On Tue, 30 Sep 2008, Max Shinn wrote:

>>> I think he has found that taking an extreme position is necessary for 
>>> him if he's going to make progress in getting people to be like you 
>>> and me.
>>
>> I disagree with that opinion. I think extremism turns people off.
>
> It depends on the situation.  RMS wants people to realize the importance 
> of Free Software.  For this, extremism is necessary.  He is trying to 
> set a good example.  If the leader of the Free Software Movement said 
> "I'm only going to use Free Software when it is practical", no one would 
> bother looking into the importance.  Actions speak louder than words, 
> especially here.  When people see how devoted Stallman is to this cause, 
> they look into it and see why it is so important to him.  This is 
> exactly what he wants.

All good points.  That's what I was trying to say too, but you did a 
better job of it.


> In some cases, though, it hurts more than it helps.  Read "Why schools 
> should exclusively use free software" 
> (http://www.gnu.org/philosophy/schools.html) for an example of this. 
> If you try to convince someone in a situation like this, it is important 
> to make them believe you only care about the practical benefits.

For me, the most important benefit of teaching students to use free 
software is that I am preparing them for the future.

Regarding grad students using statistical software:  We mostly teach 
students to use SAS and we don't teach them as much to use R.  R is an 
amazing Free Software program distributed under the GPL and it has really 
taken over among the professional, academic statisticians.  SAS is being 
used today mostly because it used to be used before -- it is vendor 
lock-in coupled with the fact that it does an adequate job on certain 
tasks that we often need to do in our research.  A few years ago I wanted 
to put SAS on a Linux box and SAS Inc. wanted $3,800.00 in the first year. 
That was the University Discount -- the regular price was $19,000.00 for 
the first year.  R costs $0 every year!  Also, R is vastly superior in 
some ways.  Judging from the rate of development, the future for R is 
extremely bright.  It will be unstoppable.  Will we drop SAS and start 
teaching R?  I'd do it but most people will prefer to retire than to 
change their data management and analysis systems that much, then rewrite 
their course handouts, slides, lecture notes, etc.  I understand that, but 
the students lose in a big way (but if we don't tell them, they won't 
know).

Another big thing is to transition from VMS to UNIX/Linux.  People who 
have been on VMS for 20 years don't want to change.  They don't want to 
learn Linux unless they have to.  So students have to learn SAS on VMS 
instead of R on Linux.  They don't have VMS at home, they won't have VMS 
at their next employer, etc., etc.  You get the idea.

Mike