Re: Matlab quirks

[John W. Eaton]

On 26-Mar-2012, Miguel Bazdresch wrote:

| I see that I was careless in what I wrote, and I implied things I didn't 
| mean to. I apologize for that.
| 
| What I meant to say is this: the unexpected difference between Matlab 
| and Octave's num2str caused a lot of grief to some of my students. They 
| did a large project, tested it on Matlab, but I couldn't reproduce their 
| results. It took some digging on my part to find the problem, and 
| meanwhile their grades were in suspense.
| 
| Students talk among themselves, and I seem to notice that in this 
| particular case Octave's reputation was somewhat tarnished. For the 
| first time in several years, I have seen more students switch to Matlab 
| than the inverse (usually less than 30% of my students use Octave at the 
| start of the class, but 80% to 90% use it by the end of the semester).
| 
| By "a quirk in Octave", I meant it was perceived as such by the 
| students, since Matlab's behavior made more sense in their eyes. Octave 
| was "guilty" that their grades were at risk. I have explained the 
| problem to the students, and some seem impressed that I could report the 
| problem and people immediately started to work on it (thanks guys!), but 
| some remain unconvinced on Octave's trustworthiness, despite my efforts.

Do you explain to them that Matlab also has bugs, but that fixing bugs
in Matlab requires the actions of a single company?

jwe