Linear #215
Linear #215
Conversation
…ore merge, plus review)
…double Conflicts: src/Diagrams/Tangent.hs src/Diagrams/ThreeD/Transform.hs src/Diagrams/TwoD/Arrowheads.hs src/Diagrams/TwoD/Combinators.hs src/Diagrams/TwoD/Image.hs src/Diagrams/TwoD/Offset.hs src/Diagrams/TwoD/Transform/ScaleInv.hs
* revert using MeasureX * various R2 -> R2Ish changes
…ish-haskell.yaml -i Although I had to tweak & untweak some files so stylish-haskell / haskell-src-exts can parse them
…t of Types.Double
…ance, so I might as well have the RealFloat constraint as well.
…rams-lib into generalize-double
It's not actually about points, but it's a utility function for vector spaces, like the rest of that module, and it doesn't belong under 2D.
| @@ -113,6 +109,9 @@ Library | |||
| safe >= 0.2 && < 0.4, | |||
| JuicyPixels >= 3.1.5 && < 3.2, | |||
| hashable >= 1.1 && < 1.3, | |||
| linear >= 1.10 && < 2.0, | |||
There was a problem hiding this comment.
In general Diagrams puts upper bounds at the smallest value that could require code changes, assuming the dependency follows the PVP. That would be something like linear >= 1.10 && < 1.11. Does @ekmettt provide some stronger promise of backwards compatibility for linear < 2.0 that should change that?
|
I just saw |
|
This looks really close (including all the other branches). Things I know still need to happen:
Anything else that needs to happen? Can we make a plan to divide up the work of updating
But the manual is probably the most work of those, so we could split it instead. |
|
Sure, I'll start on the gallery. |
|
I'll get working on |
|
Great. I'll work on the manual. |
|
It sounds like you all have pretty much decided to go ahead with this, which is great. For what it's worth I just thought I would mention that though I haven't had a chance to look through everything, from what I have seen I really like it. |
fix change of semantics made during linear migration
|
Is this (finally!) ready to merge? |
|
Think so :D I’d like it if we could hold off releasing about a week. I’m working on some other breaking changes (generalising SizeSpec and a new Measure) in another branch. If they turn out good I’d rather they be included in this realise then have more breaking changes in the next release. |
|
I was just talking about getting this into master. I have no immediate plan (or time) for a release. |
|
LGTM |
|
@cchalmers I think it's only fitting that you merge this PR |
Migrate from `vector-space` package to `linear` Make types more polymorphic.
|
OMG!! =D |
Branched version of #212.