Re: [RFC] [VOTE] Immutable/final/readonly properties

Nicolas Grekas <nicolas.grekas+php@gmail.com> wrote: (from the other thread) > there must be a way to work around the keyword - > either via reflection or another means. Can you expand on why there 'must' be a way to work around this? Can you provide some example code where not being able to change the value is going to cause problems. > I voted against the RFC because it has too many rough edges to me. > eg cloning doesn't work, the initialization needs special rules, etc. To be clear, cloning would work....but just not be useful. But being required to write a function to copy an object, and overriding the values doesn't seem like a terrible burden. And, because PHP has an evolved type system, it is always going to have some 'special rules' for certain things, just because of maintaining BC. > I don't think these issues can nor should be figured out later on: they are > low-level conceptual issues IMHO. I can't see where the edge-cases are going to bite us. What's an example of where they can't be figured out later? btw I totally agree that having them now would be nicer... cheers Dan Ack