On Wed, Sep 26, 2012 at 04:37:49pm +0200, Iustin Pop wrote:
> This will have the following advantages:
> 
> - allow base Ganeti to depend on htools, and hence promote integration
> - simplify the build configurations and requirements (hopefully they
>   will be simpler, due to the reduction in the possible combinations)
> - allow exploration of more Haskell use in the project, by allowing
>   critical components to be implemented in this language; this is not
>   possible, nowadays, due to the "optionality" of htools
> 
> The disadvantage, of course, is the requirement for GHC and said
> libraries. To compensate, we aim to ensure that our dependency list is
> sane for existing stable/long term distributions.
> 
> We'll have more precise details on the new requirements as we near the
> release (not soon, by the way).
> 
> Feedback welcome, as always.
> 

Dear Iustin,

The plan sounds well thought-out.
We've been trying to become more capable in Haskell, in an effort to
contribute to the codebase in the future.

We're fine with the Haskell dependencies, as long as they are available
for current Debian releases. It seems there are a few problems trying to
get everything right when running Squeeze. Can you comment on whether
things be better in Wheezy wrt the Haskell dependencies of Ganeti?

Finally, it would be very helpful if there was a single point detailing
all current dependencies, Haskell-related or otherwise. Do you plan on
keeping an updated list somewhere, as new features get added to master
and perhaps new dependencies are introduced?

Thank you,
Vangelis.

Attachment: signature.asc
Description: Digital signature

Reply via email to