Re: RFC 70 (v1) Allow exception-based error-reporting.

At 10:33 AM 8/9/00 -0400, Bennett Todd wrote:
>2000-08-08-16:34:16 Uri Guttman:
> > how does this work with RFC 63 (Exception handling syntax proposal - it
> > wants the Error.pm syntax) written by peter scott?
>
>As far as I can tell, my proposal stands independant of RFC 63. I'm
>not proposing that the language add a built-in provision for
>exception-based error reporting; that's clutter and bulk, and there
>are a vocal crowd who adamantly prefer the current style, where you
>have to explicitly test every operation if you don't want errors to
>get ignored.
>
> > you should look at that and work with peter on this.
>
>I looked at it, and I think it's a lovely proposal, but I don't see
>a connection with my proposal here; if RFC 63 were adopted, this
>code would still ignore errors on e.g. disk full:
>
>         print "Hello, world.\n";
>
>And similarly with mine. I'm just saying that it should be practical
>to change that behavior by preceeding with
>
>         use Fatal qw(print);
>
>which currently does not work.
>
> > and we are already discussing that rfc in flow so please followup
> > there only.
>
>Sorry for failing to heed this request, but unless I'm
>misunderstanding RFC 63, my RFC 70 is quite distinct.

Right, they are.  I think my idea of exception classes for builtins is 
orthogonal to both RFC 63 and RFC 70 now I re-read them, so I will submit 
an additional RFC.

--
Peter Scott
Pacific Systems Design Technologies