On Sun, Feb 5, 2012 at 6:18 PM, Vesa Kaihlavirta wrote:

So now I'm not so certain anymore about dropping it. We could again wait for a while until HP catches up (it should be relatively fast this time, since 7.4 is supposedly a real stable release) or just drop it and go to 7.4.1 asap (and probably face some breakage). What do you people think?

I'm not going to suggest which action Arch should take; instead, I'm going to say what I am going to do. GHC 7.4 is still to new for my production systems, and I can't develop on a different version. So, if Arch moves to 7.4 right now, I'll need to download GHC 7.0 separately and use virthualenv. Cheers! -- Felipe.

On Sun, Feb 5, 2012 at 21:18, Vesa Kaihlavirta wrote:

On Sat, Feb 4, 2012 at 7:08 PM, Magnus Therning wrote:

...

This was mainly directed to the Arch Devs on the list, and especially Vesa since he's putting in most of the work on GHC as found in [extra].

We've already discussed the question of whether we should keep HP in Arch.  *I* think there was clear what the favoured route ahead was, but in the end it's the Arch Devs who decide.  I don't like the current situation (outdated GHC and outdated HP), so a decision would be much welcome then the work on catching up can begin.

Yeah, I noticed 7.4.1's release just now.

The question of HP and Arch was left in a limbo the last time, I feel. The reason I started pondering about dropping HP earlier was that there seemed to be no progress on it with regards to 7.2. Then it was pointed out that 7.2 was sort of an unstable upstream release, so that was the reason for HP not going with it.

So now I'm not so certain anymore about dropping it. We could again wait for a while until HP catches up (it should be relatively fast this time, since 7.4 is supposedly a real stable release) or just drop it and go to 7.4.1 asap (and probably face some breakage). What do you people think?

Drop HP. The main reason is that the goals of HP are opposite the goals of ArchLinux. /M -- Magnus Therning                      OpenPGP: 0xAB4DFBA4 email: magnus@therning.org   jabber: magnus@therning.org twitter: magthe               http://therning.org/magnus

On Sun, Feb 05, 2012 at 10:18:32PM +0200, Vesa Kaihlavirta wrote:

On Sat, Feb 4, 2012 at 7:08 PM, Magnus Therning wrote:

...

This was mainly directed to the Arch Devs on the list, and especially Vesa since he's putting in most of the work on GHC as found in [extra].

We've already discussed the question of whether we should keep HP in Arch.  *I* think there was clear what the favoured route ahead was, but in the end it's the Arch Devs who decide.  I don't like the current situation (outdated GHC and outdated HP), so a decision would be much welcome then the work on catching up can begin.

Yeah, I noticed 7.4.1's release just now.

The question of HP and Arch was left in a limbo the last time, I feel. The reason I started pondering about dropping HP earlier was that there seemed to be no progress on it with regards to 7.2. Then it was pointed out that 7.2 was sort of an unstable upstream release, so that was the reason for HP not going with it.

Here's how I think: 1. The decision to not go with 7.2 was not bad. I think upstream caused a bit of confusion by giving a stable-release-number to a tech preview. 2. I think the correct path forward is to drop HP. HP is a set of libraries, and it blesses certain versions of those libraries. Each of those libraries is however released independently as well so sticking with HP feels counter to how the Arch Way is practised[1]. 3. My impression was that there was a majority of people on the ArchHaskell mailing list in favour of dropping HP.

So now I'm not so certain anymore about dropping it. We could again wait for a while until HP catches up (it should be relatively fast this time, since 7.4 is supposedly a real stable release) or just drop it and go to 7.4.1 asap (and probably face some breakage). What do you people think?

There's still a bit of work needed to keep up with HP since there's been a new release with GHC 7.0.4. /M [1]: I write it this way since I can't find anything in the Arch Way saying that it's a goal to always stay with the latest release of packages. /M -- Magnus Therning OpenPGP: 0xAB4DFBA4 email: magnus@therning.org jabber: magnus@therning.org twitter: magthe http://therning.org/magnus Perl is another example of filling a tiny, short-term need, and then being a real problem in the longer term. -- Alan Kay

I'll emerge from my lurk-cave to say: Agreed. Trying to stick to the HP is more often than not a hinderence: what would be really great is the latest "core" Haskell packages -- cabal, ghc, base libraries etc. -- be available in extra/whatever repo, but that's about all that I could reasonably ask for. Trying to maintain a huge number of binary packages from Hackage, when they all update differently, depend on different packages which themselves might update unpredictably etc. so far yields mostly-out-of-date Haskell packages. - Mike. On Thu, Feb 23, 2012 at 2:04 PM, Matthew Farkas-Dyck wrote:

Drop HP. If I want a certain package, I'll get it myself.

Cheers, MFD

_______________________________________________ arch-haskell mailing list arch-haskell@haskell.org http://www.haskell.org/mailman/listinfo/arch-haskell

Hi all, after some time trying to maintain a repository with ghc-7.2 I fell a little frustrated about the Haskell community in general: ghc developers release new version as "stable", (7.2 and 7.4) but recommend to use Haskell Platform. For this reason many packages work with 7.2 (so I suppose they will work with 7.4 as well) and many others need the older version. I've already voted for 7.2 in the past, and now I would like to see 7.4 in arch as soon as possible, but many packages will not compile with ghc-7.4 (leksah...). On the other hand Haskell Platform is too old for many new package (for example yesod or snap). Maybe we should have a ghc-7.4.1 package and a ghc-7.0.4 (in Arch we have for example python(3) and python2, ruby1.9 and ruby1.8). This would imply having two version of some other package and could be very complicated to maintain. If there are enough maintainers (4-5) IMHO we should try this. Otherwise I vote for ghc-7.4 and to drop packages that fails with it. Fabio 2012/2/5 Leif Warner

I vote for 7.4.1. Things are moving beyond 7.0.4. HP is a nice idea, but I've never really used it on Arch - I just install the Haskell packages I need and their dependencies are automatically installed. Here, I view it more of a "blessed baseline" for the larger goal of supporting as many useful Haskell packages together as possible.

-Leif

On Fri, Feb 3, 2012 at 7:28 AM, Magnus Therning wrote:

...

It's been out for a day now... this would be a good time to decide whether ArchLinux should be bold and move to a haskell-platform-free state, or trudge on with HP and the ache it causes.

If we're moving to 7.4.1 there's a lot of work with getting ArchHaskell in shape for it, so the longer notice the better :)

If we're to trudge on with HP then we really ought to move to 2011.4.0.0, which also means a new version of GHC and a lot of work on ArchHaskell, so the longer notice the better :)

Please, please, please, can we move on this soon?

I'm available to help out with [extra]&[community] packages if need be, updating PKGBUILDs, building, verifying other's changes, etc.

/M

-- Magnus Therning OpenPGP: 0xAB4DFBA4 email: magnus@therning.org jabber: magnus@therning.org twitter: magthe http://therning.org/magnus

_______________________________________________ arch-haskell mailing list arch-haskell@haskell.org http://www.haskell.org/mailman/listinfo/arch-haskell

_______________________________________________ arch-haskell mailing list arch-haskell@haskell.org http://www.haskell.org/mailman/listinfo/arch-haskell