Slashdot Mirror
Wolfram Research Releases Mathematica 7
mblase writes "Wolfram Research has released the seventh version of Mathematica, and it does a lot more than symbolic algebra. New features range from things as simple as cut-and-paste integration with Microsoft Word's Equation Editor to instant 3D models of mathematical objects to the most expensive clone of Photoshop ever. Full suites of genome, chemical, weather, astronomical, financial, and geodesic data (or support for same) is designed to make Mathematica as invaluable for scientific research as it is for mathematics."
"[It] is designed to make Mathematica as invaluable for scientific research as it is for mathematics." Cut down the advertising please. Or at least advertize some free software. It's been a while since I needes a computer algebra system. How are the free alternatives coming along? Any recommendations?
This just seems like its got so bloated that it will likely be priced beyond the budget of most students.
I don't see why we have to have these all encompassing suites anyway, what's wrong with small tools at low cost which work together?
Its most likely that students who want but can't afford this will hit the torrent trackers, which isn't really what we want.
A learning experience is one of those things that say, 'You know that thing you just did? Don't do that.' - D. Adams
>That would be nice,
It is.
>but doesn't solve the problem
It will.
>of Mathemitca's notorious copy protection.
The Pirate Bay verison of mathematica usually includes protection from copy protection.
>From what I hear, even legitimate owners often have trouble getting past it.
Legitimate owners of ANY copy protection system are generally having orders of magnitude more problems with those systems than users who just get clean copies at their Pirate Bay.
OK, perhaps I should explain something.
1. Release team
2. Initial distribution (closed FTP, IRC etc)
3. Usenet
4. P2P (Torrents, Kademlia, etc)
The pirate bay folks don't do anything except shuffle bits.
For large sets, this will be our guide even unto death, for the LORD will work for each type of data it is applied to...
You certainly don't know what it takes to program a scientific program. It's not your "usual" large-scale application. It's far more involved than you think it is. You cannot simply have some mad programming skills to make it. You also need to have some mad math skills to make some REAL optimizations.
I have found one problem with open source toolchans - producing good quality graphics. At the end of the day you have to present the data, and gnuplot just isn't cutting it anymore.
I completely agree that computer science is related to math. I personally actually like math and physics and whatnot, to the extent that I'm reading a textbook (I guess) on spacetime physics for fun. But I think you have to admit that many aspects of computer science *today* are very far removed from actual mathematical calculations or even mathematical ability (e.g., you don't have to take calculus to write a PHP script). You don't have to have any electrical engineering knowledge to "build" a computer, either, really. You have to be able to plug this into that, maybe know some basic specs about the components, etc.
I've used Maple, Mathematica, and SAS, among other products, for mathematical and/or statistical analysis. From a programming/features perspective, each has its own strengths--and weaknesses.
I'll only briefly mention cost. These things are expensive because it's not like any random programmer can build this kind of software. Especially with Mathematica, these are heavily-researched algorithms that are nontrivial to implement. Also, the market is small for such a specialized and sophisticated application. Your average person isn't ever going to be able to use something like this. They barely know what the quadratic formula is. (They should, but that's an entirely different story.) You think they need to invert a 20x20 matrix? Or compute the Galois group of a quintic? Or even do a simple hypothesis test?
As for the image manipulation stuff, I think that comparisons to Photoshop are a bit naive. Clearly, it's not supposed to be for people who want to do red-eye reduction on their family photos. It's not even for graphic designers or photographers. It's for scientists who want an algorithmic approach to adjusting their images, either for research or for purposes of publication. Could you do these things in Photoshop? Sure. Could you then say what formula or algorithm was applied to the image to produce that specific result? No. And conversely, you wouldn't do layer composition, masking, or on-the-fly tonal adjustments with Mathematica.
FWIW I hate the copy protection on it too. It's infuriating and a burden to legitimate users while doing little to deter piracy.
Maxima also sucks. Here's a session from just this afternoon.
[omf@midgar 14:45:36 ~]$ maxima .....
Maxima 5.13.0 http://maxima.sourceforge.net/
Using Lisp GNU Common Lisp (GCL) GCL 2.6.8 (aka GCL)
Distributed under the GNU Public License. See the file COPYING.
Dedicated to the memory of William Schelter.
This is a development version of Maxima. The function bug_report()
provides bug reporting information.
(%i1) Q=matrix....
(%i11) Q.T.transpose(Q);
(%o11) matrix([cos(t) (cos(t) T11 - sin(t) T12)
- sin(t) (cos(t) T21 - sin(t) T22), cos(t) (cos(t) T12 + sin(t) T11)
- sin(t) (cos(t) T22 + sin(t) T21), cos(t) T13 - sin(t) T23],
[cos(t) (cos(t) T21 - sin(t) T22) + sin(t) (cos(t) T11 - sin(t) T12),
cos(t) (cos(t) T22 + sin(t) T21) + sin(t) (cos(t) T12 + sin(t) T11),
cos(t) T23 + sin(t) T13], [cos(t) T31 - sin(t) T32, cos(t) T32 + sin(t) T31,
T33])
(%i12) trigsimp(%);
Universal error handler called recursively (:ERROR NIL
CONDITIONS::CLCS-UNIVERSAL-ERROR-HANDLER
""
"Couldn't protect")
Universal error handler called recursively (:ERROR NIL
CONDITIONS::CLCS-UNIVERSAL-ERROR-HANDLER
"" "Couldn't protect")
Maxima encountered a Lisp error:
Error in CONDITIONS::CLCS-UNIVERSAL-ERROR-HANDLER [or a callee]: Caught fatal error [memory may be damaged]
Automatically continuing.
To reenable the Lisp debugger set *debugger-hook* to nil.
(%i13) Q.trigsimp(T.transpose(Q));
(%o13) matrix([cos(t) (cos(t) T11 - sin(t) T12)
- sin(t) (cos(t) T21 - sin(t) T22), cos(t) (cos(t) T12 + sin(t) T11)
- sin(t) (cos(t) T22 + sin(t) T21), cos(t) T13 - sin(t) T23],
[cos(t) (cos(t) T21 - sin(t) T22) + sin(t) (cos(t) T11 - sin(t) T12),
cos(t) (cos(t) T22 + sin(t) T21) + sin(t) (cos(t) T12 + sin(t) T11),
cos(t) T23 + sin(t) T13], [cos(t) T31 - sin(t) T32, cos(t) T32 + sin(t) T31,
T33])
(%i14) trigsimp(Q.trigsimp(T.transpose(Q)));
Segmentation fault
[omf@midgar 14:48:25 ~]$
Computer algebra systems are not the best to begin with, but Maxima has a very, very long way to go before it can compete with Mathematica. Most of my analytical work on a daily basis is done using Maxima and I can safely say that the program could be a lot better than it currently is.
May the Maths Be with you!
I thought that Sage was quite easy to learn - and I hadn't used Python before I started. I haven't used Mathematica, but I've used Maple and Magma for a few years. There didn't seem to be any difference in learning curve between Sage and the commercial options. In some cases the tutorials and reference were a lot more helpful.
In general, open source software tends to have a crap user interface compared to the commercial application being cloned. In this case the browser based notebook is up there with maple, although the proper latex integration is more powerful. Using Python as the underlying language was genius, as it has a much cleaner design than the languages used in most CAS.
What parts of Sage did you think were difficult to use?
Slashdot: where don knuth is an idiot because he cant grasp the awesome power of php
I really think that it would solve any math problem that one would ever come across as an undergrad (and even grad level). I had a prof whom I visited once after graduation and he had just gone through one of his student's Ph.d. thesis and in a couple of hours reproduced what took the woman a year to do by hand.
$45 for a year of the most powerful math tool in existence. Just for the new functions thats $0.07 per function per month.
And may I stress here that bug reporting is one of the most helpful things you can do for any open source project? If you can provide a simple way to reproduce a problem, it is likely as good as fixed.
So, don't complain, report! (after that you may optionally complain). But don't assume developers will find it themselves, or that others will report the problem for you..
"It's too bad that stupidity isn't painful." - Anton LaVey