Front page | perl.perl5.porters |
Postings from July 2010
On Fri, Jul 2, 2010 at 6:46 PM, Dave Mitchell <davem@iabyn.com> wrote: > On Wed, Jun 30, 2010 at 10:26:06AM +0100, Nicholas Clark wrote: >> http://industry.haskell.org/collab >> >> Note, also, that their scheme forces the incoming money to be split between >> maintenance and targeted development, > > I think this is a key point. perl has a long history of new features being > added by keen people, who then (though lack of interest or changed > circumstances etc) don't hang around around to to fix the bugs. For > example (and I don't mean to pick on individuals here, these are just the > first two that came to mind), smart match and MRO. > > And new features inevitably have bugs. Perl has so many interesting > features that interact with other in interesting ways, that it's impossible > to get everything right first time. For example, overloading was added > around 1995, yet I'm only just getting it to play nicely with tie and taint. > So the overall "cost" of a new feature is far higher than just the initial > development. So maybe when quoting a price to the companies we just need to multiply it by 2 or 3 and put the rest of the money in the bank account of TPF so it can be used for later grants to fix those features. I am not sure but I think that's similar to what Nick was referring to in the SQLite Consortium. GaborThread Previous | Thread Next