concurrency (was Re: important news: refocusing discussion)

[Tomasz Zielonka]

On Tue, Mar 28, 2006 at 10:49:36AM +0100, Malcolm Wallace wrote:
> Tomasz Zielonka <tomasz.zielonka at gmail.com> wrote:
> >     http://www.uncurry.com/repos/FakeSTM/
> > 
> > Perhaps it could serve as a drop-in replacement for STM in haskell
> > compilers which don't implement STM directly.
> 
> Nice idea.  But your code already uses a whole heap of Haskell
> extensions which may or may not make it into Haskell'.
> 
>    monad transformer lib (requires MPTC)
>    exceptions
>    dynamically extensible exceptions
>    deriving non-standard classes
>    extended newtype deriving
>    pattern guards

You read the whole code? Wow! I myself would have trouble understanding
how it does what it does now ;-)

I could easily get rid of:

    deriving non-standard classes
    extended newtype deriving
    pattern guards

I used GHC's exceptions, because I wanted my STM to handle them
correctly, as in the STM paper. In a implementation without exceptions,
I could probably get away with hand made exception handling.

The rest would be a bit more difficult to remove, but I think it could
be possible more or less elegantly.

Best regards
Tomasz