IDE settings

[pjc30943]

How can the IDE be changed to not always jump to the nearest error or warning? I scroll in the file somewhere, switch to a different pane, and when switching back it immediately goes to the first error again. Very annoying

Also, is there any way to make the default debug window taller?

I scanned through the setting files, but couldn't find anything related to these.

[dkinzer]

How can the IDE be changed to not always jump to the nearest error or warning? I scroll in the file somewhere, switch to a different pane, and when switching back it immediately goes to the first error again.

I can't reproduce this by doing what I think you're describing. Can you describe more specifically each step to reproduce the problem?

Also, is there any way to make the default debug window taller?

The height of the output window (of which the debug window is a part) is stored in the User Properties file when you exit. When the IDE is restarted, the output window height is set from the stored value.

The entry in the User Properties file looks like this:

Code: Select all

app.output.height=135

[pjc30943]

Edit: Changing "app.output.height" didn't alter the window size even after exiting; it was set to 0 by default, so I assume the value stored is added to the current default window size (about five lines)? I tried values from 20 to the 135 mentioned.
-------------------------------

Ah, okay, thanks.

As for jumping to the nearest error or warning, here is one sequence:

After compiling and downloading, the IDE highlights and jumps to the line of the first warning.

Say I'm working on another piece of code. I scroll away from this highlighted line to check something. Then I switch back to the header file (another pane), to--for example--check the value of a constant.
When switching back to the first pane mentioned, the cursor is no longer where I left it, but instead the file has scrolled back to the first warning/error again. If the code I want is at the very end of the file, then every time the file must be scrolled through again to find it--this is quite time consuming.

I'm sure there is just a setting to flip this feature off! Obviously, if there are no warnings or errors, no jumping occurs.

[dkinzer]

Changing "app.output.height" didn't alter the window size even after exiting

The app.output.height value is supposed to be updated automatically - no need to manually adjust it. If yours is not getting updated when the IDE exits then something is happening that is different than we anticipated. What OS are you using - Vista perhaps? Also, what means do you use to exit the IDE (e.g. click on the close box, select Exit from the File menu, right click on the task bar icon and select close, etc.).

[pjc30943]

EDIT: Anyone find or know solutions for either of these?

The app.output.height value is supposed to be updated automatically - no need to manually adjust it. If yours is not getting updated when the IDE exits then something is happening that is different than we anticipated. What OS are you using - Vista perhaps? Also, what means do you use to exit the IDE (e.g. click on the close box, select Exit from the File menu, right click on the task bar icon and select close, etc.).

XP.
After exiting using the menu, the window retains its size--until the program is recompiled or loaded again.
The window then defaults back to the small size, and all debugging is through the small windows.
Should the larger (saved) window size be desired, the window has to be closed, then opened again using two presses of F8. It would be best if it remembered after loading, since--presumably--that is when most debugging takes place.

[pjc30943]

Anyone find or know solutions for either of these?

[stevech]

guess: XP logged in user lacks permissions so that the registry can be updated? Most people use admin priv. login so this may be n/a

[pjc30943]

guess: XP logged in user lacks permissions so that the registry can be updated? Most people use admin priv. login so this may be n/a

Thanks for the thought. I'm already admin as you guessed...

The main time-consumer is the focus always being taken back to errors when switching panes; this should be a local IDE setting that wouldn't be affected by the registry, presumably.

[twesthoff]

As for jumping to the nearest error or warning, here is one sequence:

After compiling and downloading, the IDE highlights and jumps to the line of the first warning.

Say I'm working on another piece of code. I scroll away from this highlighted line to check something. Then I switch back to the header file (another pane), to--for example--check the value of a constant.
When switching back to the first pane mentioned, the cursor is no longer where I left it, but instead the file has scrolled back to the first warning/error again. If the code I want is at the very end of the file, then every time the file must be scrolled through again to find it--this is quite time consuming.

I'm sure there is just a setting to flip this feature off! Obviously, if there are no warnings or errors, no jumping occurs.

Has anyone found a solution to this problem? After compiling, I'd like it to leave the cursor where it was in my code. Then if I double-click on the error/warnings take me to that line. I spend a lot of time going back and finding where I was working in the code each time I compile...

[dlh]

I use UltraEdit for all editing and only use the IDE to compile. Not only does that keep the cursor where I was, I can edit from any machine on my network as the compiler checks for changes before compiling.

[pjc30943]

Anyone figure out yet how to turn off jumping to errors after compiling?

[mikep]

Anyone figure out yet how to turn off jumping to errors after compiling?

Yep. Write code that doesn't have syntax errors

[pjc30943]

Anyone figure out yet how to turn off jumping to errors after compiling?

Yep. Write code that doesn't have syntax errors

Thanks Mike

They're actually warnings though: such as "statements __ will never exectute" etc., which clearly I know if I intentionally have a test case with an infinite while loop...

[mikep]

Anyone figure out yet how to turn off jumping to errors after compiling?

I think displaying compile errors is a good thing. If you want to turn off warning messages then you can do that with the --warn compiler option. The one you probably want to add to your PJT file is:

Code: Select all

--warn=no-useless-code

[Don_Kirby]

I think displaying compile errors is a good thing.

I agree Mike, but the issue here isn't displaying the errors, it's that the IDE focuses the cursor on the first error after compiling, relocating it from wherever it was previously in the code.

I think the key here is to keep the output/debug window from getting focus during the compile.

-Don