Is Mathcad Dying?

[ptc-1886369]

Hi all

I have just written a blog post with the above title and would value the opinion of regular Mathcad users such as you.

http://www.walkingrandomly.com/?p=1425

Cheers,
Mike

[StuartBruff]

On 6/16/2009 7:59:40 AM, MikeCroucher wrote:
== I have just written a blog post with the above title and would value the opinion of regular Mathcad users such as you.

If it carries on like it has been recently, then yes. Too little benefit for the support costs and a product that just hasn't kept up with the times (in fact, it should be ahead of the times!)

Mathcad is already regarded as 'light-weight' by comparison to other products, particularly Matlab in the engineering world - however, the core of this criticism is unjustified IMO. Mathcad is capable of doing much of the work that Matlab is used for and with a quicker overall turnaround plus the benefit of a self-documenting work-product.

But there may be light at the end of the tunnel and it may not be an express train. I believe (only from a PTC conference poster), that PTC intend to release a product called Mathcad Spirit that may, perhaps, represent the major step forward that some of us have long asked for. I have no idea what Spirit actually is, and given some of the capability decisions that have happened in the past (eg, SUC and recursive-depth limiting), I'm not getting my hopes up. For Mathcad's sake, I will, however, keep my fingers crossed.

Stuart

See the bit entitled Roadmap in http://www.ptcuser.org/2009/mathcad.html

[ptc-1886369]

Thanks for the info Stuart. Do you mind if I take the liberty of copying and pasting your comment into the comments section of the blog please?

Of course, feel free to comment there yourself if you wish (if you don't want me to have your email address, just use a fake one).

Cheers,
Mike

[RichardJ]

A few comments:

1) In answer to your statement "personally I cannot see why anyone who can also choose from Mathematica, MATLAB and Maple (and I am lucky enough to be in this position) would ever bother with it", given your access to all the packages I'm (really) surprised that's not obvious. It's the interface. I can print a Mathcad document to pdf and send it to someone and they can understand what I did even if they have never seen Mathcad in their life. I can have a diagram, some formatted explanatory text, some math written in standard notation, a table of results, etc, all laid out on a page. The same is not true for any of the other packages. Maybe that's of no importance or use to you, but it is to a lot of people. Until there is a competing package with such an interface, and math capability that at least approaches that of Mathcad, Mathcad will always have a market.

2) You seem to conclude from the fact that PTC has not released a major update in the last couple of years that they are not working on it. That's not a valid conclusion. Unless you are privy to PTCs internal development plans (which you are obviously not, since you didn't even know about aspects of those plans that have been made public) you only know that either they are not working on it, or they are putting all the resources into something major that they are not talking about. Since they only recently purchased an entire company to get Mathcad as a product, the first option seems by far the least likely of the two. Especially given that they are in fact talking about something (even if with limited details).

3). Mupad as the symbolic engine. Yes, compared with Maple it's a poor option. I noticed that you have a couple of "Mathcad" symbolic bugs posted on your site. Actually, they are Mupad bugs, and if you use Matlab a lot you are going to get to enjoy them too! Like the Mathcad users, you will look longingly back on the days when you had Maple as the symbolic engine. I think that should also address your post on whether or not you should care about the switch in Matlab. Yes, you should. A lot.

It will indeed be interesting to see what PTC does about the symbolic engine in Mathcad now that Mathworks has purchased Mupad, but any change away from Mupad couldn't be all bad.

As an aside, regarding your post

http://www.walkingrandomly.com/?p=33

the polylog function is in the Mathcad help. Try searching for the word "polylog" (a fairly obvious choice, I would have thought). The help points you to a quicksheet that defines the function, and points out that polylog(2,x) is equal to dilog(1-x) (not dilog(x), as you have on your website).

As a final comment, having looked at your website I think the statement "the writer is heavily biased" is something of an understatement. Perhaps if you looked more objectively at the merits of the different products you would realize that they all have their pluses and minuses. What is "best" depends on what you are trying to do, and what the end goals are. The fact that a hammer doesn't do a good job of putting in screws doesn't make it a bad tool.

Richard

[ptc-1368288]

>As a final comment, having looked at your website I think the statement "the writer is heavily biased" is something of an understatement<<br> ______________________________

In my reply, just removed a shorter statement because I was about sure Mona wouldn't have like it, but clearly the writer is confused with:

1. maths are scalar.
2. numerical maths.
3. knows nothing or near nothing about Mathcad.

I have a deeper question to myself: how much is the writer paid to detract what he knows nothing about. Detracting Mathcad is detracting > 2.5 millions users, all assholes ?

jmG

[ptc-1886369]

Hi

I am sorry if I offended - I certainly didn't mean to. Regarding the comment

'how much is the writer paid to detract what he knows nothing about. Detracting Mathcad is detracting > 2.5 millions users, all assholes?'

I am not paid anything for the blog (In fact it costs me money) and I am not on the payroll of any math software company. Instead I represent a customer of many of them (PTC, Wolfram and Mathworks included). I don't use Mathcad myself very much but I am often called upon to help deal with its bugs - The print preview bug for example, the fact that it couldn't symbolically evaluate 2^31 and so on.

I am not suggesting that the user of any application is an asshole. I am asking for opinions.

Regarding the comment

' It's the interface. I can print a Mathcad document to pdf and send it to someone and they can understand what I did even if they have never seen Mathcad in their life.'

You can do the same in Mathematica and Maple.

Your second point is very valid:

'You seem to conclude from the fact that PTC has not released a major update in the last couple of years that they are not working on it.'

You are, of course, probably right. I really didn't want to offend anyone here - I just wanted to open a discussion and maybe get an inkling on what that work might be.

Best wishes,
Mike

[RichardJ]

On 6/16/2009 1:48:14 PM, MikeCroucher wrote:

>Regarding the comment
>
>' It's the interface. I can
>print a Mathcad document to
>pdf and send it to someone and
>they can understand what I did
>even if they have never seen
>Mathcad in their life.'
>
>You can do the same in
>Mathematica and Maple.
>

I have never seen such a page produced by either Mathematica or Maple. Show me a screen capture with an annotated diagram (it doesn't matter what; a bullet trajectory, a mechanical part, a picture of Winne the Pooh, whatever), a paragraph of explanatory text next to it (preferably formatted), a function definition in normal math notation (to include more than just basic operators, let's say an integral), an evaluation of that function placed to the right of the function, and a graph of the function below it.

Richard

[Ninetrees]

A Mathcad user since the mid-1980s (if I recall correctly), for the past few years, I have been increasingly dissatisfied with PTC's handling of bug releases and their support costs. My impression has been of a product that was nearing EOL, notwithstanding the comment that PTC had just purchased it.

I use Mathcad now for the same reason that I began using it so many years ago: its self documenting capabilities.

I use MATLAB and Maple regularly, partly because I must, and partly because they provide different benefits from Mathcad.

Maple is a very strong player in many ways, but V13 just released doesn't address its shortcomings compared to Mathcad. I regularly use MATLAB because it is so easy to code in, and its vector and matrix handling capabilities are superb.

But MATLAB is not a documenting tool, and even though it provides traceability via scripts, they are simply more difficult to read. The scripts are not meant to be read as an engineering document. I recently was approached by a coworker who needed help preparing a document. He was using data and plots from MATLAB, and writing in Latex. He could have done the whole thing much easier in Mathcad and been done with it.

Several coworkers have seen some of the documents that I've produced in Mathcad, and, thinking that I did them in MATLAB, have labeled a whiz with MATLAB. Others have looked over my shoulder as I developed work in Mathcad, and were impressed with the speed and effectiveness of the Mathcad interface.

I've made a substantial effort to move to Maple, but Maple, despite its significant advantage in a powerful UI, has some significant drawbacks in its documenting capabilities. The latest blurbs from Maple tout "advanced users can use the tutorials to learn about document creation," but Ricard's example above is a good challenge that they don't meet, or at least not easily. Formatting a document "your way," not "Maple's way," is difficult. Maple still insists that units come in these funky brackets [[]], and its 4 different math entry modes can be confusing.

OTOH, Maple's ease of working with equations and plotting, for example, irrespective of my dissatisfaction with the finished product, are superior to Mathcad's, and Maple's right-click and find a tool to help solve a problem leave Mathcad in the dust.

In the end, I want to do the work just once, and this is where Mathcad has been historically the winner. If I need a fast programming tool, MATLAB beats Mathcad IMO. If I want intuitive ODE solvers that recognize that I am using dot notation for time drivatives, then I am OOL with Mathcad, but in luck with Maple.

My Maple productivity suffers, as one might expect, from may lack of familiarity with it, and from my familiarity with Mathcad; I frequently use the wrong keystrokes in Maple. But that would be true in reverse were I a long-time Maple user. I recently had to do some work with transfer matrices, and it was far easier to do with Mathcad, and clear for others to see what I did and why.

Rich
http://www.downeastengineering.com/

[Ninetrees]

Here are a few examples...

Rich
http://www.downeastengineering.com/

[ptc-1368288]

Pretty neat, lovely but nothing to make Mathcad envious. The 2nd *.PDF, such a huge graph deserves some grid lines.

jmG

[RichardJ]

Thanks. They are good examples, especially the tutorial one. For someone like me, who does not have Maple, it gives a much better idea of what can and can't be done.

Richard

[mikej1980]

On 6/16/2009 2:04:47 PM, rijackson wrote:
>On 6/16/2009 1:48:14 PM, MikeCroucher
>wrote:
>
>>Regarding the comment
>>
>>' It's the interface. I can
>>print a Mathcad document to
>>pdf and send it to someone and
>>they can understand what I did
>>even if they have never seen
>>Mathcad in their life.'
>>
>>You can do the same in
>>Mathematica and Maple.
>>
>
>I have never seen such a page produced
>by either Mathematica or Maple. Show me
>a screen capture with an annotated
>diagram (it doesn't matter what; a
>bullet trajectory, a mechanical part, a
>picture of Winne the Pooh, whatever), a
>paragraph of explanatory text next to it
>(preferably formatted), a function
>definition in normal math notation (to
>include more than just basic operators,
>let's say an integral), an evaluation of
>that function placed to the right of the
>function, and a graph of the function
>below it.
>
>Richard



I have access to mathematica and maple and since someone already posted a maple document I thought I'd post a mathematica document.

FWIW the two things that made me want to try out mathcad were units and document capabilities (I understood these to be the strengths relative to other programs).

IMO when it comes to creating documents:

mathematica >> maple > mathcad

when it comes to handling units

mathcad > maple >> mathematica

[mikej1980]

On 6/16/2009 2:04:47 PM, rijackson wrote:
>On 6/16/2009 1:48:14 PM, MikeCroucher
>wrote:
>
>>Regarding the comment
>>
>>' It's the interface. I can
>>print a Mathcad document to
>>pdf and send it to someone and
>>they can understand what I did
>>even if they have never seen
>>Mathcad in their life.'
>>
>>You can do the same in
>>Mathematica and Maple.
>>
>
>I have never seen such a page produced
>by either Mathematica or Maple. Show me
>a screen capture with an annotated
>diagram (it doesn't matter what; a
>bullet trajectory, a mechanical part, a
>picture of Winne the Pooh, whatever), a
>paragraph of explanatory text next to it
>(preferably formatted), a function
>definition in normal math notation (to
>include more than just basic operators,
>let's say an integral), an evaluation of
>that function placed to the right of the
>function, and a graph of the function
>below it.
>
>Richard



I have access to mathematica and maple and since someone already posted a maple document I thought I'd post a mathematica document.

FWIW the two things that made me want to try out mathcad were units and document capabilities (I understood these to be the strengths relative to other programs).

IMO when it comes to creating documents:

mathematica >> maple > mathcad

when it comes to handling units

mathcad > maple > mathematica

[ptc-1368288]

On 8/12/2009 1:22:26 PM, mikej1980 wrote:
>On 6/16/2009 2:04:47 PM, rijackson
>wrote:
>>On 6/16/2009 1:48:14 PM, MikeCroucher
.........
>IMO when it comes to creating documents:

==> Mathcad beats them all [jmG]

[mikej1980]

On 8/12/2009 3:35:06 PM, jmG wrote:
>On 8/12/2009 1:22:26 PM, mikej1980
>wrote:
>>On 6/16/2009 2:04:47 PM, rijackson
>>wrote:
>>>On 6/16/2009 1:48:14 PM, MikeCroucher
>.........
>>IMO when it comes to creating documents:
>
>==> Mathcad beats them all [jmG]
>

Is the plot supposed to be proof of that? ...and surely a document is more than just a plot? Most people seem to talk about mixing graphics, code, math typesetting, text etc.

Anyway here is a mathematica version of your plot. Since I don't have your data and don't know this field it is the best I can do. Looks better though.

[ptc-1368288]

On 8/12/2009 5:22:37 PM, mikej1980 wrote:
....
>Anyway here is a Mathematica version of
>your plot... Looks better though.
____________________

Your Matematica plot looks better than mine (Mathcad), let people appreciate.



I would be much surprised if my Plot32 (attached) wouldn't transmit correctly. No collab have ever mentioned problem with my images, rather the opposite, i.e: of superb quality. That point is important if your system is not properly tuned.

About documentation, take a universal example : the Breather. Make a picture of the Mathematica work sheet, not *.PDF. Or, if you make a *.PDF of Mathematica, make same of the Mathcad. Show all the steps in Mathematica. Show the timing, that's for my curiosity if your recent Mathematica version is still "reputed slow". It is known in advance that Mathematica can produce images of very high quality, but not running under current PC's capabilities.

jmG

[ptc-1368288]

Oh ! you have just learned to ask about how to do in Mathcad c/w a work sheet attached. Engineers learn after the academic passage and as they are responsible to the public, that's why they need a traceable tool like Mathcad. They calculate, they model ... and hand superb documents in universal notations, not like Mathematica 4.0 from which it was not possible to get copy/paste a data table nowhere else !

Mathcad 11 and earlier versions get



What do you expect the answer should be ?
The approximation to 2^ or the continued fraction ?
What do you get from Matlab, Mathematica ?

jmG

[ptc-1368288]

On 6/16/2009 7:59:40 AM, MikeCroucher wrote:
>Hi all
>
>I have just written a blog
>post with the above title and
>would value the opinion of
>regular Mathcad users such as
>you.
>
>http://www.walkingrandomly.com
>/?p=1425
>
>Cheers,
>Mike
_______________________________

You can only compare products by their final utility and efficiency to users. Ignoring Mathematica 4.0 as "academic" and as such useless for Engineers. Mathcad is a "super mathematical scalar structure", the word is scalar. As long as PTC will keep it as such with the simple programming structure and a valid symbolic, you can consider Mathcad as a tool for Engineers and Scientist easily manageable with nearly limitless possibilities and the best document producer.
The best of Mathcad is this forum. It is a mine, full of expertise to quick "project done by users", especially for data fitting and symbolic (Maple versions). You tend to glorify so many new versions of Mathematica, Matlab ... to me it rather proves the original product was incorrectly structured. The attached sheet contains 8 exercises for you to code and do in your preferred one. There is one thing for sure with Mathcad: It has not been exhausted yet (reasonably) nor scratched. It has a limited number of "built-in", but that was not your point because Mathcad is infinitely "built-in".

Mathcad will RIP if PTC kills it but soon an equivalent will get birth, because Engineers need such a gorgeous quick and productive tool of that extraordinary high quality documentation. If you can't do any of the attached exercises within a reasonable amount of hours, then question yourself and question the product you use.

jmG

PS: the last plot [Map] is truncated because of the 2 MB limit.