DIP29-3: function calls

[WalterBright]

This is a large step forward in better implicit conversions.

[yebblies]

Did you forget to #undef LOG at the end? Please just comment these out instead of putting #ifs everywhere.

[WalterBright]

done

[yebblies]

Walter, what exactly are you doing here? You do know that mutable return values of pure functions have converted to immutable for ~3 years now, right? I wouldn't have to ask if you would write a meaningful description for the pull request.

[WalterBright]

You do know that mutable return values of pure functions have converted to immutable for ~3 years now, right?

Try the test cases I supplied with the PR - they fail otherwise. The trouble is that the existing version did not check if the arg expressions were convertible - only that the types were.

[WalterBright]

I wouldn't have to ask if you would write a meaningful description for the pull request.

See http://wiki.dlang.org/DIP29

[yebblies]

See http://wiki.dlang.org/DIP29

So it implements the entire DIP?

[WalterBright]

So it implements the entire DIP?

No, just the function calls. See the previous DIP29 bits I've done, the ptr+offset and array[i] expressions:

#3336
#3319

I'm doing it piecemeal because it's much easier to review it that way - each change is fairly standalone.

[9rnsr]

LGTM.

[9rnsr]

Auto-merge toggled on

[wilzbach]

@WalterBright I know time is very short, but would you be so kind to update the DIP's status?
I just went through the git log to see that it's actually not "Draft", but "Implemented"?

http://wiki.dlang.org/DIP29

(might be a good idea to do a short house keeping and see whether there are other DIPs with changed status out there - thanks :) )