Re: Ensuring postconditions in the face of exceptions

["Randy Brukardt" <randy@rrsoftware.com>]

"Robert A Duff" <bobduff@shell01.TheWorld.com> wrote in message 
news:wcciq8wxpkv.fsf@shell01.TheWorld.com...
> "Randy Brukardt" <randy@rrsoftware.com> writes:
...
>> Repeat after me: "limited" is a property of a view! It's not constant for 
>> a
>> particular type! That's a common mistake that even those of us on the ARG
>> make from time-to-time.
>
> I don't understand why the definition of "by reference" is subtly
> different from the definition of "immutably limited".

Two obvious reasons: "by-reference" predates "immutably limited" by more 
than 10 years, and besides there are non-limited types where are 
"by-reference" (non-limited tagged types). The latter also means that types 
with a component of a non-limited tagged type are "by-reference" as well. 
Finally (OK, I can't count), "by-reference" breaks privacy (thus only 
dynamic rules ought to depend on it; legality rules should never depend on 
the contents of the private part), while "immutably limited" doesn't break 
privacy and thus can be used in legality rules. "Immutably limited" is a 
direct property; it doesn't depend on type of components.

The net effect is that there isn't much similarity between the two concepts, 
nor their use.

                             Randy.