Fred Emmott
Fred Emmott
They're now in a pseudomain, but this is still invalid and should raise a type error: ``` + // The key 'z' doesn't exist in $point so our explicit default...
Though I guess that's partially a dupe of #375, the change commit you made for this issue should cause test failures
It should remain compatible (or 'close enough') so that people working on features can contribute documentation examples in a format they're familiar with.
Actually a big chain of blockers here: 1. run tests in parallel 2. ... add a zero-config way to do parallelism in hack 3. decide on an entrypoint mechanism for...
> You can't multiplex queries This is assuming that transactions are used, right?
Also the `Asio\v` recommendation should be replaced with HSL
I'm not sure how to design this in a way that works reasonably well for both all-args-have-same-renderer and different-renderer-per-arg. It might be best to add a `HackBuilderValue` type which includes...
This is probably worth looking at v5, which hopefully won't be as long away as v3 => v4 :)
I don't think we'd consider things like `Map` to be type arguments - except when passed as a generif type - and I'm not really sure what this page is...
we should probably actually write introductions again, even if brief , or replace these with codegen. As they are now, they're a maintainability problem