You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I don't know if any of this has merit. You know the whole storage versus execution trade off.
Is there compile time metadata on the arity or number of methods in the vtable?
If users were given a way to tag types as immutable then the following template constuctor pattern may be worth exploring. If the immutable object type is smaller or equal in size of its pointer then store a copy of the object instead. Also it may be possible to store a object of size 2 * pointer size - 1 byte if the number of virtual methods are less than 256 which will practically always be the case. In any case, the 2 pointer storage type should always be the same type. This logic, or at least part of it, being performed in the concept map instead of the poly.
The text was updated successfully, but these errors were encountered:
I don't know if any of this has merit. You know the whole storage versus execution trade off.
Is there compile time metadata on the arity or number of methods in the vtable?
If users were given a way to tag types as immutable then the following template constuctor pattern may be worth exploring. If the immutable object type is smaller or equal in size of its pointer then store a copy of the object instead. Also it may be possible to store a object of size 2 * pointer size - 1 byte if the number of virtual methods are less than 256 which will practically always be the case. In any case, the 2 pointer storage type should always be the same type. This logic, or at least part of it, being performed in the concept map instead of the poly.
The text was updated successfully, but these errors were encountered: