|
On 02/20/2015 06:36 PM, Carnë Draug
wrote:
On 19 February 2015 at 18:07, Olaf Till <address@hidden> wrote:
If Matlab code gets run in Octave and then is subject to automatic
broadcasting, it may "work" (in the sense of "runs w/o error messages) but
there's no guarantee it'll yield the same (correct or not) results as in
Matlab. Right?
No, wrong. It will return exactly the same results as Matlab. Unless you
count throwing an error as a "result". If the function in Matlab works
with the input you give it, then there is no automatic broadcasting done.
Automatic broadcast only happens in cases that in Matlab would cause an
error. If the code works in Matlab, then it works the same way in Octave.
Looking into the future, Mathworks might see the usefulness of
broadcasting, and implement it in Matlab. If past performance is to
guide us, it might be different from Octave's current approach.
Having some sort of warning infrastructure might end up being good,
although of course eventually Octave is supposed to roll its eyes
and adapt to whatever Matlab ends up doing.
This doesn't have to be the case, of course, except that Mathworks
seems to studiously pretend that Octave does not exist. I never
understood that: whenever I talked to their reps and salespeople, I
suggested that they engage in cooperation with Octave and see it
more as a R&D tool, in the sense of the relationship between
RedHat and Fedora. I actually think that Octave has a positive
impact on their market share, but I am afraid they don't see it that
way. Did anyone from the development team have any positive
interactions with Matlab developers? Maybe it'd be a good idea to
make another contact attempt?
|