Re: Symbolic Calculator for Mobile Devices

I'm unclear on why you are spending time on this. The smallest
laptop probably weighs < 2 kilograms and is the size of a paperback
book. It can run Maxima, Maple, Mathematica. It has a keyboard
and mouse-equivalent. Sometimes it has a wireless internet and
phone access like Skype.

Is there an application you have in mind where someone needs to
take that extra step and do calculus homework on a cellphone?
Would it be simpler to use instant-messaging could get you access
to a much larger computer? Or interact with a web page somehow?

Robert is right that you would be more than welcome to
contribute to advancing the art, if you tire of hacking on the current
mobile computing API.

By the way, the original Macsyma on a PDP-10 time-sharing
system occupied, along with Lisp, a maximum of 1.2megabytes (that was all
the address space of the PDP-6 computer): 256K words X
36 bits per word.

Regards,
Richard Fateman

..
"Helmut Dersch" <der@xxxxxxxxxxxxxxxx> wrote in message
news:20998343.1147596798158.JavaMail.jakarta@xxxxxxxxxxxxxxxxxxxxxxxxx
I'd like to take this opportunity to suggest that
reinventing the wheel
in this case might be suboptimal, if the goal is to
get a working
symbolic computation system on some device. Maxima
embodies
a fair amount of mathematical and quasi-mathematical
knowledge,
and many details have been hashed out over a period
of decades.
So I recommend pretty strongly that new projects
start by working
with Maxima itself, rather than trying to reimplement
it.


Hi Robert,

I agree, but still had to reinvent the wheel.
There are a few mobile devices which run
clisp and maxima, but the universal language in this
area is Java, which almost any cellphone can handle,
as opposed to the 1% (guess) Lisp-capable.
There are some translators of Lisp to Java but
(a) most can only handle Scheme
(b) none of the more extended translators translates
Maxima
(c) none of the more extended translators translates into
mobile Java.
Point (c) is more severe than it might look. Eg all java-based
Lisp-systems heavily depend on Java Reflection
which is not available in mobile devices.
Given the typical hardware constrains (200kByte
binary size, 1 MByte free RAM) also makes Maxima
not the ideal candidate for a port.


As outlined in the docs, I also tried
to port a few existing Java projects to the mobile
API without success.

Finally, I have some experience in porting stuff
to mobile devices from my previous project FnattLabME,
which is a port of the Matlab-clone FnattLab to
mobile devices. I ended up rewriting much of the
program, because mobile devices aren't just
small PCs. Eg: It doesn't suffice to get a program
running if startup takes 5 minutes etc.

Disclaimer: I'm working on the Maxima project, so I
have a
personal interest in getting people on board rather
than striking
out on their own ....


At least, I choose to use Maxima syntax as opposed
to maybe Mathematica or Maple, so people are encouraged
to use this on their PCs.

Regards

Helmut Dersch


.

Relevant Pages

  • Re: Symbolic Calculator for Mobile Devices
    ... with Maxima itself, ... area is Java, which almost any cellphone can handle, ... none of the more extended translators translates into ... which is not available in mobile devices. ...
    (sci.math.symbolic)
  • Re: Future of MFC?
    ... put the MFC into the system... ... Yes, it's not uncommon to find a JVM in ROM on mobile devices, but it tends ... to be the J2ME version of Java, ... runtime environment that they could all implement. ...
    (microsoft.public.vc.mfc)
  • Re: Future of MFC?
    ... put the MFC into the system... ... Yes, it's not uncommon to find a JVM in ROM on mobile devices, but it ... to be the J2ME version of Java, ... runtime environment that they could all implement. ...
    (microsoft.public.vc.mfc)
  • Re: Beginner question on Huffman implementations
    ... data that's meant to be stored on the persistent storage portion of J2ME (Java 2 ... MicroEdition) based mobile devices. ...
    (comp.compression)