* Steffen Schwigon <ss5@renormalist.net> [2010-11-25 16:50]: > Anyway, in this particular case, the “deprecated for 10 years” > might be one of those few rare exceptions... I fear them > anyway... I think the new deprecation policy is sane. Putting in the warning is a good way to shake lose the code that’s relying on the behaviour, without actually breaking it. If there is a lot of backlash, it can be undeprecated and we know it has to stay. If there is very little, it can be removed without breaking much code. Some way forward is necessary in any case. (In distantly related notes, it’s also important to undo new mistakes quickly. Not just because this creates less opportunity for wide use, but more importantly because it creates less opportunity for wide use in *unmaintained* code. Recently introduced features will be used in recently written code, which is likely to be actively maintained. So tweaking or withdrawing recent features is less of a problem than breakage that affects code written 15 years ago. Perl does not have a good track record in removing mistakes, because of a confluence of factors. But I feel the new deprecation policy and release schedule provide the conditions to enable it, and I’m very glad of that.) Regards, -- Aristotle Pagaltzis // <http://plasmasturm.org/>Thread Previous | Thread Next