On 24/01/2014 06:40 PM, Warp wrote:
> Orchid Win7 v1<voi### [at] devnull>  wrote:
>> Fun thing: Apparently if you buy a Raspberry Pi, it now comes with a
>> free copy of Mathematica.
>
>> Not some crippled version. The entire thing. For free.
>
>> (Recall that to *buy* Mathematica is currently over ?8,000.)
>
> I really have to wonder how they can afford doing that (I'm assuming
> this whole thing is completely legal.) Maybe they got some kind of
> odd deal with Wolfram Research.

It was Mr Wolfram's idea. He thought given that the Pi is geared towards 
education, it would be a cool idea to get all the kids used to using 
*his* premium-grade math package, so that when they grow up big and 
strong, they'll beg their employers to buy a full-price copy.

Kind of like how MS lets you have a free 30-day trial of Visual Studio, 
even though the thing actually costs thousands of pounds.

> It's just a bit odd, because there are free math software out there
> as well. Sure, most of them is not even nearly as good as Mathematica,
> but still...

I've never seen anything else that does what Mathematica does which 
isn't also klunky, kludgy and generally useless. (Then again, I haven't 
looked very extensively.)

Having said that, the GUI for Mathematica is real ugly. I suppose you 
have to expect that from such ancient software. But Jesus, put some 
anti-alias on that thing!!

The thing is, the symbolic computation engine in Mathematica can 
probably be implemented in a day or two of coding in just about any 
programming language by a half-competent person. What you *can't* easily 
duplicate is the man-decades of R&D that Wolfram Research has put into 
the libraries you get with Mathematica. I have no idea what the hell a 
generised hypergeometric function even *is* - but I can promise you 
Mathematica has a function for it...

It's easy to write a program that can do trivial algebra rearrangements. 
(I did it myself, when I was 17, using Borland Turbo Pascal.) Getting it 
to actually work right for _all_ simple expressions, though... that's 
really tricky. Getting it to be 100% correct all the time, in the face 
of possible division by zero, different number fields (complex vs real), 
taking into account the branch cuts of various functions... All of this 
is way more tricky than it initially appears.

Post a reply to this message