Most appropriate sub-area of Processing 4?

PDE

Processing version

4.5.6

Operating system

Windows

Bug description

When size() is called from setup() using variables for its width and height, Processing reports the error only when the sketch is run/compiled:

size() cannot be used here, see https://processing.org/reference/size_.html

However, the PDE does not report this as a logic error while editing the sketch.

This is inconsistent with other invalid size() usage and makes the problem harder to identify before running the sketch.

For comparison, using the same variables from settings() works correctly, and arithmetic expressions such as size(400 + 400, 900 - 300) are accepted.

Expected behavior

The PDE should identify the invalid size() usage as a logic error while editing, before the sketch is run.

Actual behavior

The invalid usage is not reported as a PDE logic error while editing. The error is encountered when the sketch is processed/run.

Steps to reproduce this

1. Open Processing 4.
2. Create a new Java Mode sketch.
3. Add the following code:
4. Wait for the PDE to process the sketch.
5. Observe that the invalid size() usage is not reported as a logic error in the editor.
6. Run the sketch.
7. Processing reports:

size() cannot be used here, see https://processing.org/reference/size_.html

For comparison, moving the size(width, height) call into settings() allows the sketch to run successfully.

snippet

int width = 800;
int height = 600;

void setup() {
  size(width, height);
}

Additional context

The issue appears to originate in the Java Mode preprocessing of size() arguments.

I investigated the existing sizeParamValid() handling and implemented a change that reports invalid size() arguments as a logic error during preprocessing. I also added regression tests for invalid method and variable arguments.

I manually verified the following cases:

• size(width, height) in setup() → logic error
• size(width, height) in settings() → works
• size(400 + 400, 900 - 300) → works

The Java test suite passes with 194 tests and 0 failures on my local setup.

I would be interested in working on this issue and submitting the corresponding fix.

Would you like to work on the issue?

Yes, I’d like to help with this