Release 0.18?

[aloucks]

Would it be possible to see a new release sometime soon? I'm trying to trim down dependencies and the rand crate was made optional about a year ago. It would be nice to exclude that when it's not needed.

[kvark]

Looks like cgmath needs a maintainer. I've not been doing any math stuff for ~ year, so I don't feed incentivized invest time in it right now. Would you want to help? We'd need:

1. a pass through the issues to see which of them need to be addressed for 0.18
2. a pass through the current dependencies, considering removal or bumps
3. a pass through the API to see which parts need to be updated

[aloucks]

Would you want to help?

Yes.

All of the dependencies on master are now up-to-date. However, I'd like to see a new release of mint that includes kvark/mint#50 (and we'd update cgmath Quaternion to match).

The simd crate is questionable at this point, but it's an optional nightly-only feature. I'd like to see this updated but I'm not sure it's something I'd be comfortable trying to take on right now.

I did a quick scan of the issues and saw a few that I might be good candidates for fixing. I'll take a closer look this evening or later in the week.

[kvark]

See/review kvark/mint#53 please

[kvark]

@aloucks mint-0.5.5 is up

[aloucks]

@kvark I scanned through the issues and these are the ones I'm thinking are good candidates.

These issues are all related to the "handedness" of various functions. I'd like to resolve and clarify them once and for all by adding _rh and _lh variants to all things with "handedness":

#448
#390
#350

Matrix3::look_at_rh
Matrix3::look_at_lh
Matrix4::look_at_rh
Matrix4::look_at_lh
Decomposed::look_at_rh
Decomposed::look_at_lh
perspective_rh
perspective_lh

Nice to have:

Non-uniform scale for Decomposed:
#434

Low hanging fruit:

Slerp result needs normalized:
#498

Matrix3::from_translation:
#484

normalize can produce vector of NaN (add a note in the documentation that the result needs normalized if the magnitude > 0):
#450

Run rustfmt on project:
#415

Doc formatting (is this still an issue?)
#373

[kvark]

Handedness proposal looks good to me.
#434 is something we perhaps not need at the moment.

[kvark]

Actually, I think it could be best to have enum Handedness, since this would make up for less entry points, and internally cgmath would do something like that anyway.

Also, there may need to be a enumeration of depth for perspective generation:

enum Depth {
  Signed, // OpenGL style, -1 to 1
  Unsigned, // D3D style, 0 to 1
  UnsignedReverse, // reverse-depth, 0 to 1
}

[kvark]

Looks like #501 should also be addressed before we release

[nstoddard]

Would it be possible to release 0.18 fairly soon, with relatively minimal fixes for issues like #501, putting off larger changes (like the proposal to eliminate Point entirely) until 0.19? I'd like to stop depending on the git version of cgmath, but some of my code depends on new features.

I might be able to help if there's just a few things that need to be done before release.

[kvark]

Yes, I think it's a good idea! So we are only blocked on #501 atm

[aloucks]

I've submitted #508 for feedback on look_at_[lh|rh]. I don't think a Handedness enum is the right path forward.

[hayashi-stl]

Non-uniform scale on Decomposed would be tricky because it would be possible to concat two Decomposeds into a transform that can't be represented losslessly by a Decomposed.

[kvark]

@josh65536 can you explain this, please? My understanding is - if you have any sequence of uniform scales, translations, and rotations, you aren't squeezing the object improportionally, so there is an equivalent Decomposed for it, losslessly.