Hi Leslie,

I’ve just noticed your latest push: the implementation of repeated element 
tuples. Nice!
It appears both you and hciR (you know who!) are keen on optimising various 
tuple shapes.

I do wonder how all the tuples specialisations interact at runtime.
For example, is it possible for a TreeTuple to hold non-TreeTuples as childs 
(i.e. reversed, repeated element tuples, etc) ?

I’m especially interested to know whether the endless concatenation and slicing 
of specialised tuples doesn’t incur too much copying, but instead results in a 
lot of sharing between immutable (sub)tuples.
Can we be sure that the (amortized) runtime complexities of combining the 
specialised tuples are not worse than TreeTuple’s runtime complexities?

cheers,
Robbert.

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

Reply via email to