Address typo in type inference

[Nigel-Ecma]

Fix for issue #787

The PR uses “corresponding parameter” rather than the “ith parameter" from #787. I think this is better, avoids HTML tags, and matches the use of corresponding parameter elsewhere; but either choice is valid.

[jskeet]

I suspect it has to be corresponding parameter, and that we should change the "Ti" part to "the type of the corresponding parameter" or similar.

Otherwise, using named arguments in a way such that the order of the arguments doesn't match the order of the parameters breaks things... unless there's a notional reordering earlier on in the proces, of course.

[BillWagner]

This LGTM

[Nigel-Ecma]

I suspect it has to be corresponding parameter, and that we should change the "Ti" part to "the type of the corresponding parameter" or similar.

Otherwise, using named arguments in a way such that the order of the arguments doesn't match the order of the parameters breaks things... unless there's a notional reordering earlier on in the proces, of course.

You were not meant to notice that ;-) I too realised it hasn't been updated for named arguments but choose just to fix the typo you found. We can always address named arguments later, or do you want to do this now?

[jskeet]

Let's discuss in the meeting - if someone has a suggestion for a quick way to do it properly, let's do that; otherwise, this is better than it was before.

[jskeet]

Merging as better than what we've got.