[Date Index] [Thread Index] [Author Index]
Re: Mathematica 220.127.116.11 and some General Comments
I couldn't disagree more. The core of Mathematica is still the same and very easy to learn. Difficulties arise mostly from the enormous amount of functions now included, but once used to the documentation system, that isn't a big deal. Backwards compatibility is very good. Most of the things I wrote 17 years ago still work or need only a small (semi-automated) conversion, whereas colleagues of mine are burned because their mathematical tool of choice (that is not to be named here) changed so dramatically between versions that they were busy for weeks before they could used their own programs again. In the earlier versions, finding possible values for options (or even the options themselves) was often pretty difficult. Now, most options are extensively documented (although unfortunately even in 7.0.0 you can still find the feared option description saying "Typical values include..."). I admit some function descriptions are rather dense. I remember reading the LictConvolve description over and over again, and I still need to study its doc page whenever I want to use it, but this is more the exception than the rule. The UI keeps on improving. Syntax coloring is a tremendous help (it's a pity though that the options that are strings instead of symbols do not profit from this). Pallettes, function templates, preventing of large output... The dynamic capabilities are just fantastic. I have been very pleased with them and use them almost on a daily basis. On Mar 7, 9:40 am, AES <sieg... at stanford.edu> wrote: > In article <goqphr$lt... at smc.vnet.net>, > "David Park" <djmp... at comcast.net> wrote: > > > > > The 2 January 2009 issue of Science was largely devoted to 'Education a= nd > > Technology'. There were many articles about the use of 'computer games'= , > > 'virtual reality', java applets and such but I could not find a single > > mention of Mathematica or even CASs in general. To me, this is absolute= ly > > incredible! I cannot see anything that can come even close to Mathemati= ca > > for students who actually want to DO some mathematics or learn some > > technical subject matter. I'm thinking of actually organizing calculati= ons, > > derivations or proofs, trying examples and alternative methods, and > > describing and explaining the methods. Isn't this what 'STEM' education > > should be all about? > > > A year or so ago, on sci.math.symbolic a high school student stated = that he > > was interesting in studying pure mathematics and wanted to know what > > computer language he should learn to further that goal. There were hund= reds > > of responses with some threads going to great depth. As far as I could = find, > > there was not a single mention of Mathematica and only one reference to > > using a CAS. Everything else concerned the merits of C++, Perl, List, > > Fortran etc. To me, this again was incredible. > > I am absolutely as saddened and dismayed by the above observations as I > believe David Park is -- but unfortunately also not much surprised by > them. > > I would especially endorse every word of the last three sentences of the > first paragraph above. I have long believed in, and personally and > extensively used Mathematica for exactly those purposes, in research, > teaching, educating, learning, publishing, lecturing, and personal > avocations, ever since the day I got my hands on Mathematica version 1 > (and, in fact, I pursued all those objectives using RealBASIC, Igor, > Excel, and other even more primitive tools well before that). > > On the other hand, I am also (equally sadly) not at all surprised by the > above observations. The bizarre and idiosyncratic and dysfunctional wa= y > that Mathematica has evolved, in its structure and syntax, in its user > interface, in its pricing, and in its documentation, among other > factors, all contribute significantly to the dismal situation summarized > by David above. I don't understand how things can have gone so wrong > for Mathematica -- but they have. > > Ten years ago I was writing enthusiastic memos to my Deans and Vice > Provosts for Research, and my university's IT people, urging them to > invest university funds in Mathematica licenses and bookstore discounts; > to sponsor (or pay for) free training courses for students, faculty and > staff through the university; to promote the use of Mathematica > throughout every even mildly technical area of the university; and to > make it a kind of semi-official university standard for all areas of > research and teaching (as they had, at that time, semi-officially done > for the Mac platform). > > To be blunt, if a proposal like that were actually to be taken under > consideration by my university today -- not that there's any chance that > this could actually happen at this point -- I would actively oppose it, > or at least recommend that a major effort be undertaken to explore every > other possible substitute (and there are some out there) to meet the > same needs. > > > This all leads me to believe that Mathematica is poorly positioned in t= he > > marketplace. At the present time, static printed documents are the domi= nant > > method of publication and communication in the technical world. Math= ematica > > is primarily an ancillary tool to produce calculated results and starti= ng > > graphics for these static documents. I believe that static documents ar= e a > > dying technology and they will sooner or later be replaced by active an= d > > dynamic Mathematica notebooks - or their equivalent. Why? Because they = are > > orders of magnitude superior for presenting and explaining mathematical= and > > technical ideas. WRI seems to have this concept in mind but one feels t= hat > > they don't take it with full seriousness, or realize the impediments in= the > > way. > > Unfortunately here we have to part company. Your second sentence above > makes it seem as if you've never heard of PowerPoint -- which is very > possibly "the dominant method of publication and communication in the > technical world" today -- along with QuickTime, HTML, Flash, Word, TeX, > LaTeX, the entire Adobe product line, and the ever expanding set of > online and multimedia journals and innumerable personal and > institutional web sites, all of which, working together quite > constructively in most cases, are nearly obliterating "static printed > documents", in the technical world and everywhere else. > > Mathematica is (or at least in earlier versions was, from the beginning) > absolutely an insanely great "ancillary tool to produce calculated > results and starting graphics" (including all kinds of animations) for > all the objectives we've both agreed upon. Adding a certain level of > interactivity to these graphics in recent versions is quite wonderful > also. > > But the idea that these results, along with all the underlying analysis, > will then also have to be "delivered" (presented, displayed, put on > screen) (and also organized, maintained, cataloged, re-purposed, > modified, edited, annotated), using (and _only_ using) Mathematica and > Mathematica notebooks, is just, in my considered opinion, an absolute > non-starter -- and a _highly_ destructive concept as well, in its impact > on the usability of Mathematica for what it's really good at (and where > it in fact really has very limited competition). > > I'm afraid that I'm strongly enough convinced of this point that > examining this idea further is just of no interest to me. We should be > focusing on what's needed to keep Mathematica highly _usable_ for the > particular things it's really near uniquely _useful_ for, not trying to > make it a monster tool for everything in the world.