|Subject:||Re: discuss the odepkg package update for octave 4.1|
|Date:||Thu, 13 Nov 2014 09:57:24 -0500|
On Thu, Nov 13, 2014 at 9:48 AM, Jacopo Corno <address@hidden> wrote:Dear Doug,
during the last two SoCiS programs Roberto Porcù and myself worked on odepkg. The idea was to change the inside structure of the solver to unify them and add some simplectic solvers and, as you said, to ensure compatibility with Matlab, to add the main function to the core (odeset, odeget, and the main solvers like ode45 for example) and make a new release of the package.
I just finished my SoCiS period and I was going to submit my modification to the community in the next few weeks.
You can have a look at the current state of the code at:
Thank you. This is why I asked before I spent too much time on it.
Nella citazione in data giovedì 13 novembre 2014 15:16:18, Doug Stewart ha scritto:
DASCertificate for 206392A question was asked in the octave chat room about odepkg, so I
tried it with the development branch and could not install it from
-forge. I then tried 3.8.1 and it installed OK from forge.
I discussed with Carandraud the idea that I would look into updating
the pkg to work with 4.1. Carandraug said the maybe we should be
moving these functions into Octave core, because they are in Matlab core.
Should they be moved to Octave core????
In looking at the files in odepkg I see that there are some tared
files and that the make file tries to untar the files and the apply
some patches to these FORTRAN files. there are 3 different tared files.
As a start to straighten out this pkg I propose to untar these files
and apply the patches, then change the make file to use the patched
files. I will then make pkg that works with the development branch.
When this is working then should we keep it as a pkg or move it to
Any suggestions (guidance) are most welcome.
Jacopo Corno M.Sc.
Technische Universität Darmstadt
Graduate School of Computational Engineering
64293 Darmstadt / Germany
Phone: +49 6151 16 - 76877
|[Prev in Thread]||Current Thread||[Next in Thread]|