[Gnu-arch-users] Re: Nit

[Tom Lord]

    > From: Miles Bader <address@hidden>

    > Dustin Sallings <address@hidden> writes:
    > > You're dismissing an entire methodology because you think
    > > you've found a small hole that might cause an exception to be handled
    > > in the wrong place.

    > I think it's not a `small hole', it's a rather large one.

Thank you.


    > >         You appear to be an exceptional developer in your areas of
    > > expertise, but you don't seem to have the experience with writing
    > > reliable java applications that some of the ``java-fans'' have.

    > Java's exception model is hardly unique, it's exemplary of a very common
    > exception model.  So one doesn't need to be a java expert to say
    > something useful about this issue; java is just an example.

What is most unique about Java is the interaction of the flawed
exception model with its safety model and features for execution of
"untrusted" code.

What is somewhat unique about Java is the contexts (economic and
computational) in which it is deployed.

Big hole with big stakes.

-t