Joseph Brill

Bear in mind that I don't know what I'm talking about. The runs that pass seem to consider node `build\\debug\\lib\\core\\static\\core` before any individual file nodes (e.g., `build\\debug\\lib\\core\\static\\corF.cpp`). What does the...

Just a WAG and likely misinformation, but it seems the difference in evaluation order of the DAG could be responsible for the difference in behavior. Maybe?

> What's head-scratching there is the line which should end up building the pch comes out ordered a bit differently, though it appears to have identical contents. The lines to...

Stupid question time. In the abstract, it appears that there is a difference when processing the DAG by the taskmaster whether or not the static library node is considered first...

@mwichmann When you have time, would you please review to see if the following makes any sense. *My comment about the commands in your runs is not quite true. In...

Based on the generated argument order and pch switch difference for the pch source file, I'm starting to think that PCHCOM is not called/applied when the pch source file is...

Old-school print-style debugging. The builds that pass generate a command to build the `pch` and `obj` file. The builds that fail generate a command to build the `obj` file. **PASS**...

Edit: * change `FILENAME.pch` to `FILENAME.cpp` in 3rd paragraph * change `a-pch.cpp` to `a-pch.pch` in annotation for first listing. In the builds that fail, the taskmaster is not evaluating the...

Excerpts from output (listings above without annotations): [scratch-fail.txt](https://github.com/SCons/scons/files/13260071/scratch-fail.txt) [scratch-pass.txt](https://github.com/SCons/scons/files/13260072/scratch-pass.txt)

Updated annotations in listings above. I made a mess of them originally. Apologies.