Le lundi 24 juillet 2006 12:21, vous avez écrit :
> On Mon, Jul 24, 2006 at 11:56:18AM +0200, Erich Focht wrote:
> > Hmmm, your WARNING looks for me like an unsafe GO decision. But it is a
> > GO decision. Only it has attached with it a warning message. So you are
> > well within the GO / NO GO pattern.
> >
> > Actually a WARNING is only necessary because the check is not precise
> > enough about the state of the system. If you'd have precise information
> > about the system, you could make a clean decision and _know_ whether this
> > is a GO or a NO GO. Without the warning message. (And I understand that
> > such cases with imprecise checks can exist).
>
> My feeling is that a WARNING should, by default, imply NO GO. The user
> should then have the option to repeat the attempt with a parameter
> implying: "I have looked into that (specific) WARNING and have decided
> to proceed anyway."

If you do that in the example i gave, just forget any check: it is impossible 
to be sure at 100% the apt repositories are setup according a specific 
parttern... in that case, if we do what you said, it may never work even if 
the setup is correct (it is just impossible to check it). 
If a warning implies a stop, the warning is then equivalent to an error, there 
is different between them. Therefore doing that we do not need a Warning 
statement, just GO and NOT-GO... and back to the initial question.

So i tend to agree with Erich a warning is more "message + GO".

Regards,
-- 
Geoffroy

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Oscar-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/oscar-devel

Reply via email to