > specifically for the > purpose of helping me develop JDEbug. I wrote the Lisp side of the > debugger. Amit's Java code is basically a first draft. It > does the basic > job but there are many rough spots. I am currently in the process of > learning, revamping, and upgrading Amit's code. This should > result in a > steady, gradual increase in functionality and robustness over upcoming > releases. My plan is to make many small releases in order to get the > improvements into your hands ASAP. And I do believe that is the correct model to follow for stuff like this. > > You can contribute in any way you like, including fixing bugs > in the lisp > and/or Java code, implementing new features or enhancing > existing ones, > writing documentation, or simply reporting bugs and providing > suggestions > for enhancements. I can use any help you can give me. Ya know, I keep forgetting that parts of the JDE are written in Java. I just might be able to help out a bit on that side. And as always, Thanks.
