Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!uupsi!cmcl2!lanl!bedney@lanl.gov
From: bedney@lanl.gov (Bill Edney)
Newsgroups: comp.sys.next
Subject: Re: Experiences with BugByte
Message-ID: <23533@lanl.gov>
Date: 10 May 91 17:06:40 GMT
References: <506@nwnexus.WA.COM>
Sender: news@lanl.gov
Lines: 48

In article <506@nwnexus.WA.COM> adonis1@nwnexus.WA.COM (Adonis Corporation )  
writes:
> How about a general summary of its features?  What do they really mean
> when they say 'graphic'.  Is it really graphic, and in what respect?
> Is it robust?
> 
> How does one set breakpoints, examine the contents of variables, 
> does it expand structures?  How do you trace through executing code?  
> Can you examine/trace through at the assembler level? 
> 
> Just a few simple question like that...
> 
> Thanks,
> 
> Doug Kent
> Independent NeXT Developer (what does that really mean?)
> adonis1@nwnexus.wa.com

Doug -
BugByte is essentially a graphical interface on top of the GNU debugger. I  
think it's about as graphical as a debugger can get. A separate window has a  
browser which lists the various object classes in the first column. When you  
click on one of the classes, the second column in the browser fills with all of  
the methods in that class. When you click on one of those, the source code for  
that method appears in the main window. One sets breakpoints by double clicking  
on the line and clicking the "set" icon, after which a small stop sign appears  
in a column to the left of the selected line. You can then click "run" to run  
the program and when your program hits the breakpoint, you are put back into  
BugByte. A black arrow shows the line that you are currently at. From here, you  
can step through, either by clicking "step" (which will step you into all  
functions), or "next" which will step over functions and keep you in your  
current source file. You can examine the contents of variables by double  
clicking them and then clicking either the "Print" button or the "Print *"  
button. I'm not sure what you mean by "expanding structures". I can click on a  
pointer to a struct and then click "Print *" and it gives me back all of the  
values of the fields in that struct. You cannot examine/trace at the assembler  
level, as far as I know. One nice feature is that you can edit your source  
right in the main window, save it and click a "Make" option. It will "make  
debug" your code and then reload the libraries automatically. You can then just  
click run. It also has a neat feature for examining the stack.

Anything else you wanted to know?

- Bill Edney
- Los Alamos National Laboratory

"I don't speak for my employer and they don't speak for me. Kind of a nice  
arrangement."