develooper Front page | perl.perl5.porters | Postings from June 2008

Re: Fatal keeps its ugly messages, autodie smells like roses.

From:
Aristotle Pagaltzis
Date:
June 4, 2008 07:50
Subject:
Re: Fatal keeps its ugly messages, autodie smells like roses.
Message ID:
20080604145026.GW22427@klangraum.plasmasturm.org
* Paul Fenwick <pjf@perltraining.com.au> [2008-06-04 14:35]:
> My main motivation for improving messages in Fatal is that
> whenever I mention it, there's huge enthusiasm from all the
> people who hate Fatal's errors, but who like the idea. Because
> autodie sounds different, there's an immediate assumption that
> it must require more work to use/learn, even though one can
> essentially s/Fatal/autodie/;

Then let’s educate them. :-)  If those are the people we want to
reach, it would be a good idea to market autodie as “a drop-in
replacement for Fatal with lots of extra goodies packed in.”

> With autodie […] you're interrogating the information it
> exposes through methods. […] With Fatal, we care hugely if the
> strings change significantly, even if it's our own code that's
> doing it.

Well, “hugely” is probably an exaggeration – as I said right when
I brought the up the backcompat issue, it’s not a very strong
argument, and you gave reasons for why changing the messages may
not even affect code at all in a range of cases.

The way I see this is rather that we simply don’t stand to gain
enough to make up for even the small risk of modest breakage we’d
take, considering that the same gains are available on an opt-in
basis that carries no risk of breakage at all.

Which is essentially the same as your conclusion, if somewhat
differently emphasised.

Regards,
-- 
Aristotle Pagaltzis // <http://plasmasturm.org/>



nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About