Re: Review ode csets

[c.]

Rik,

On 18 Oct 2016, at 20:40, Rik <address@hidden> wrote:

> There is also a patch I submitted to https://savannah.gnu.org/bugs/?49364
> that needs review.

This one looks fine to me, merging with a loop seems to also remove the need
of protecting against the case of fields with cell-array value so if it is 
faster
I like the change.

On the other hand, if we are going for performance optimization, fixing the 
issue
discussed in this thread:

http://octave.1599824.n4.nabble.com/Default-merged-to-stable-for-upcoming-4-2-release-tp4679934p4680015.html

would have a much larger impact in real-life scenarios.

Incorrect time-step estimation may result in time-step rejection and usually 
function evaluation
is the real bottleneck in any ode simulation.

And it is even more important to avoid unnecessary function evaluations because 
we don't have JIT.

c.