Re: crash_message
Gabriel Boehme wrote:
>Interesting. Nobody seems to want to respond to the points I raised earlier.
>Well, I suppose it's a lot easier to argue nebulous conceptual ideas than it
>is to actually come up with specific, detailed solutions about how this idea
>could work within the language.
Well you did make a good point Gabriel; the possible need
of error codes had come across my mind as well. But since
I was inclined to investigate all other options before
supporting the crash-routine concept to begin with, I was
waiting to see how others countered your point.
As far as error codes go, I think it depends entirely on how
and why each of us would prefer to see a crash routine
implemented. I have no preference as of yet, although I can
see the addition of error codes complicating the matter
further.
Concerning re-entering the code, I think it makes more sense
to just execute the routine, then exit the interpreter. This
prevents compounding of errors (especially since it'll be
literally impossible to tell where the error occured), and
lets the programmer either recover from the error, or abort
as he chooses.
To Ralf: yes, like I said, I could tolerate a crash-routine
myself, and would even use it. My point though, is that I
think we should investigate every option before proposing it.
I know I *personally* can't think of any other way to do
what needs to be done, but considering others on the list
have experience in languages and procedures that I do not,
someone might present an alternative.
To David & Ralf: I suppose the simplicity, clarity, and power
of Euphoria is enough to attract people to it. But I still
think what makes it stand out even more than that is it's
rigid, prohibitive design and double-checking, interpretive
implementation. Again, I'm not saying we don't need some way
to take care of errors in Euphoria, I'm just trying to stress
the need to explore all possibilities before opting for a
quasi-Euphorian mechanism like crash_routine (just like I'm
sure other possibilities were explored before the adoption of
routine_id.)
Rod
|
Not Categorized, Please Help
|
|
