I am not meant to be harsh, but note that it fails on a small number of test cases, on v0.21 that is ~900 out of ~50k. Strictly speaking it cannot be described as standard-comforming unless there is some reason behind every failed test. A better way to strive on standard conformance, like QuickJS takes, is to pin down the ecma262 revision and make it 100% compliant.

▲

nekevss 12 hours ago | parent | next [-]

That may be fair. The better wording may even be specification compliant, but none of the Boa maintainers made this post (we found out about it after the fact).

We, the maintainers, could probably sit down some day and try to triage the failures into a documented txt file, but the overhead on that might make it hard to maintain for an open source project. Plus a large amount of the failures are just new specification features that nobody has implemented yet.

	▲	qbane 4 hours ago \| parent [-]
		I really appreciate the clear and detailed project description, the only concern is the title on HN. Failing on tests is fine, especially when it demands huge efforts to triage and reiterate on the project structure for them.

▲

bakkoting 9 hours ago | parent | prev [-]

I don't think there's literally any conforming implementations of modern ECMAScript by that definition.

	▲	qbane 5 hours ago \| parent [-]
		If you mean "latest" ECMAScript then that is true. Even latest gcc or clang does not support all features from C++23: https://en.cppreference.com/w/cpp/compiler_support.html.