Add support for magit next

[calve]

This commit adds support for the forthcoming next branch of magit, which should be
merged in master in the near future.
magit-next defined new faces for it's diff hunks that do not play well
with the solarized dark, and this commit intend to fix it.

[sellout]

Sorry for the delay here …

Could you describe a bit more what the problem is? I am not yet using magit next, but I don’t know why the existing inheritance of diff-add, etc. isn’t good enough. Should those fundamental faces be changed in some way, rather than redefining the magit ones? I much prefer package faces inherit from some built-in face when possible.

[Havner]

Any chance this could be merged soon? I've stayed for a long time on the old magit because the theme did not work with the new one. But I need it now and the other solarized color theme out there that supports magit >= 2.0 is rather poor. Doesn't follow color specification and doesn't work in text terminal.

If you're short on time just give me a shout, I'll rebase it and prepare the commit for you.

[sellout]

@Havner Thanks – I can try to get something like this in soon (or you can submit a new PR). I want this to look more like

(magit-diff-added (:inherit diff-added))

though. (Which doesn’t quite line up with the definitions in this PR.)

[Havner]

I had a quick look at this, and those definitions seem not to work properly anymore (there are some missing faces in the diff view). Additionally they are not nearly enough for the current magit version. There are many, many more incorrect faces. I'll try to sit down to it at some point, but can't promise this to be soon.

[Havner]

Ok, I did about half the theme and with my workflow I don't see any non-solarized colors anymore. I'll try to finish this as some point. For now, can you have a look, whether I'm going in the right direction? When I'm done I'll do a PR.

                ;; magit next branch
                (magit-branch-current (,@fg-violet))
                (magit-branch-local (,@fg-blue))
                (magit-branch-remote (,@fg-green))
                (magit-diff-added (:inherit diff-added))
                (magit-diff-added-highlight (:inherit (highlight diff-added)))
                (magit-diff-context (:inherit default))
                (magit-diff-context-highlight (:inherit highlight))
                (magit-diff-file-heading (,@fmt-bold :inherit default))
                (magit-diff-file-heading-highlight (,@fmt-bold :inherit highlight))
                (magit-diff-hunk-heading (,@fg-blue))
                (magit-diff-hunk-heading-highlight (,@fg-blue :inherit highlight))
                (magit-diff-removed (:inherit diff-removed))
                (magit-diff-removed-highlight (:inherit (highlight diff-removed)))
                (magit-diffstat-added (:inherit diff-added))
                (magit-diffstat-removed (:inherit diff-removed))
                (magit-dimmed (:inherit shadow))
                (magit-filename (:inherit default))
                (magit-hash (,@fg-red))
                (magit-head (,@fg-cyan))
                (magit-header-line (:inherit magit-section-title))
                (magit-log-author (,@fg-cyan))
                (magit-log-date (,@fg-magenta))
                (magit-log-graph (,@fg-blue))
                (magit-section-heading (,@fmt-bold :inherit magit-section-title))
                (magit-section-heading-selection (,@fg-blue))
                (magit-section-highlight (:inherit highlight))
                (magit-section-secondary-heading (,@fg-base1))
                (magit-tag (,@fg-yellow))

[sellout]

I think that looks pretty good, at a glance.

[sambrightman]

Is the above comment going to be merged into this PR with a rebase or should a new PR be made?