Re: Routine_ID Is No Doubt My Solution...
On Tue, 15 Apr 2003 15:30:21 +1000, Derek Parnell
<ddparnell at bigpond.com> wrote:
>> 4) Save the status of the program in order to continue execution =
later.
>
>Hmmm. Never really needed to do this.
Possibly badly worded. Resume the status of a debug session
definitely. I want to start the next debug session with result,
params, local var x etc showing their current values.
I'll not be the first dumb numbskull that needs more than one debug
session to solve a thorny problem & I freely admit that.
The debugger should allow a "permanent view" to be defined - not as it
stands overwrite randomly/sequentially those variables you are
interested in tracking with "fluff"; eg for i=3D1 to 10; you *know*
what i is, cause you're single-stepping the loop; when it blats the
current value of i over a value you are waiting to see go wrong, it's
a pain, 'cos you have to "?" it again to get it back on the screen.
Some kind of ini file, only read when the debug/trace window opens?
The debugger is smart enough to cope if I define some variable such as
"result" which is not in scope at the trap point & say "<undefined>"=20
The other one I'd vote for is viewing the *end* of a large sequence in
the permitted 80-char view. I know 2.4 allows you to see the whole
thing, but when you are single-stepping through, waiting for it to go
wrong...
Pete
|
Not Categorized, Please Help
|
|
