Timothee Cour

> Well yes, I'm aware of the cross module inlining issue but I hope link-time optimization is good enough these days to make it a non-issue. `--passc:-flto` (i had originally...

At least we now agree we need 2 (at least 2) inlining annotations (whether it's my recommendation (hint = {.inline.}, force = {.forceinline.}) or (hint = {.hot.}, force = {.inline.});...

just hit this bug again; `import nre` messes up toSeq. >> Wouldn't it be possible to solve the issue by avoiding an untyped implementation of toSeq in sequtils and explicitly...

> I think the proposal is to have various overloads of toSeq - one for iterators and the other ones for specific collections. As far as I can tell, this...

right, I moved over to https://github.com/AndreiRegiani/INim

IMO this is better: https://github.com/prasmussen/chrome-cli/issues/55 allows eg: `chrome-cli --json info | jq -r .title`

could that be due to not-nil strings / seqs and the changes from https://github.com/oswjk/nimpb/pull/10 ?

> to make the IR easier to read, it'd be better not to inline expect though if keeping IR clean is a concern, maybe also optionally making expect `nothrow` to...

question: how to run a specific test? the number of tests failed is misleading, since, for eg, druntime-test-shared comprises of multiple tests (subtests); would be nice to report on these...

excerpts from Testing/Temporary/LastTest.log: ``` ============================== Test failed: expected: ---- fail_compilation/fail10346.d(9): Error: undefined identifier 'T' ---- actual: ---- fail_compilation/fail10346.d(9): Error: undefined identifier 'T' fail_compilation/fail10346.d(13): Error: template fail10346.bar cannot deduce function from...