[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Scheme-reports] Exception handling

Vincent Manis scripsit:

> I would like to know what the metric is for `this is too much to
> ask'. As presented in the Draft, exception handling is purely
> ornamental;

That is, handling of error-objects, whether created by calling `error`
or by some implementation-dependent means, is ornamental, because you
can't detect them.  I agree that that's bad.

However, a portable program can raise its own condition objects and
detect and dissect them, and an implementation can do the same using
implementation-dependent procedures.

> there is essentially nothing that a compliant program can do in
> reporting an error except write out `SOMETHING IS WRONG' (it's agreed
> that a program will know that an error or exception occurred; here I
> am concerned about reporting). I attempted to repair this by proposing
> a type predicate, [...] and two accessors.

There are really only two bars to pass: a WG member has to file a ticket
(which I have done) and the WG has to vote the proposal in.  WG members
may or may not provide explanations of why they vote the way they do; if
so, these are published with the ballots.

John Cowan              cowan@x          http://www.ccil.org/~cowan
Historians aren't constantly confronted with people who carry on
self-confidently about the rule against adultery in the sixth amendment to
the Declamation of Independence, as written by Benjamin Hamilton. Computer
scientists aren't always having to correct people who make bold assertions
about the value of Objectivist Programming, as examplified in the HCNL
entities stored in Relaxational Databases.  --Mark Liberman

Scheme-reports mailing list