On 4/25/07, Xavier Hanin <[EMAIL PROTECTED]> wrote:
On 4/25/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> Sorry - just few cents from the dont-know-ivy-yet man ...
Don't be sorry, your input is very valuable!
>
> >So I see 3 options:
> >- rename to haltonfailure and keep code as is
> >- rename to haltonfailure and actually halt on any failure
> >- add haltonfailure which halts on any failure, and keep haltonunresolved
>
> What about having an enumeration here?
Indeed, I often forget about using enumerations instead of boolean
(Stephane Baillez already complained about that, I should learn!).
So this is a fourth option which would IMO at least be a better option
than the third one.
So now maybe the question is more: do we want to have the halt on
unresolved feature (and implement it using option 4), or only halt on
failure (option 2)?
IMO the answer depends on use case justification by Jeffrey.
I'm not sure Jeffrey is following this list, so I've just pinged him
directly to get his opinion.
Xavier
