Re: No exceptions?

From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Thu, 29 Jun 2006 20:20:16 GMT
Message-ID: <4CWog.3832$pu3.91040@ursa-nb00s0.nbnet.nb.ca>

paul c wrote:

> Bob Badour wrote:
>

>> paul c wrote:
>>
>>> Bob Badour wrote:
>>> ...
>>>
>>>> The way to avoid exceptions is to treat them as compile-time errors.
>>>
>>>
>>> I take it you really mean "ONE way to avoid exceptions ...".  For one 
>>> thing, I am interested in being able to express "x join y" in advance 
>>> of defining a header for "x" (and I would like to evaluate it as well 
>>> if that is logically possible!).
>>
>>
>> In what context? I am sure once you specify that, you will find the 
>> appropriate context for the compile-time error too.

>
> It may sound silly, but the context I have in mind is the most minimal
> of languages possible, one that strips away as much context as possible.
> One reason it may sound silly is that I don't want to insist that
> even something as "obvious" as a spelling error should necessarily abort
> evaluation.
>
> (While I am looking for a logical argument for or against my application
> of the CWA, my motivation, if that's what you mean by context, is, I
> think, outside the logical realm so it isn't silly AFAIAC.)
>
> Even if somebody can offer support for handling the syntax and semantic
> exceptions in the way I suggested, I can see that avoiding what I think
> of as operational exceptions, like the old fortran compiler behaviour of
> issuing different diagnostics depending on how much memory it was
> allocated (or sometimes, no diagnostics if it was given enough memory)
> can probably not be handled in the same way.

In the context of 'nothing is ever an error', then nothing is ever an error. Received on Thu Jun 29 2006 - 15:20:16 CDT