[Haskell-cafe] Re: Exception handling in numeric computations

[Robert Greayer]

Henning Thielemann <lemming at henning-thielemann.de> wrote:
> The usual example against clear separation of exceptions and errors is the web server which catches 'error's in order to keep running. 
> However, the web server starts its parts as threads, and whenever one thread runs into an 'error', it is terminated, just like an external shell
> program, that terminates with a segmentation fault. So, yes an error might be turned into an exception, but these are rare cases. In 
> general it is hard or impossible to correctly clean up after an error, because the error occured due to something that you as programmer
> didn't respect. The "error handler" could well make things worse by freeing memory that is already deallocated and so on.

I don't see that as an argument against 'clear separation', really.  Having _some_ way of dealing an error (from within a program), in special circumstances doesn't preclude clearly separating how it's done from exception handling.  I always find it jarring when an HUnit test I've run tells me it encountered an 'exception', when I'm testing pure code (nevertheless I'd also find it annoying if the entire test run terminated because of a failed pattern match).

With respect to the last point - isn't proving that a given program can't corrupt its own RTS possible, even in the presence of errors?