Re: [eigen] backporting the vectorized quaternion product to 2.0 ?

[ Thread Index | Date Index | More Archives ]

(Gael, feel free to disagree with the very idea of 3.0. Just because
we started discussing that doesn't mean it's too late to oppose it.)

2009/9/22 Rohit Garg <rpg.314@xxxxxxxxx>:
> 3.0 sounds better. Apart from API, what else do you plan to put in 3.0?

The "ideal world" answer:

If we can just get the API right, that's enough for 3.0 as
features/optimizations can got into 3.x.

The "real world" answer:

Getting the API right is only possible if we have spent enough time
implementing optimized algorithms. Take the Householder module:
currently there is a risk that we have to completely change the API
when we implement blocked Householder, which in itself is just used as
an optimization.

So the answer is that in addition to finalizing the API we have to do
well-chosen optimized algorithm that are likely to trigger API
improvements elsewhere.

>> We are well on our way already, and I think it's realistic to aim for
>> a beta in 3 or 4 months, with a near-finalized API, and final release
>> in Q2 2010.
> I think it is important to setup the features spec for 3.0 now and
> stick to it. Otherwise we'll suffer from feature creep. Since the gap
> has been long, I suggest that we keep the API changes for 3.0. We
> don't need to make 3.0 block for sparse module since it wasn't in 2.0.
> Also, it is unlikely that we'll get the sparse API right the first
> time, so let's not throw the entire kitchen sink of sparse module at
> it.

Yes, it's OK to keep Sparse as experimental in 3.0.

On the other hand, I would like to have all Dense modules finalized in
3.0, otherwise there's no point in releasing something.

If needed, we can make Householder and PlanarRotation into internal
things, so we don't have to worry about their API.

Let's start a wiki page on 3.0 goals.


Mail converted by MHonArc 2.6.19+