Topic: Reactor causes fatal error.

[BlackBox]

Being relatively new to TheSwamp, and development in general... It's unfortunate that the positive characteristics which you have attributed to James, characteristics that I know to be included in the reputation that precedes him, were not represented here at all.

Instead James' flaming, baseless accusations, name dropping, and pettiness prevailed; behavior that I expect from an adolescent, and not such a well respected veteran... I look forward to when this is not the case.

[mr_nick]

So, in amongst all this squabbling has anybody actually tested to see if 'their' method, be it considered right or wrong by anybody else, allows the use of the MVSETUP command to behave with a CommandWillStart reactor in AutoCAD 2012/2013?

[Kerry]

mr_nick

Forget about MVSETUP.
Have you tested your reactor with a LINE command and hit ESC part way through. Then close the drawing.

Regards
kdub.

[mr_nick]

Have you tested your reactor with a LINE command and hit ESC part way through. Then close the drawing.

Nothing untoward happens for me, but that's beside the point as I wanted to know if anybody could get MVSETUP to work using their method. I know that the code I posted is incomplete as it was merely meant as a quick example. My 'proper' code did actually have other reactors to handle other events (as has been discussed at length) but MVSETUP, or as it actually turns out command-s, crashes my system regardless of what I try.

[Kerry]

The code you posted crashes AutoCAD 2013 when closing after I run the LINE command.

lets try again.

Can you post the complete code you are using.

[mr_nick]

Odd because my code snippet causes no error with an escaped line command in either 2012 or 2013 (at my end anyway).

As a side note, I have just loaded RenderMans 'foo' reactor code into 2013. Again, no problems escaping out of a line command but if I escape from MVSETUP then I get a fatal error.

[BlackBox]

I believe you mentioned that MVSETUP is a LISP file; open that in VLIDE and place a breakpoint at the top of the defun call. Load the code with the breakpoint, and use F8 to step through the code in order to identify what line exactly is causing your error.

FWIW - I agree with Kerry; you should try posting the complete code you're using after anonymizing it of necessary.

** Edit - my 'foo' reactor does not include CommandCancelled, CommandEnded, or CommandFailed Events or a suitable Callback for same.

[mr_nick]

I already tracked the exact cause and posted this on the very first page - it is a call to command-s which was introduced in 2012. You already said that your machine doesn't crash but you reported that you were using Civil 2011 hence you won't have command-s on your system and thus no fatal error.

And for further reference, the code posted by jbuzbee also crashes AutoCAD 2013 with a fatal error when using command-s.

[BlackBox]

I recalled your posting that, and did reread your posts from the first page. I then tested MVSETUP on Civil 3D 2012, and AutoCAD 2013, both of which result in ultimately fatal errors following several Exceptions on my end.

As an aside, Civil 3D 2012 as AutoCAD (i.e., Map), reported 'unknown command' for MVSETUP, using an OOTB Profile, yet Civil 3D 2012 (as Civil 3D) includes MVSETUP out of the gate.

[mr_nick]

The error is obviously not isolated to MVSETUP - it is anything that uses command-s. From what I've found, that includes MVSETUP, EDGE, AI_UTILS.

Using this simple command will trigger a fatal error when a command reactor is loaded:

(command-s "_.line" "0,0,0" "10,10,0" "")

[TimSpangler]

I remember reading somewhere (not being a master of reactors my self) that you can not use "command" in a reactor?

From AfraLisp

A word of warning! Did you notice how I used ActiveX statements and functions to "Save" and "SaveAs". You cannot use interactive functions from within a reactor as AutoCAD may still be processing a command at the time the event is triggered. Therefore, avoid the use of input-acquisition methods such as getPoint, ensel, and getkword, as well as "selection set" operations and the command function.

*EDIT
Try changing it to vl-cmdf

[mr_nick]

Command is NOT being used within the reactor. Command-s is being used in a drawing which has a reactor running. Doesn't matter what the reactor is actually doing - just it's actual presence is enough to cause the crash.

Hence loading RenderMans or jbuzbees code samples and then using command-s you will crash AutoCAD.

[jbuzbee]

jeez you guy's can't be this stupid.  You CAN'T use ANY form of COMMAND in a reactor.

Stop.

[TimSpangler]

Command is NOT being used within the reactor. Command-s is being used in a drawing which has a reactor running. Doesn't matter what the reactor is actually doing - just it's actual presence is enough to cause the crash.

Hence loading RenderMans or jbuzbees code samples and then using command-s you will crash AutoCAD.

Still the same thing.

[mr_nick]

If that is the case, why does the old version of MVSETUP not crash AutoCAD  when it uses a command call but the new version which uses command-s triggers the crash.