Re: C<strict> as default (was Re: RFC 16 (v1))
[email protected] (Michael Stevens) Sun, 6 Aug 2000 19:57:15 +0100
| Newsgroups | perl.perl6.language.strict |
|---|---|
| Message-ID | <[email protected]> |
On Sun, Aug 06, 2000 at 02:27:00PM -0400, [email protected] wrote: > > Because perl is for the lazy. I want an easy way to arrange for all my > > programs to run under strict/warnings, should that be my inclination. > > It already has a run-time option; You can set the environment variable > PERL5OPTS to contain '-Mstrict -W'. Of course, you've got to fiddle enough to ensure this variable is set in all possible contexts where you're going to be running perl. I'd imagine a wrapper script will be the Right Thing. > Now, do you, Michael Stevens, actually *have* PERL5OPTS set to contain > '-Mstrict'? Or is this lazy person you're talking about totally > hypothetical? He's a person I can see myself wanting to be in the future. > > Can't we make everyone happy by making it a build-time option? > > No, and I explained at great length yesterday why I think that's a > stupid idea. > But I think it would be extremely valuable research if someone would > actually try living with PERL5OPTS='-Mstrict -W' for a month or two. > That will simulate what things will be like for the people who are > stuck on systems where Perl has had the 'always strict' build-time > option enabled. Then we will be better able to judge what some of the > effects will be, at least for people on the receiving end. > Since you just said that you wanted an easy way to arrange for all > your programs to run under strict and warnings, why don't you carry > out this experiment and let us know how it works out? (Or, if I've > misjudged you and you already *are* carrying out the experiment, why > don't you tell us how it is going?) I wasn't actually trying it, but it sounds like an extremely good idea. As soon as I'm back in a position to try that, I will experiment.