Front page | perl.perl5.porters |
Postings from August 2010
On Mon, Aug 30, 2010 at 01:56:18PM +0100, Nicholas Clark wrote: > On Mon, Aug 30, 2010 at 01:36:50PM +0100, Tim Bunce wrote: > > On Mon, Aug 30, 2010 at 06:30:47AM +0100, Nicholas Clark wrote: > > > > Perl is kind. Perl has autovivification and uninitialised value warnings. > > > C has segfaults and undefined behaviour. > > > > > > They are not the same. > > > > Obviously true. However, I believe his key point is a philosophical one. > > I simply don't think that it is applicable to C. I'm not refering to the language. I'm refering to the philosophy of Perl as a tool that avoids putting obstacles in the way of "getting your job done". > > I think it's worth *exploring* ways to avoid *inadvertent* use of such > > modules and dependencies, rather than trying to use a shotgun to kill them. > > I don't think that it's worth it. I disagree. I agree with you on the risks of allowing unrestricted access, but I believe that hard restrictions will limit innovation and/or create worse abuse and problems. I also believe there's a middle ground that can work well for both perl maintainers and module developers. > And you're already using loaded terms. Yeah, sorry about that. > Other people are welcome to have this discussion, but don't expect me to play > a part in it. > Whether a less cautious approach would have worked, I don't know. > But I know that my job would have been easier if I could have assumed that > "private" meant private. I agree entirely. I want to see more things be private to allow greater innovationion within the core. > It's clear that we no longer have people with the same level of technical > skill prepared to make maint releases, and this doesn't look like it will > change in the future. So, I infer that we *have* to make maint job easier. I agree entirely. You seem to be arguing from a viewpoint that assumes blame will fall on perl developers when dependency chains fail after a maint upgrade. I'm arguing that there's a way to mitigate that. (And that if we don't we'll still have dependency chains fail after a maint upgrades because developers will copy-n-paste code from the core instead.) Tim.Thread Previous | Thread Next