This seems to me to be a significant decrease in the environment's or
sandbox's ability to control what happens in error cases. In my own
project, I am not even sure I can handle this safely; if the environment
can't exert any control over this feature, it might be impossible for me
to support 5.4 at all, which would be a great shame.
Can you give a more concrete example of how that would be a problem?
Would it be better to keep the original error after running the closing
methods?
(We have no strong reasons to raise the last error instead of the first
one; this is just how "most" languages (at least Java :-) do.)
-- Roberto
