comp.lang.idl-pvwave archive: archive » Re: .trace not working?

Home » Public Forums » archive » Re: .trace not working?

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: .trace not working? [message #67588 is a reply to message #67584]

Thu, 06 August 2009 18:40

David Fanning
Messages: 11724
Registered: August 2001

Senior Member

jkj writes:

> I have always used ".cont" to resume execution and when I want to know
> the nesting of calls by which I arrived at the given stop I enter
> "blah", giving me a traceable path up through the calling code.
>
> It certainly sounds like ".trace" is exactly equivalent to ".continue"
> but nobody wants to address it clearly - weird. I can't imagine why
> IDL supplies ".trace" and ".continue" without giving a clear reason
> about the differences - .trace is clearly misnamed if all it does is
> continue execution... must be something more with a bit deeper
> digging, but I'm not getting my shovel out for that one.
>
> For object code I create a function call for each object
> "::get_hack_val" that returns a '1' when I want to trace execution
> paths, so every routine checks "->get_hack_val()" and if it returns
> "1" then a print of the routine's name and any other helpful
> information (status of keywords, for example) is printed.

It is always interesting to me to see how people debug
their code. I'm always most interested in the people who
don't read their error messages. They seem almost a different
breed to me.

But if you get to a stop, and you want to know how you
got there, I don't see how "blah" can help. I would
think HELP, /TRACEBACK might do more good. :-)

I debug in the Workbench, of course, so when I come to
a stop I usually just step over routines to the next
line of code (F6) or step into routines (F5). Or, if
I want to run awhile, just put my cursor on the line
I want to run to and type CNTL-R. I don't usually feel
a need for anything fancier than that.

Debugging mostly involves what goes on between your
ears, and it especially helps if you are not closely
wedded to your own ideas of how you expect things to
be. A certain ability to not take yourself too
seriously is what makes a good debugger, I think.

Cheers,

David

--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Sepore ma de ni thui. ("Perhaps thou speakest truth.")

Report message to a moderator

[Message index]

		Re: .trace not working? By: BLemire@ittvis.com on Sat, 08 August 2009 20:14
		Re: .trace not working? By: Jeremy Bailin on Sat, 08 August 2009 19:19
		Re: .trace not working? By: penteado on Fri, 07 August 2009 20:06
		Re: .trace not working? By: David Fanning on Fri, 07 August 2009 13:36
		Re: .trace not working? By: JDS on Fri, 07 August 2009 13:11
		Re: .trace not working? By: penteado on Fri, 07 August 2009 10:44
		Re: .trace not working? By: David Fanning on Thu, 06 August 2009 18:40
		Re: .trace not working? By: jkj on Thu, 06 August 2009 17:52
		Re: .trace not working? By: Bruce Bowler on Thu, 06 August 2009 06:44
		Re: .trace not working? By: David Fanning on Thu, 06 August 2009 05:18
		Re: .trace not working? By: Bruce Bowler on Thu, 06 August 2009 05:13
		Re: .trace not working? By: Michael Galloy on Wed, 05 August 2009 17:39
		Re: .trace not working? By: BLemire@ittvis.com on Wed, 05 August 2009 14:16
		Re: .trace not working? By: JDS on Mon, 10 August 2009 10:33

Previous Topic:	Re: Plot multiple axes with log and linear scales
Next Topic:	3D UserSym?

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sun Oct 12 15:13:29 PDT 2025

Total time taken to generate the page: 1.13747 seconds