MathCad for Mac?
Keith W. Whites
the tech support people, but nobody returned my call.
Also, does anyone know if MathCad6 (the latest version on the PC ?)
supports vector plots or animation?
Thanks for any information you provide.
______________________________________________________________________
Keith W. Whites e-mail: whi...@engr.uky.edu
Assistant Professor
Department of Electrical Engineering Tele: (606) 257-1768
University of Kentucky FAX: (606) 257-3092
453 Anderson Hall
Lexington, KY 40506-0046
Christopher Wright
<whi...@engr.uky.edu> wrote:
> Does MathSoft support MathCad for the Mac?...nobody returned my call.
No kidding? Anyway you're lucky. What they do have isn't much good.
Support comprises suggestions to buy Windows machinery
--
Christopher Wright P.E. |"They couldn't hit an elephant from
chr...@skypoint.com | this distance" (last words of Gen.
Voice phone (612)933-6182 | John Sedgwick, Spotsylvania 1864)
eric larson
> the tech support people, but nobody returned my call.
No. MathSoft dropped it's Mac support some time ago. They have a lot to prove
to me before I would touch their products again.
David Gutierrez
Inazaki) wrote:
> Rumor has it that MathSoft is getting back into Macs.
> They claim that they'll be putting out a new version of MathCad
> for Macs.
Lots of luck finding customers. I wouldn't touch any product from one of
these off-again/on-again developers. Ditto for Lotus, unless IBM does a
good job convincing us that they've made real changes at the company.
--
David Gutierrez
d...@biomath.mda.uth.tmc.edu
"Only fools are positive." - Moe Howard
Eric Inazaki
: > Does MathSoft support MathCad for the Mac? I've tried to get a hold of
: > the tech support people, but nobody returned my call.
: No. MathSoft dropped it's Mac support some time ago. They have a lot to prove
: to me before I would touch their products again.
Rumor has it that MathSoft is getting back into Macs.
They claim that they'll be putting out a new version of MathCad
for Macs. Much remains to be seen but it sounds like they have
a _hell_ of a lot of work to do.
We'll see...
Mike Hamlyn
larson) wrote:
> > Does MathSoft support MathCad for the Mac? I've tried to get a hold of
> > the tech support people, but nobody returned my call.
>
> No. MathSoft dropped it's Mac support some time ago. They have a lot to prove
> to me before I would touch their products again.
As far as I know MathCad 6 will be available for mac next year (probably
only Powermac from what I understand). For more info look at their websites
http://www,mathsoft.com
http://www.adeptscience.co.uk
At least we'll be able to use the package without resorting to Windows
--
Mike Hamlyn tel +44 1785 353243
School of Engineering fax +44 1785 57073
Staffordshire University UK e-mail M.G.H...@staffs.ac.uk
ST18 0AD
occasional Webpage http://bsmmgh.staffs.ac.uk
Bob Arthur
> Does MathSoft support MathCad for the Mac? I've tried to get a hold of
> the tech support people, but nobody returned my call.
I'm not surprised they didn't return your call. MathSoft dumped the
Macintosh market and their Mac customers, of which I am one.
MathCad 3.0 is the latest for the Mac, and I bought it a few years ago
along with a few "notebooks". What a waste of money! The thing runs too
slow to be usable. The last I heard from MathSoft was an announcement of
ver 4.0, for Windows, along with a cheerful offer for an "upgrade" to the
Windows version for their Mac customers. I told them to stuff it.
Now they might be producing version 5.0 for the Mac. Who cares. They
lost me.
--Bob
Stuart Beaton
(Bob Arthur) wrote:
I might be inclined to agree with Bob's sentiments, but for what it's
worth, a friend just told me that he saw the old edition, V3.0 at the
campus store, and a Student Edition of the new version there as well. Both
may have been < $100, with the new Student Version slightly more
expensive. With the way MathSoft hung Mac user out, you might consider
switching to Mathematica (first introduced on the Mac) or Maple, or
Matlab, depending on your needs.
Stuart
Alex Morando
Stuart Beaton <sbe...@du.edu> wrote:
>In article <BArthur-3107...@ppp029-vand.wis.com>, BAr...@aol.com
>(Bob Arthur) wrote:
>
>> MathCad 3.0 is the latest for the Mac, and I bought it a few years ago
>> along with a few "notebooks". What a waste of money! The thing runs too
>> slow to be usable. The last I heard from MathSoft was an announcement of
>> ver 4.0, for Windows, along with a cheerful offer for an "upgrade" to the
>> Windows version for their Mac customers. I told them to stuff it.
>>
3.X) and I use it to do some of my calculations.
>I might be inclined to agree with Bob's sentiments, but for what it's
>worth, a friend just told me that he saw the old edition, V3.0 at the
>campus store, and a Student Edition of the new version there as well. Both
I've seen version 3.X going for about $20 at surplus stores.
--
! Alex Morando, Engineer-at-large - a...@netcom.com,amo...@aol.com
! My employer does not usually agree with the views given above.
Louis M. Pecora
> In article <sbeaton-0308...@sl-8.ducomm.du.edu>,
> Stuart Beaton <sbe...@du.edu> wrote:
> >In article <BArthur-3107...@ppp029-vand.wis.com>, BAr...@aol.com
> >(Bob Arthur) wrote:
> >
> >> MathCad 3.0 is the latest for the Mac, and I bought it a few years ago
As a former, unsatisfied, MathCad user I recommend:
DON'T WASTE YOUR MONEY ON MATHCAD.
For heavy duty mathematical work it's worth it to learn Mathematica or
Maple.
For lighter work, perhaps Theorist.
For strictly numerical work check out HiQ. I had a demo that looked pretty
good (way better than MathCad), but I have never really used it. Any
opinions on HiQ?
Also check out MATLAB (do not confuse with MathCad -- different products,
different companies). I hear good things about MATLAB, but I think it is
expensive.
--
Louis M. Pecora
pec...@zoltar.nrl.navy.mil
/* My views and opinions are not those of the U.S. Navy.
If you want those, you have to start a war. */
Keith W. Whites
>worth, a friend just told me that he saw the old edition, V3.0 at the
>campus store, and a Student Edition of the new version there as well. Both
>expensive. With the way MathSoft hung Mac user out, you might consider
>switching to Mathematica (first introduced on the Mac) or Maple, or
>Matlab, depending on your needs.
> Stuart
I totally agree. I've been using Mathematica since version 1.0 and really like the
program (though I admit I still can't remember most of the syntax -- that's what
laps are for!) Unfortunately, in a project I'm working on, someone want's me to use
MathCad. Version 6plus is out now and it supports animation and vector plots
together with hyperlinking between sheets on the same disk as well as the net.
Pretty neat -- all's I need is a PC. Does anyone know what "Pentium" means? I
think it's some kind of five-bladed flower, but I'm not sure.
AARenshaw
wrote:
> As far as I know MathCad 6 will be available for mac next year (probably
> only Powermac from what I understand). For more info look at their
> websites
> http://www,mathsoft.com
> http://www.adeptscience.co.uk
MathSoft is coming out with Mathcad 6.0 for Mac early next year. I have
been hired to help them in defining the needs of this market. It was
interesting that you said you'd heard we are only going to have a Power
Mac version. That decision has not been made, although right now we are
leaning towards developing for both Power Mac and 680x0-based Macs.
It would be very useful to get feedback from the members in this forum on
that issue and other features that they would like to see in a Mathcad for
Mac product. Also, from reading through the messages on this topic, it's
obvious that MathSoft is going to have to do a lot to prove our commitment
to Mac. Please let us know what you want in a Mathcad for Mac product and
how we can prove to you that we are committed to the platform.
FYI, below is a list of some of the features in the released version of
Mathcad 6.0.
Thanks in advance for your feedback.
Amy
***************************
NEW: Indicates features new in versions of Mathcad 6.0.
PLUS: Indicates features available in Mathcad 6.0 PLUS Professional
Edition only.
Numeric Calculation Functions:
* Operators and built-in functions for manipulating numbers,
vectors, matrices.
* Units of measurement and dimension checking.
* Simultaneous equation solving.
* Derivatives, integrals, summations and products.
* Trigonometric, hyperbolic, exponential and Bessel functions.
* Find roots of a polynomial.
* Fast Fourier transforms.
* Wavelet transforms. (PLUS)
Symbolics/Calculation Functions:
* "Live" symbolic solutions to individual equations or systems of
equations. (PLUS)
* Symbolic integration and differentiation.
* Expand, simplify and factor expressions.
* Inverse, transpose and determinant of a matrix.
* Laplace, z, Fourier Integral transforms and their inverses. (PLUS)
Graphics and Visualization Tools:
* (NEW) 2-D vector plots.
* 2-D rectangle and polar plots.
* (NEW) 3-D scatter and bar plots.
* 3-D contour, surface and parametric surface plots.
* Log and linear axes.
* (NEW) 3-D axis options.
* Trace and zoom.
* (NEW) Animation.
* (NEW) Image viewing and processing. (PLUS)
Statistics and Data Analysis:
* (NEW) Statistical distributions: chi-square, F, t, uniform, normal,
binomial, Poisson.
* (NEW) Additional statistical distributions: beta, Cauchy,
exponential, gamma, geometric, log-normal, logistic, negative
binomial and Weibull. (PLUS)
* Standard methods of analysis, including parametric and non-
parametric hypothesis testing, confidence interval construction,
computing standard errors, analysis of variance, principal
component analysis and Monte Carlo techniques.
* (NEW) Curve fitting (polynomial regression).
* (NEW) Median Data Smoother.
* (NEW) Data Smoothing (3 additional data smoothing functions in
PLUS: Loess, Supsmooth, Ksmooth).
Differential Equations:
* 4th-Order Runge-Kutta method. (PLUS)
* 5th-Order Runge-Kutta fixed and adaptive step-size methods.
(PLUS)
* Bulirsch-Stoer fixed and adaptive step-size methods. (PLUS)
* Bulirsch-Stoer methods for stiff systems. (PLUS)
* Linear Shooting and Relaxation methods for boundary value-
problems. (PLUS)
Advanced Matrix Functions:
* Matrix decompositions such as QR, LU and singular value. (PLUS)
* Linear systems solver. (PLUS)
* Matrix analysis tools for row reduction, computing rank, and
finding matrix norms and condition numbers. (PLUS)
* Eigenvalue/eigenvector capabilities for real and complex
matrices. (PLUS)
Document Preparation Features:
* Print preview and technical spell checker.
* (NEW) Export worksheets to Microsoft Word.
On-line Learning Aids and Help:
* (NEW) QuickSheets.
* On-line tutorial and context-sensitive help.
Programming Operators:
* (NEW) Procedural operators let you build functional programs.
(PLUS)
* (NEW) Define local variables, nested arrays, branches, loops, and
recursive functions. (PLUS)
* User-defined functions using C, C++ programming interface.
(PLUS)
* (NEW) User-defined notation with live operators. (PLUS)
Communications Capabilities:
* (NEW) Send Mathcad worksheets to colleagues via e-mail systems
(VIM and MAPI) Supported systems include: cc:Mail, Microsoft
Mail, NotesMail.
* (NEW) Save and retrieve Mathcad worksheets from Lotus Notes
databases.
* (NEW) Link to worksheets on the World Wide Web or a local Web
server. Supports standard Microsoft compatible Internet access
method.
Authoring:
* (NEW) Create libraries of frequently used functions. (PLUS)
* (NEW) Create locked sections. (PLUS)
* (NEW) Define hyperlinks to Mathcad documents locally, in Lotus
Notes or on the World Wide Web. (PLUS)
* "Live" Mathcad equations, text and graphics with word processing
documents and spreadsheets.
Dirk Wright
: It would be very useful to get feedback from the members in this forum on
: that issue and other features that they would like to see in a Mathcad for
: Mac product. Also, from reading through the messages on this topic, it's
: obvious that MathSoft is going to have to do a lot to prove our commitment
: to Mac. Please let us know what you want in a Mathcad for Mac product and
: how we can prove to you that we are committed to the platform.
Another useful feature to draw people away for mathematica would enable
mathcad to read mathematica files. I assume mathcad doesn't already do this
(I haven't used it in awhile).
Dirk
Kirk A. Stork
>
>: It would be very useful to get feedback from the members in this forum on
>: that issue and other features that they would like to see in a Mathcad for
>: Mac product. Also, from reading through the messages on this topic, it's
>: obvious that MathSoft is going to have to do a lot to prove our commitment
>: to Mac. Please let us know what you want in a Mathcad for Mac product and
>: how we can prove to you that we are committed to the platform.
>
I haven't followed all of this thread, so please forgive me if I'm redundant.
Make all the MathCad modules that are available for windoze available on
the Mac. Make the Mac, Windoze, and Unix versions file compatable - make
the modules cross-platform compatable - I want to be able to send/receive
anything to/from a MathCad for Windows user and be able to use it without
modification.
Bring your other software to the Mac (S-Plus, Maple) in a serious way.
Support A/UX for 68K and also AIX for the Power Mac. Use the unique
features of the Mac - publish and subscribe, Applescript, Apple Guide,
Quickdraw GX, Quickdraw 3D. Make it easy to use Mac development tools to
make externals for your software, like Metrowerks C/C++, Symantec C++, and
MPW... Make it easy to use an external editor of the users choice (or
just Alpha - sorry, my own pref). Help us convince Apple to keep A/UX
alive - and to bring it to the PowerMac.
Most of all, take the time to read Apple's Human Interface standards, and
comply with them (okay, they are getting a bit dated, but they still have
good things to say).
Hire someone who believes in the Mac to be a part of the project - you
probably don't have one of those.
------------------------------------------------------------------------------
Opinions expresed here are my own, and do not represent those of the
United States Navy, Department of Defense, or the Government of the United
States.
+----------------------+----------------------------------------+
| Kirk A. Stork | Another day... |
| LT, USN | Another independent Bernoulli Trial... |
| Operations Research +----------------------------------------+
| Naval Postgraduate School, Monterey, CA. Class of 96 |
| email: kas...@nps.navy.mil |
| web: http://dubhe.cc.nps.navy.mil/~kastork/ |
| fax: (408) 647-1693 |
| SMC: 2783 |
+---------------------------------------------------------------+
Christopher Wright
(AARenshaw) wrote:
> MathSoft is going to have to do a lot to prove our commitment
> to Mac. Please let us know what you want in a Mathcad for Mac product and
> how we can prove to you that we are committed to the platform.
You've got a lot of nice features listed, which I gather exist on the
current Windows version. Very tempting. OTOH, you've jerked me around once
for real, and were damn patronizing when I asked some plain questions
about the product. I don't like being patronized and I still don't like
the guy who talked to me on the phone that day. Start by making a real
commitment to user service. Talk straight, keep promises.
Prove to me you're not putting out another lame DOS ported version by
sticking 100% to Mac human interface guidelines. Make it a real Mac
product. Make it scriptable. Support Open Doc. Act like a Mac isn't some
kind of troublesome anomaly. Then get back to me.
Eric Inazaki
Dirk Wright (wri...@uspto.gov) wrote:
: AARenshaw (aare...@aol.com) wrote:
: : It would be very useful to get feedback from the members in this forum on
: : that issue and other features that they would like to see in a Mathcad for
: : Mac product. Also, from reading through the messages on this topic, it's
: : obvious that MathSoft is going to have to do a lot to prove our commitment
: : to Mac. Please let us know what you want in a Mathcad for Mac product and
: : how we can prove to you that we are committed to the platform.
: Another useful feature to draw people away for mathematica would enable
: mathcad to read mathematica files. I assume mathcad doesn't already do this
: (I haven't used it in awhile).
: Dirk
Might not be a _bad_ idea. Keep in mind, though, MathCAD wasn't
designed to compete with the likes of Mathematica/Maple. It was
and, I feel, should be a quick, painless way to do "back of the
envelope" type calculations. It shouldn't cost as much as the
more heavy duty apps either. Having said that, the more features
the better.
The thing that bugs me about MCAD is that it doesn't deal with
continuous value functions very well. Seems everything needs
to be in the form of a matrix of one kind or another.
eric
Louis M. Pecora
A few days ago I told you that you had an uphill battle in bringing MathCad
back to the Mac. Here are some typical replies to your inquiry about what
Mac users wanted:
(1) In article <armDD2...@netcom.com>, a...@netcom.com (Alex Morando)
wrote:
[cut]
> The only thing MathSoft can do to regain their credibility among non-Wintel
> technical users is to create the best version of Mathcad ever. As MathSoft
> probably knows with their 3.1 experience - "PORT" is a four-letter word.
>
> I don't think they can pull it off, but I hope I'm wrong.
(2) In article <kastork-0908...@slippc5.cs.nps.navy.mil>,
kas...@nps.navy.mil (Kirk A. Stork) wrote:
[cut]
> Most of all, take the time to read Apple's Human Interface standards, and
> comply with them (okay, they are getting a bit dated, but they still have
> good things to say).
>
> Hire someone who believes in the Mac to be a part of the project - you
> probably don't have one of those.
(3) In article <chrisw-0908...@pm4-118.skypoint.net>,
chr...@skypoint.com (Christopher Wright) wrote:
[cut]
> Prove to me you're not putting out another lame DOS ported version by
> sticking 100% to Mac human interface guidelines. Make it a real Mac
> product. Make it scriptable. Support Open Doc. Act like a Mac isn't some
> kind of troublesome anomaly. Then get back to me.
You've probably seen many more like this in the mean time. I am not posting
this to gloat, but to ask you, in light of these sentiments, how does
MathSoft expect to succeed?
I mostly agree with Mr. Morando above. I hope you can succeed. Most Mac
users deep down would agree. We would like to see good software on the Mac.
A loser really benefits no one, especially users. But these aren't
soreheads, MathSoft dug their own hole. I'm just wondering how you think
they can get out of it?
K. John Timm
>
> MathSoft is coming out with Mathcad 6.0 for Mac early next year. I have
> been hired to help them in defining the needs of this market.
>
> It would be very useful to get feedback from the members in this forum on
> that issue and other features that they would like to see in a Mathcad for
> Mac product.
Well this is good news and about time too. Like some others I am somewhat
bitter about the way Mathsoft dropped us Mac users. Yes, I do understand
that the Mac community accounts for only a small portion of their
business, but still.......So I will put my Mathematica manual and intent
to examine Maple on hold for a while.
Talking about features, it sure would be nice to see a split window
capability in Mathcad. This way one could look at the top of the
worksheet, where the input information is normally entered, and the bottom
where the answers usually appear.
Alternatively, and this is more of a bandaid, allow the document to opened
twice, each could then be scrolled appropriately to see start and finish.
*******************
John (Dallas TX)
tur...@metronet.com
*******************
David B. Joyce
> It would be very useful to get feedback from the members in this forum
on
> that issue and other features that they would like to see in a Mathcad
for
> Mac product. Also, from reading through the messages on this topic,
it's
> obvious that MathSoft is going to have to do a lot to prove our
commitment
> to Mac. Please let us know what you want in a Mathcad for Mac product
and
> how we can prove to you that we are committed to the platform.
>
> FYI, below is a list of some of the features in the released version of
> Mathcad 6.0.
>
> Thanks in advance for your feedback.
> Amy
>(Followed by numerous posts from people who are apparently really good at
holding a grudge.)
I for one am very happy to hear that MathSoft is returning to the Mac
Market. Much of the criticisms I've read are valid, but I don't
understand the level of animosity. My experiences with customer support
at Mathsoft were positive and prompt. I still use MathCad whenever I can,
and Mathematica when I absolutely have to.
A couple of quick suggestions:
* By all means follow the mac interface guidelines, *especially* for input
and output.
* Find the Macintosh toolbox, and use it. You should never have to use
alert boxes that say "Building dialong box".
I look forward to MathCad's new release, and hope I will be able to
recommend it even more highly than 3.x
Alex Morando
David B. Joyce <d...@blencoe2.chem.ornl.gov> wrote:
>>
>>(Followed by numerous posts from people who are apparently really good at
>holding a grudge.)
>
I probably wouldn't mind getting a copy of the next version of
Mathcad/Mac for personal use.
>Market. Much of the criticisms I've read are valid, but I don't
>understand the level of animosity. My experiences with customer support
>at Mathsoft were positive and prompt. I still use MathCad whenever I can,
>and Mathematica when I absolutely have to.
However, I am responsible for making sure 200 engineers have
the right tools. When MathSoft abandoned the Mac, we started
having problems as the PC users started to use 4.0 and couldn't
exchange files with the Macs. Needless to say, my job was on the
line. We dumped Mathcad rather than half our computers.
You can probably see why my company would be hesitant to get
back into Mathcad.
Michael Kahlow
"MathSoft is coming out with Mathcad 6.0 for Mac early next year. I have
been hired to help them in defining the needs of this market. It was
interesting that you said you'd heard we are only going to have a Power
Mac version. That decision has not been made, although right now we are
leaning towards developing for both Power Mac and 680x0-based Macs.
It would be very useful to get feedback from the members in this forum on
that issue and other features that they would like to see in a Mathcad for
Mac product. Also, from reading through the messages on this topic, it's
obvious that MathSoft is going to have to do a lot to prove our commitment
to Mac. Please let us know what you want in a Mathcad for Mac product and
how we can prove to you that we are committed to the platform."
A reply:
>
> Prove to me you're not putting out another lame DOS ported version by
> sticking 100% to Mac human interface guidelines. Make it a real Mac
> product. Make it scriptable. Support Open Doc. Act like a Mac isn't some
> kind of troublesome anomaly. Then get back to me.
>
Alelluia and amen. Putting it into non-Mac terms:
All Mac applications behave in a similar fashion. The mouse does the same
thing, key combinations do the same thing, the menus are arranged in a
similar fashion, etc. Special cntl-alt-shift-whatsits combos are
verbotten (one particularly onerous feature of prior MathCad for Mac).
Because the previous Mac version did not obey these rules,my of MathCad
spent exactly three days on my hard drive before deletion. Worst $100 I
ever spent.
I have a colleague who is a fan of MathCad and Windows/DOS from the
earliest days. I have had Mathematica on my Mac since version 1.1.
Neither of us will budge. If you who make MathCad listen to all of the
people's comments on this list and actually make a Mac MathCad (instead of
a dreadful Windows port), I can guarantee you'll sell at least 25 copies a
year here. Because we'll make our students buy it with their texts.
michael kahlow
chemistry department
university of wisconsin-river falls
michael....@uwrf.edu
Gerald Recktenwald
wrote:
>MathSoft is coming out with Mathcad 6.0 for Mac early next year. I have
>been hired to help them in defining the needs of this market. It was
>interesting that you said you'd heard we are only going to have a Power
>Mac version. That decision has not been made, ...
[edits]
> ... Please let us know what you want in a Mathcad for Mac product and
>how we can prove to you that we are committed to the platform.
Your hesitation suggests that you don't understand the Mac market
at all. Don't even think twice about bringing out a PowerMac version!
I'd go so far as to develop the PowerMac version *first*, and make
it clean and fast for the PPC chip. Then worry about 68k Mac
compatibility and performance. Yes, there is a large installed
base of 68k Macs, but they are not the future.
If, like that software monolith (gee, the name escapes me, but it's
located somewhere in Warshington, the state, you know), you put out a
bloated, and inefficient PPC version you will squander any interest
you may have among Mac users.
>FYI, below is a list of some of the features in the released version of
[edits]
>Programming Operators:
>
> * (NEW) Procedural operators let you build functional programs.
> (PLUS)
> * (NEW) Define local variables, nested arrays, branches, loops, and
> recursive functions. (PLUS)
> * User-defined functions using C, C++ programming interface.
> (PLUS)
> * (NEW) User-defined notation with live operators. (PLUS)
These are all very good ideas. Though others like the notebook
metaphor, I was always annoyed that the results depend on where
the formulas are located on the page. IMHO, only a spreadsheet addict
could desire such hidden dependencies in a calculation tool.
If fact, because of the lack of real programmability and other
interface clumsiness I took my registered version of MathCAD 3.1
off my hard disk in favor of some real analytical tools.
>Communications Capabilities:
>
[edits]
> * (NEW) Link to worksheets on the World Wide Web or a local Web
> server. Supports standard Microsoft compatible Internet access
> method.
What is a "Microsoft compatible" anything?! Microsoft compatibility
should be a subset of compatiblity with the Web or any other internet
standard (such as they exist). Though it controls the largest market
share of desktop OS, Microsoft is a special case. If I can read between
the lines here it seems that you are proposing to support MS Network.
Fine. Just don't call that a standard.
You may want to refer to the small article at the bottom of p. 10 in
the Aug. 14 edition of InformationWeek. There, the CEO of UUNET,
the company that is providing backbone service to MS Network says,
"The network we're building can handle only 1 million users in the first
month, and only 2 million users by year's end--if every single thing
goes right." It sounds to me that you, and your potential customers,
would be farther along if you provided support for standard web access
and tools (CGI's and the like).
+-+-+-+-+-+-+-+-...
Gerald Recktenwald (503) 725-4296
ge...@me.pdx.edu http://www.me.pdx.edu/~gerry/
PSU Mechanical Engineering, P.O. Box 751, Portland, OR 97207
The lottery: an investment strategy for the statistically impaired
eric larson
> sticking 100% to Mac human interface guidelines. Make it a real Mac
> product. Make it scriptable. Support Open Doc. Act like a Mac isn't some
> kind of troublesome anomaly. Then get back to me.
I saw today that Maple is going to support Open Doc. To me that means MathCAD
is going to have to be one better, if possible.
Kirk A. Stork
>>
> [edits]
>> * (NEW) Link to worksheets on the World Wide Web or a local Web
>> server. Supports standard Microsoft compatible Internet access
>> method.
>
What the heck is "Microsoft compatible" internet access? The Mac has
(arguably) better internet tools than any platform, and Microsoft has
nothing to do with it. Windows 95 is the LAST OS on the market to make
internet access easy. Microsoft doesn't own the Net, they have no
standard to push. If you are going to make a product internet aware - do
it using the excellent tools already available in the Mac OS, or from the
many sainted shareware authors that have brought us this fine set of
tools.
Eric Moyer
(David B. Joyce) wrote:
> > It would be very useful to get feedback from the members in this forum
> on
> > that issue and other features that they would like to see in a Mathcad
> for
> > Mac product. Also, from reading through the messages on this topic,
> it's
> > obvious that MathSoft is going to have to do a lot to prove our
> commitment
> > to Mac. Please let us know what you want in a Mathcad for Mac product
> and
> > how we can prove to you that we are committed to the platform.
> >
> > FYI, below is a list of some of the features in the released version of
> > Mathcad 6.0.
I didn't see the list of features, so I don't know if this is planned:
* I need more powerful symbolic math features (but I don't want to pay for
the full implementation of Maple)
Lou Pecora
76331...@compuserve.com (Eric Moyer) wrote:
[cut former message]
>
> I didn't see the list of features, so I don't know if this is planned:
> * I need more powerful symbolic math features (but I don't want to pay for
> the full implementation of Maple)
Symbolic math programs may be expensive, but symbolic math is orders of
magnitude harder to program than numerical algorithms. My guess is that
things like Maple and Mathematica are a bit overpriced, but only a bit.
--
Lou Pecora
code 6341
Naval Research Lab
Washington DC 20375
USA
pec...@zoltar.nrl.navy.mil
/* My views are not those of the U.S. Navy.
If you want their views, you have to go to war. */
Jeff Tupper
pec...@zoltar.nrl.navy.mil (Lou Pecora) wrote:
>In article <76331.3366-15...@dd24-007.compuserve.com>,
>76331...@compuserve.com (Eric Moyer) wrote:
>
>[cut former message]
>>
>> I didn't see the list of features, so I don't know if this is planned:
>> * I need more powerful symbolic math features (but I don't want to pay for
>> the full implementation of Maple)
>
>Symbolic math programs may be expensive, but symbolic math is orders of
>magnitude harder to program than numerical algorithms. My guess is that
>things like Maple and Mathematica are a bit overpriced, but only a bit.
>
I have no idea why you think this. Can you defend this in any way? There are
"hard" algorithms for both numerical and symbolic computation. There isn't
even a clear dividing line between numerical and symbolic computation.
jeff
Louis M. Pecora
moon...@cs.utoronto.ca (Jeff Tupper) wrote:
> In article <pecora-18...@esp225.nrl.navy.mil>,
> pec...@zoltar.nrl.navy.mil (Lou Pecora) wrote:
> >In article <76331.3366-15...@dd24-007.compuserve.com>,
> >76331...@compuserve.com (Eric Moyer) wrote:
> >
> >[cut former message]
> >>
> >> I didn't see the list of features, so I don't know if this is planned:
> >> * I need more powerful symbolic math features (but I don't want to pay for
> >> the full implementation of Maple)
> >
> >Symbolic math programs may be expensive, but symbolic math is orders of
> >magnitude harder to program than numerical algorithms. My guess is that
> >things like Maple and Mathematica are a bit overpriced, but only a bit.
> >
>
> I have no idea why you think this. Can you defend this in any way?
The short answer (since you are making the complaint) is to go look up the
literature on symbolic algorithms and their implementation. Don't depend on
my defense. Go to the source. Compare to numerical methods and judge for
yourself. You'll see what I'm saying is true.
But I will address it without getting into a full-blown debate, just
consider integration. Schemes for numerical integration have been around
for literally centuries. They're very direct--almost geometrical. We can
easily implement them on any computer with floating point math. Algorithms
for symbolic integration of elementary functions (sine, log, etc.) (the
analogs to symbolic differentiation) were not finished until the 1960's,
when attempts to do symbolic computation showed vast shortcomings there.
The algorithms (unlike their derivative counterparts) are not simple.
Just stop and think about it for a moment. You can always "simplify"
numerical calculations by constantly reducing each part to a number (I'm
speaking loosly here). Symbolic manipulation requires much more
sophisticated and complex pattern matching to replace one set of symbols by
another. You're brain is good at it, but algorithms to do it are hard. Try
it. Write an alorithm to numerically solve a quadratic equation, then write
one to symbolically solve it and simplify the result. You can do it in one
line numerically, but require several tests for symbolic simplification and
don't forget such issues as number representation.
Symbolic math requires keeping track of lots of information and
*efficiently* matching patterns. That's lots harder than multiplying.
>There are
> "hard" algorithms for both numerical and symbolic computation. There isn't
> even a clear dividing line between numerical and symbolic computation.
The concepts behind any algorithm can be hard. That's not the point. The
point is that most of the numerical operations people would need in a
general package are not hard to implement. Most of the symbolic ones are.
It's not for nothing that all symbolic manipulation programs that even
attempt to be general are large (Macsyma, Mathematica, Maple, ScratchPad,
etc.).
You just need to think about it a little more. Programming a computer to
calculate exp(-3.5) is easy. Programming it to symbolically handle the
exponential function is a lot harder.
Anyway, I said more than I planned to.
Jeff Tupper
pec...@zoltar.nrl.navy.mil (Louis M. Pecora) wrote:
>In article <1995Aug18.0...@jarvis.cs.toronto.edu>,
>moon...@cs.utoronto.ca (Jeff Tupper) wrote:
>
>
>But I will address it without getting into a full-blown debate, just
>consider integration. Schemes for numerical integration have been around
>for literally centuries. They're very direct--almost geometrical. We can
Well, I am not sure how old the algorithms are has to do with complexity
of modern algorithms, but I believe that symbolic schemes have been
around as long. People had to rely on symbolic schemes because high
speed computers were not around. (some numeric schemes for integration
existed before "integration" did, but if you really consider them
"integration" schemes then they should be around after Newton/Liebnitz [sp?] )
>easily implement them on any computer with floating point math. Algorithms
>for symbolic integration of elementary functions (sine, log, etc.) (the
>analogs to symbolic differentiation) were not finished until the 1960's,
>when attempts to do symbolic computation showed vast shortcomings there.
>The algorithms (unlike their derivative counterparts) are not simple.
Well, a primitive numerical algorithm is simple. So what? A primitive
symbolic alogirithm is simple.
>Just stop and think about it for a moment. You can always "simplify"
>numerical calculations by constantly reducing each part to a number (I'm
>speaking loosly here). Symbolic manipulation requires much more
>sophisticated and complex pattern matching to replace one set of symbols by
>another. You're brain is good at it, but algorithms to do it are hard. Try
Just stop and think about it for a moment. You can always "simplify" a
symbolic algorithm by saying "I won't handle this case". A proper
numerical algorithm will give good results - hacking things to constants
which aren't is usually not a good idea.
>it. Write an alorithm to numerically solve a quadratic equation, then write
>one to symbolically solve it and simplify the result. You can do it in one
>line numerically, but require several tests for symbolic simplification and
>don't forget such issues as number representation.
Well, I have in the past :) Not sure about the comment regarding one line
numerically - how about guaranteed error bounds? Numerical computation isn't
that great without at least an estimation for the error.
>Symbolic math requires keeping track of lots of information and
>*efficiently* matching patterns. That's lots harder than multiplying.
It may be harder because a library is not available. Ever tried implementing a
modern hardware multiply part for an ALU?
>>There are
>> "hard" algorithms for both numerical and symbolic computation. There isn't
>> even a clear dividing line between numerical and symbolic computation.
>
>The concepts behind any algorithm can be hard. That's not the point. The
>point is that most of the numerical operations people would need in a
>general package are not hard to implement. Most of the symbolic ones are.
Actually, the concepts behind some numerical algorithms are quite simple,
unfortunately very tough to program. My research is on generalized interval
arithmetic - some things are very easy to prove/understand as to what the
algorithm is doing at a high-level, but quite difficult to implement.
For example, CIA division is a couple hundred lines [and still not
optimal] in most implementations. LIA division is several times as
complicated. QIA is again several times as complicated.
>It's not for nothing that all symbolic manipulation programs that even
>attempt to be general are large (Macsyma, Mathematica, Maple, ScratchPad,
>etc.).
Ever seen a generalized interval arithmetic package? Not small. Even for doing
things as basic as just evaluating functions.
>You just need to think about it a little more. Programming a computer to
>calculate exp(-3.5) is easy. Programming it to symbolically handle the
>exponential function is a lot harder.
Have you written a fast, arbitrary precision IEEE754 compliant numerical
library? Care to post the code for exponentiation? Storing "exp(-3.5)" is easy
in a symbolic package...
>Anyway, I said more than I planned to.
Still looking for some reasons as to why "symbolic hard, numeric easy"...
Both can be easy or difficult, it all depends on what you expect from your
algorithms (how robust, efficient, general they are).
Jeff
Eric Moyer
> Symbolic math programs may be expensive, but symbolic math is orders of
> magnitude harder to program than numerical algorithms. My guess is that
> things like Maple and Mathematica are a bit overpriced, but only a bit.
I fully understand and agree... I wasn't trying to say that software like
Maple is overpriced, just more expensive than what I'm willing to pay. I
would be happy to find a go-between, something with more symbolic power
than MathCAD, but not necessarily as much as Maple, and for a medium price.
--
Copyright, Eric Moyer, 1995. Microsoft Network is prohibited from
redistributing this work in any form, in whole or in part. License to
distribute this post is available to Microsoft for $1000 US. Posting
without permission constitutes an agreement to these terms. Please
send notices of violation to <er...@engsys.com> and
<postm...@microsoft.com>.