Wither Haskell Platform 2013.4.0.0

[David Luposchainsky]

I think making the HP release cycle purely time-based is not the best
option. For a large amount of users, the HP and what comes with it
(including the associated compiler) is what "Haskell" is. If something
does not work with the HP, then Haskell does not work.

- Support for the latest OS is crucial, probably most importantly to
avoid bad publicity. "Hard to install" is an image that's not easily
changed when search engines are full of hacks that make it somehow work
again.

- Are there significant changes in the HP libraries compared to the last
release? Is there a good reason to make people update to the latest
packages?

- In the light of the AMP, I think a relatively early adoption of 7.8
would be very beneficial in order to prepare Hackage. Releasing this
year with 7.8 would not be a good idea, but maybe next spring (which
would be halfway between the usual HP release dates) is possible?


In short, my vote is "delay until it works with all OS and 7.8".

David