IDE did hang after the exit ("waiting-cursor" forever).
Can't reproduce it anymore. Now it just exits, like you said. Strange ....
Minor "issue": if you start IDE.exe from a CMD prompt, afterwards the window title is set to "ide" and the prompt is not shown. [Enter] fixes that last one.
Minor "issue": if you start IDE.exe from a CMD prompt, afterwards the window title is set to "ide" and the prompt is not shown. [Enter] fixes that last one.
Assuming you're in a console session (not a TCMD tab window), the title change is being done by CMD, not IDE.
The prompt issue is unreproducible here (and cannot possibly be related to IDE, which is a GUI app running in a separate session). When you start IDE.EXE from a CMD prompt, CMD displays a new prompt immediately, because it's not waiting on the GUI app for anything.
I created a "TCApp.BTM" that just echos something and waits for a key before quitting. I tried starting AppDispatch.BTM with an "auto" argument and ran the batch debugger both from a TCC console window and a TCMD tab window. It worked perfectly chaining directly in both cases (and also when using CALL).
You say that chaining is "crashing the debugger" - are you actually getting a crash dialog from Windows? If so, please send the contents of that dialog. (And if you are, it's almost certainly coming either from Windows or from third-party injected code, as IDE does its own exception handling.)
If IDE is crashing in its internal code, it will create a log file named "IDE.exception.log". If you have that file, please post it here. (If you don't have it, the crash isn't in IDE.EXE.)
IDE will load yout TCMD.INI and your plugins and run TCSTART before launching the batch file, so try starting IDE with the /I option to disable TCMD.INI, plugins, and TCSTART.
You are right, the title change *is* done by CMD. When you strart "ise.exe", everything OK; starting "ise" (without the .exe) gives the new ise title right away. To make sure I copied ide.exe to abc.exe and started "abc" : title = abc.
Regarding the prompt: 100% reproducable here :-) . But I don't concider this as something worth looking further into. It might even be some quirk in my system.
Enclosed two screenshots. "Crash" should be understood as an unhandled exception in IDE.exe, followed by an Access violation, followed by a restart of the IDE/debugger.
If you want I could install a screen recorder and mail you the video. However, sending more source files would probably add to the confusion.
Please let me know if you plan to look into this any further and what info would be helpful. Note that I have a workaround for my problem.
I do think there is a vulnerability here, somewhere. If I may suggest something - what might be a rare condition and perhaps therefor untested - is:
process A (deferring something) - chaining to B - chaining to C
When is the deferred action executed? Before B or after C? Just a thought.
The deferred action is executed when the current batch file exits. That means if you chain to another batch file without a CALL, the DEFER will be done before executing the second batch file.
Enclosed two screenshots. "Crash" should be understood as an unhandled exception in IDE.exe, followed by an Access violation, followed by a restart of the IDE/debugger.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.