The time before Mathematica (2013)

http://blog.stephenwolfram.com/2013/06/there-was-a-time-before-mathematica/

by geyang • 11 years ago

27 10 11 years ago

blog.stephenwolfram.com

codeulike• 11 years ago

Modesty is not his strong point, but if you've never played with Mathematica, have a play with it - the way it works is very interesting.

zitterbewegung• 11 years ago

Mathematica / Wolfram Language I would describe as you take the idea of M-expressions from John Mccarthy's paper and program nearly every single Mathematical algorithm that you can think of.

For instance you can do these things without calling an external library or datasource:

Classify some data using the examples. Oh yea the text is included in the program or automatically downloaded. Also you don't have to choose a model for this classification step it just attempts to classify based on the preprogrammed models. Accept user input to classify the unknown. (Yea it has its own UI code).Deploy to the cloud (this is sort of in beta but it works). Export your notebook to latex / a slideshow once you have submitted your work to someone so you can explain it.

mnl• 11 years ago

I like Mathematica quite a bit (the Front End/Kernel makes a lot of sense), but I don't really like Wolfram after he left Physics. The way he downplays Reduce, Macsyma (Maxima), the pioneering work that was Veltman's Schoonschip and then FORM... He doesn't even mention the folks at Maple. Computer Algebra was a well established field way before Mathematica appeared in 1988[1]. He's smart and very good at selling his product, that's fine, and for a long time his company has had the resources to hire a lot of good people and you can tell/take advantage of it when using Mathematica. But honestly, I think we could do without.

Personally, I find more interesting David R. Stoutmeyer, who made the first CAS for microcomputers in the late 70s, muMath for CP/M, and then the tiny PICOMATH-80 for the TRS-80. He went on to write Derive, that was much loved in the early 90s, because it was very good and you could afford it while Mathematica has always been expensive and for some years (until version 2.2 really) you needed either a souped- up Macintosh or a workstation to run it. Then Texas Instruments hired him in order to develop the Derive-based TI-92/89 (Derive was written in Lisp, its version for those in C BTW) and ended up buying Soft Warehouse and shutting Derive for good... which is just sad and nobody cares.

I'd like to ask for some love for Reduce, now open sourced and hosted at SourceForge -not cool any more, but it lives there-. It still makes a few calculations easier than Mathematica.

[1] See for instance this very interesting review of the early Mathematica: http://www.cs.berkeley.edu/~fateman/papers/mma.review.pdf

ska• 11 years ago

He was much the same about his scientific work as I recall, it's standard Wolfram M.O. : discount, ignore, or attempt to outright claim credit for the fundamental work of others (preferably without mentioning them by name) while puffing up your own contributions.

It's an unfortunate and deep character flaw, but he still has made some interesting contributions that are worth evaluating on their own merits (neither as fundamental nor as impactful as he would suggest, typically)

ssivark• 11 years ago

I do not have a personal grouse against Wolfram, but it takes a level of self-importance to (re)name every product of your company after yourself, even if you might be coming up with some key ideas and the vision. Eg: Wolfram Alpha, Mathematica --> Wolfram Language, etc.

It seems wrong for him to take sole credit for what I'm guessing a talented and hard-working team at Mathematica has achieved together, over the years.

psychometry• 11 years ago

At this point, I'm pretty sure a list of Stephen Wolfram's favorite pie recipes would begin with a history of his childhood academic accomplishments. It's almost nauseating reading this crap.

kyberias• 11 years ago

There's an invitation to solve a puzzle hidden in this text. Has it been solved yet?

mayoff• 11 years ago

I hope nobody solves it "just for fun" without a promise from Wolfram to free the decrypted SMP source code.

oneJob• 11 years ago

In the time before Mathematica, mathermatics was non-proprietary and open source.

hga• 11 years ago

No so in the case of Macsyma and Wolfram's own SMP (https://en.wikipedia.org/wiki/Stephen_Wolfram#Symbolic_Manip...): "Wolfram led the development of the computer algebra system SMP (Symbolic Manipulation Program) in the Caltech physics department during 1979–1981. A dispute with the administration over the intellectual property rights regarding SMP—patents, copyright, and faculty involvement in commercial ventures—eventually caused him to resign from Caltech.

It's not infamous enough, but Symbolics managed to arrange to solely license Macsyma, as far as we could tell to keep it out of everyone else's hands. Wikipedia seems to be accurate in this section: https://en.wikipedia.org/wiki/Macsyma#Commercialization based on my communications with Danny Hillis after the sole licencing when I was working for LMI (which also was selling SMP), working for Gold Hill in 1984, and working for Joel Moses in the late '80s (he was indeed none too happy about how MIT had treated him WRT to Macsyma).

The time before Mathematica (2013)

Related Stories