MathGroup Archive 2011

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Royalty free runtime for Mathematica

  • To: mathgroup at smc.vnet.net
  • Subject: [mg117864] Re: Royalty free runtime for Mathematica
  • From: Fonseca <public at fonseca.info>
  • Date: Sat, 2 Apr 2011 02:41:26 -0500 (EST)
  • References: <il50rp$n9u$1@smc.vnet.net> <il7q5n$dpc$1@smc.vnet.net>

Hi,

Although there was almost no feedback on this one, I did get some
information and opinions by e-mail, both private and from Wolfram
Research.

Unfortunately most opinions tend to point to the greatness of cdf, and
although I can see a bright future for the cdf technology, the way it
is today is completely not fitted for most of the corporate tools I
can relate to:
- no integration. There is no way to call the cdf functionalities from
another environment and get back an answer. This is absolutely not
fitted in the current corporate working environment. This is the way
we used to work on the 90's...
- not possible to hide code. A simple compilation would be already
something, but I can't even do this without very astute tricks (I can
compile, and trick the compiled function so that the uncompiled code
it stores is just a message to the user, and not the original
proprietary algorithm, although this is not very safe coding...).
- not possible to control access to the software/tool. Normally this
is done, in the corporate world, by the use of an usb dongle (or a
computer id signed license), but since the integration is non
existent, I cannot figure out a way of implementing this kind of
strategies on a cdf (either way, even if I could, I cannot hide
code...)

I know player pro allowed encryption, and most probably, cdf will end
up having DRM. But the use I'm mentioning needs the three listed
functionalities putted together.

Integration is the present and is the future. Who better than WR to
know this when they keep integrating everything into the same package?
What they cannot integrated are my business tools, and since not all
of them should be built with Mathematica (imagine an ERP built with
Mathematica...), the end result of the tools built with Mathematica
definitely need integration capabilities with the rest of the
corporate environment.

Someone told me that I cannot say Mathematica lacks integration.
Definitely Mathematica has a lot of integration capabilities, but I
cannot install Mathematica on 200 computers of the company. Unless I'm
building Wolfram|Alpha, I don't see how it would be economically
feasible compared to building the tool in another programming
environment (I know with Mathematica things get done quicker, but not
that quicker...). Most of you are probably thinking of player pro
(less expensive). Well, 200 licenses of player pro is also very
expensive, and there are great alternative high end programming
languages that definitely end up winning an economical balance (and I
also need better integration than the one made available by the player
pro).


Is there really a great risk to WR on having a compiled Mathematica
library be called from a different environment, eventually contact the
installed free cdf player, and return the result?

And, on the other way around, why can't I build a cdf that can
interact with other outside tools and files, in the same way I can do
with Mathematica?

Only a full Mathematica could build either the cdf or the compiled
library. Would this reduce the number of Mathematica licenses sold or
actually increase it? How big would be the new market?

Regards,
P. Fonseca


  • Prev by Date: Re: NonlinearFit for specific data
  • Next by Date: Re: why extending numbers by zeros instead of dropping precision is a good idea
  • Previous by thread: Re: Problem: Approximate complex values in Mathematica 8.0.1
  • Next by thread: SumConvergence mistake?