Parital fix #212 stale gutter, anyTargetChanged->true

[fredsa]

Comment out implementation and have anyTargetChanged always return true.

Holding off on deleting the function
entirely, in case it needs to be brought
back in the immediate future.

[fredsa]

I don't see any downside to always refreshing the views. Perhaps there's performance reason to do this on slower platforms I'm not aware of? Happy to hear your thoughts on what's needed.

[sehugg]

The intent was to keep the emulator from reloading while editing comments and whitespace. The gutter sometimes gets out of sync but only temporarily, are you seeing differently?

[fredsa]

I've been focused on the editor experience, slightly ignoring the emulator, trying to fix issues when for examples moving existing lines of code, here's the clc has been separated from its gutter bytes:

I see now that resetting the emulator is a bit jumpy.

I'll see about updating the gutter but not resetting the emulator when the binary does not change.

[sehugg]

It's not only about resetting the emulator, when you have multiple object files like C or CA65 it takes a relatively long time to do a build, so you only run the full build if an object file changes. Unfortunately the debug info doesn't run until the full build completes. I never worried about it since you'll eventually sync up when you make a proper change.

[fredsa]

Ack. I hadn't realized the behavior for multi file C projects.

How's this updated PR?

For DASM, on source code only change:

• Gutter is refreshed, but emulator is skipped, and console prints:

unchanged, skipping loadROM

For multi-file C, behavior is still the same:

• Stale gutter can happen with certain edits.