Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Already on GitHub? Sign in to your account

Bump base upper bound #36

Closed
singpolyma opened this Issue Jan 12, 2013 · 6 comments

Comments

Projects
None yet
3 participants

Include support for base-4.7.0.0

Member

hsenag commented Jan 12, 2013

Are there any code changes required?
I'm not all that keen on promising support for base 4.7 in a hackage release until it's stable, though I realise it's also inconvenient for people working on the bleeding edge. Would having a branch that tracks 4.7 be enough?

No code changes required (at least as it stands right now).

A branch would be nicer than the current situation (where I have to hack the *.cabal file to install it), but means I can't use cabal install HTTP, which is sad.

Member

hsenag commented Feb 18, 2013

There's a suggestion of "double bumping" package versions: http://www.haskell.org/pipermail/libraries/2013-February/019409.html

e.g. with base 4.6 as the current stable, 4.7 would only be used for development (and never released on hackage), and 4.8 would be the next stable release.

With that scheme I'd be happy to bump to 4.7 eagerly and re-upload to hackage, and then bump to 4.8 when the API is promised to be stable.

bgamari commented Aug 23, 2013

@hsenag it seems that link is broken. Would you mind turning up a new one?

Member

hsenag commented Aug 23, 2013

The mailing list archives got renumbered recently - there's been some discussion of trying to retrieve the original numbers so maybe the link above will start working again.

In the meantime, I think this is the precise post I linked to: http://www.haskell.org/pipermail/libraries/2013-February/019140.html

This was referenced Sep 1, 2013

Member

hsenag commented May 5, 2014

Base dependency is now bumped, though the general principle of dealing with development dependency bumps is still an issue.

@hsenag hsenag closed this May 5, 2014

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment