From:	root AT jacob DOT remcomp DOT fr (root)
Subject:	gdb: a positive note
13 Oct 1998 09:08:05 -0700 :
Message-ID:	<m0zSkq4-0002fdC.cygnus.gnu-win32@jacob.remcomp.fr>
To:	gnu-win32 AT cygnus DOT com

OK, enough complaints.

Now lets look at the good sides of gdb.

First of all, it has superior expression parsing software as my debugger. The
expression parser is older and more debugged than mine. 

Second, you can get a list of all defined types easily. (Of course you
should not attempt to type <q> to stop it... :-). This is now not possible
in my debugger. Obviously I am not going to leave this like that forever... but
for the time being this points to a lack of functionality of my debugger.

The disassembler display can receive an argument. You can disassemble any
known (to the debugger) address. Mine displays always only the current
function. On the other side, I can show the disassembly even if I am in a
system function. Gdb can't.

The watch variable (display in gdb's vocabulary) feature definitely works.
The same can't be said from my debugger...

The 'step' command should step into the called subroutines. Gdb does this,
but it is unable to trace into indirect subroutine calls like:

	hwnd = CreateWindow(...)

CreateWindow produces a WM_CREATE message, that calls MainWndProc, a procedure
that is known to gdb. This is not detected by gdb. To be fair, I learned how
to do this only last week...

Very impressing is the 'call subroutine' feature... I am planing to implement
that later this year, but gdb has it now.

-- 
Jacob Navia	Logiciels/Informatique
41 rue Maurice Ravel			Tel 01 48.23.51.44
93430 Villetaneuse 			Fax 01 48.23.95.39
France
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request AT cygnus DOT com" with one line of text: "help".

- Raw text -