Fabrizio Montesi

Sounds good to me! :-) Please make sure that the java8 branch is also updated with this if you decide to work on master. Otherwise, you can just work on...

Was this merged or am I remembering incorrectly?

I agree it's a problem. How about a rethrow primitive or some token that distinguishes this from the normal throw? throw( default ) looks like it would throw a fault...

How about the following? - `rethrow` : it rethrows the current fault with the same attached data - `rethrow(data)` : it rethrows the current fault with the new attached `data`

The behaviour will be dependent on the protocol configuration anyway, because it will access a variable that is set by the protocol. However, @mawolf's proposal also changes the interface, which...

Yes, that's also my belief, we should move the problem to the architecture. Also, in general, I think it's far more interesting to make the language powerful enough to handle...

PS: we should also have a notion of type aliasing, like in other languages, such that we can simply get surfaces directly in Jolie without having to use external tools....

Not a bug: it is valid to write a variable path on multiple lines. So you've basically written .b = .b.c = "yes", which is a valid statement. We could...

Just a small clarification: it's incorrect to call it a bug simply because the language works as intended, i.e., variable paths can be expressed over multiple lines, just like in...

I agree, see my last comment. It's just that solving it in a consistent way is nontrivial, see the possible solutions. Another possible solution is to use another symbol than...