Re: [Haskell-cafe] Control.Monad provisional?
Roman Cheplyaka <roma <at> ro-che.info>
2013-02-11 09:53:29 GMT
Alfredo is right. Most users don't pay much attention to it, and
developers have preferred to be conservative and mark the modules as
experimental (and provisional at best), because there's no incentive to
commit to keeping the modules stable. And there's no policy about stable
The Haskell Platform is a much more successful experiment in the same
direction, because there actually is an incentive (and clear need) to
keep the packages stable.
"Portability" is another obsolete field. The same information nowadays
can be robustly inferred from the LANGUAGE pragmas in the modules and
the .cabal file.
For my packages, I stopped putting these fields quite some time ago, and
I'm looking forward to their removal.
* Alfredo Di Napoli <alfredo.dinapoli <at> gmail.com> [2013-02-11 08:03:30+0000]
> I might be wrong,
> but the impression I always had is that that field is something that most
> developer struggle to keep in sync. Maybe it
> was provisional ages ago, but they simply forgot to upgrade to "stable" or
> they are simply too humble and think
> "How am I to judge if a package is stable or not"? :)
> My 2 cents :P
> On 10 February 2013 20:38, Petr Pudlák <petr.mvd <at> gmail.com> wrote:
> > Dear Haskellers,
> > Looking at Control.Monad:
> > http://hackage.haskell.org/packages/archive/base/18.104.22.168/doc/html/Control-Monad.html
> > I see: "Stability provisional". I checked some older versions and it's
> > the same. This feel somewhat unsettling - if Control.Monad is provisional,
> > do we have any "stable" packages at all?
> > Best regards,
> > Petr Pudlak
> > _______________________________________________
> > Haskell-Cafe mailing list
> > Haskell-Cafe <at> haskell.org
> > http://www.haskell.org/mailman/listinfo/haskell-cafe
> Haskell-Cafe mailing list
> Haskell-Cafe <at> haskell.org