Re: ode45 inclusion into core

[Richard Crozier]

On 07/10/15 08:02, Olaf Till wrote:
> On Tue, Oct 06, 2015 at 11:03:04PM +0100, Richard Crozier wrote:
> > On 06/10/15 17:52, Olaf Till wrote:
> > > On Tue, Oct 06, 2015 at 04:09:54PM +0000, Carlo De Falco wrote:
> > > > On 6 Oct 2015, at 18:02, Richard Crozier <address@hidden> wrote:

<snip>

> I still wan't to explain my view a bit better than just shouting
> 'ugly':
>
> - I think anonymous functions, being a general mechanism, are to be
>    prefered over explicite extra arguments, which would have to be
>    implemented separately for each case.
>
> - The only issue seems to be that currently there are bugs preventing
>    anonymous functions from always being saved correctly. It may be
>    acceptable for external packages to work around such bugs. But
>    within core Octave, the right way is to fix the bugs (which should
>    be feasible in these cases, the code is there), instead of working
>    around them at other places.
>
> This should not negatively affect you (Richard) (except for a need to
> alter your code to use anonymous functions for extra arguments), since
> at the time Octave is released with the new ODE functions, the above
> bugs should also be fixed.
>
> Olaf


ok, but one question I have, if I create anonymous functions that all 
use the same data, does this result in the creation three copies of the 
data in memory? Do variables in anonymous functions obey the 
copy-on-write semantics even when they are passed into functions via 
anonymous functions?

Richard

--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.