I'm wondering if there is a workflow that allows developers to share due dilligence of work in a pull request.
To give you some insight, I noticed that if you have a network parameter in your config, like chain=main and then use chain=test in the commandline, you get an error message but It doesn't really tell you what is going on.
This change provides a little bit more verbosity letting the user know where and how the collision is happening.
For example, my change breaks the test util_tests/util_ChainMerge
You can run it as per the instructions
CHAIN_MERGE_TEST_OUT=results.txt {build_dir}/src/test/test_bitcoin --run_test=util_tests/util_ChainMerge
an excerpt for an example
Current:
-regtest=1 -noregtest=1 noregtest=1 notestnet=1 || main
-regtest=1 -noregtest=1 testnet=1 regtest=1 || error: Invalid combination of -regtest, -signet, -testnet and -chain. Can use at most one.
Mine with a more verbose error message:
-regtest=1 -noregtest=1 noregtest=1 notestnet=1 || main
-regtest=1 -noregtest=1 testnet=1 regtest=1 || error: Invalid combination of -regtest, -signet, -testnet and -chain. Can use at most one./n Too many network flags being set in the commandline.
So I would like to show that I have a script that compares these results appropriately and shows that my change doesn't break anything.
What is the best way to package helper scripts up to help reviewers? Should I make a separate repo for reference?