Mathcad Prime 5.0: some new updates

[VladimirN]

I decided to start a new topic about the future version 5.0. Based on the new timeline (was dated February 2017), the release is scheduled for early 2018 year. And there is nothing new about the planned features of the 5.0 release.

Old thread "PTC Mathcad Prime Roadmap (2015/2016 update)": https://community.ptc.com/t5/PTC-Mathcad-Blog/PTC-Mathcad-Prime-Roadmap-2015-2016-update/ba-p/446448

[Fred_Kohlhepp · Accepted solution]

The steadily increasing time between releases (up to 4.0) suggests that the "early 2018" estimate for Prime 5.0 is management wishful thinking.

[sflamant]

I just keep thinking about what would have happened if PTC had spent their effort continuing on mathcad 16 instead of going towards Prime, which was a useless effort ... What a bad decision from some managers at PTC !

I never looked at Prime, as people were saying that it was not good, and light years from Mathcad 15... But now that there was Prime 4.0 , I wanted to compare with Mathcad 15 that I know well... so I did that today.

First I converted a mathcad "capability" file I had for mathcad 15 with all kinds of features to Prime 4.0.

First conclusion : the converter is quite bad, and basically useless for the numerous complex files I have written in mathcad15 through the years... I had to copy and paste pieces of the converted file to see what was working and what was not...

1) The new "look" of mathcad :  I don't know what people think, but mathcad 15 looks more professional to me than Prime, and the rubban don't seem very practical compared to the menus of mathcad15 that seem more intuitive to me, easier and faster to use.

2) The speed ?  On the same computer , large loops with a log computation in the loop are now almost twice slower with prime 4.0 than they were with Mathcad 15 ! On the other hand, large matrices computation seem to have improved in speed by a factor of 3. So speed might be a wash, which is not good. You want newer software to be always faster not slower ...

3) Matlab component is not supported ( but it was not working well in mathcad15), components like the slider and the buttons are not supported.

4) The symbolic equation solver is not supported yet.

5) Some of my 3D graphs ( scatter plot for example) need to be substituted for contour plots that are not as good... The picture operator needed also to be replaced with contour plots in greyscale, for which the greyscale granularity is not good enough and cannot be modified. Certainly not impressed with the plots .

6) The hyperlink capability that was allowing me to jump quickly in large files doesn't seem to be supported.

7)  ... and where are the e-books, and the quick links ?!  They were great, you could just drag and paste the examples in a split second... For example if I need to look at some DSP, I see that some of the DSP examples that used to be in the e-books are now on the net , but the drag and paste doesn't work, and the copy and paste doesn't seem to work well either... One of the great idea of mathcad ( the e-book) is gone !

😎 I stopped looking at Prime 4.0 before investigating any further to see if maybe, Prime had fixed some of the memory leaks of mathcad 15... Maybe Prime 4.0 is better than mathcad 15 in that area, but because of the Prime 4.0 numerous shortcomings I will never know... We will see if Prime 5.0 or 6.0 one day come back to the level of mathcad 15... It looks to me like 10 years of development have been lost !

To me, a professional tool that was mathcad 15, was replaced by a product Prime 4.0 ,  that looks "unfinished" and not really usable , what a pity ! 

[Fred_Kohlhepp]

Your note 4) is (I believe) incorrect; Prime does support symbolics.

With that exception, I think you're "spot on."  Which is why there are so many posts in this forum discussing what other options there might be . . .

[RichardJ]

Symbolics in Prime are one of the few things that are in fact almost the same as Mathcad 15.

[sflamant]

I don't think that all the symbolic functionality that we have in mathcad 15 are supported in Prime 4.0:

When I converted some of my files, I encountered a few messages telling me that my symbolic operation was not supported : for example:

Given

x + a*y = 2

x+3*z = 4

x^2 +a*z =2

Find (x,y,z) ->

Gives  3 pairs of responses that are functions of the parameter a , when using mathcad 15

but with Prime 4.0 I get no response and the following message instead : "The symbolic evaluation of solving functions is not supported in the current version of PTC Mathcad" .   And this is with Prime 4.0 loaded a few days ago from scratch on a new computer... But I must say that otherwise you are correct, most of the symbolic operations are supported and with exactly the same format as with Mathcad 15, which tells me that they are using the same symbolic processor core. But for some odd reason you cannot solve a system of 'symbolic' equations, which is a big miss in my opinion... When in the past colleagues who were using matlab for which they didn't have the symbolic toolbox, were coming to me for help with Mathcad, most of the time it was to solve a system of symbolic equations...

[Werner_E]

What you encountered is a limitation of the solveblock. It was not only modified to look uglier but in Prime its also not allowed to evaluate a solve block symbolically.

In Prime its also not allowed to copy a symbolic evaluation inside a program. A trick which made some nice things possible in Mathcad but being undocumented we always knew that we can't rely on it.

[Werner_E]

Addendum (for some unkown reason I am not allowed to edit my posts - the option has vamished from the menu):
You still can solve systems of equations symbolically, but you can't use a solve block to do so. You will have to put all equations in a vector and use the symbolic solve command. And yes, in Mathcad this symbolic solve sometimes behaved differently compared to an equivalent solve block evaluated symbolically.

[LucMeekes]

You're not alone: https://community.ptc.com/t5/Community-Feedback/Editing-a-post-not-possible/m-p/493475#M154.

[Werner_E]

---

@LucMeekes wrote:

You're not alone: https://community.ptc.com/t5/Community-Feedback/Editing-a-post-not-possible/m-p/493475#M154.

---

Looks like PTC has a good hand for breaking things (like Mathcad) and making ugly things (like this new forum) even worse.

[LucMeekes]

Can you do the lower one in Prime?

Luc

[Fred_Kohlhepp]

Prime 3.0:

[sflamant]

Probably, since you seem to be able to do it...

But frankly, I would first have to find out how to edit what you did, and when I face the menu of Prime 4.0 I don't even know where to start as I find the new menus not intuitive at all... And if I need to spend more than 5 seconds trying to find the symbolic menu, that is not a good start... And besides I have 30 years worth of mathcad files, and after trying to convert one of them that would be typical of the complexity of my files, I know that it would not be practical, as the conversion result was horrendous, with pieces of computations everywhere, not counting the fact that a significant amount of computations had no equivalent in Prime 4.0 and came up with blanks and warning messages ( due to the functionality not existing in Prime 4.0 , or as you are showing me exists but done in a completely different way) .

At the limit if Prime 5.0 ends up being worth something, I would have to translate manually from the converted file piece by piece each time I really need an old file. Meanwhile , after 30+ years of mathcad ( the first version I saw at work was a pre-window area, when mathacd was a Dos version) 

Another issue, that I might not have mentionned earlier, is that I don't believe mathcad Prime 4.0 supports inserting routines I would be writing in C, in order to speed up the core of my algorithm as I had to do for one of my telecomunication algorithms with great success 10 years ago, when mathcad was way too slow for the amount of computations that were required  ( was it mathcad 13 at the time ? I don't recall) , and I was able to reduce my simulation time to about 10 hours when writing the critical inner loops of my algorithms in C. Now, no need to say that the fact of not being able to insert "C" routines would be a no no for my new algorithms, as I am now dealing with deep learning algorithms that can takes weeks in simulation time, in my new company...

A long time ago, I also wrote some benchmark files to compare the same loops with log computations or matrices computations written with mathcad and matlab... Mathcad used to always outperform matlab... significantly .  I hadn't run my benchmark files for about 5 years, and a couple of days ago, I was dissapointed to see ( but not surprised) that now matlab is quite faster than Mathcad running on my latest work computer ! Not surprised, because matlab is always evolving while Mathcad is still at the same place as it was 10 years ago... Things could have been different if the dreadful decision of going to Prime had not been made, as by now we would probably have versions 16, 17, 18 running ... and still be competitive against matlab.  

[RichardJ]

---

@sflamant wrote:
Mathcad is still at the same place as it was 10 years ago...  

---

I wish. Now the only license you can get is an annual subscription, which is another step backward for the user. I assume that when that expires it kills not just Prime, but also Mathcad 15. So going forward, PTC can hold new users IP for ransom. Don't pay, get cut off from your worksheets. I guess that does solve the glacial development problem of Prime though. You have to pay every year, even if there are no upgrades at all!

[LucMeekes]

A small correction to your statement "...3 pairs of responses...":

Luc

[sflamant]

Not really, given the fact that you didn't write my initial equations correctly...

The first equation was    x + a*y = 2   ,

                              not    a + a*y = 2       🙂

Sylvain

[LucMeekes]

Sylvain,

sorry about that. Corrected here:

I still think it is two sets, rather that three pairs.

You have to take all three: x, y AND z, from either the top row or the bottom row.

Other than that, I fully agree with your complaints about Prime, and as far as I can tell most of the long-time users of Mathcad-as-we-know-it feel the same way.

I have used Mathcad from version 2.01. And I have the luck to still being able to use Mathcad 11, which I think is the best there ever was; and it is the tool that I use for fun, creative and also for serious work.

I have two versions of Prime (3.1 and 4) on my machine that I almost only fire up to read other poster's sheets when trying to help them. But the ( lack of !! ) speed, in firing the program, actually using it, and shutting it down should bring tears to my eyes. Combine that with the knowledge that PTC now owns Mathcad for about 10 years and has not been able to get Prime at the level of Mathcad 15 and continues to tear down the functionality of this forum, is not a good sign for the future of Mathcad (nor Prime).

BTW. I think that the same UserEfi DLL's that you could write for execution with Mathcad (11, or 15) will or should work with Prime. I remember having tried that once, long ago. Might be that it's limited to the 32-bit version, and it might not work with (free) Express Prime.

Luc

[TimMixon]

Bingo, they blew it.    I am trying to figure out how I can use 15 forever.   I am doubtful Prime will ever compare.

[RichardJ]

---

@TimMixon wrote:

Bingo, they blew it.    I am trying to figure out how I can use 15 forever.

---

Simple. Find a company with competent staff to buy Mathcad from PTC. Then maybe we don't even have to use MC15 forever. Maybe we could have MC16!

Yeah, OK. Maybe I've had too much to drink

[Werner_E]

---

@RichardJ wrote:

---

---

Simple. Find a company with competent staff to buy Mathcad from PTC. Then maybe we don't even have to use MC15 forever. Maybe we could have MC16!

Its nice to see that you still have dreams (& drinks)

[LucMeekes]

Based on past experience, this would be my prognosis for Prime 5:

December 2022 or January 2023.

Luc

[RichardJ]

LOL!

And all new customers will have to pay a subsription every year until 2023 for no updates

[JamesCahill]

Maybe by then it will be capable of producing a plot with gridlines like yours has.

[MJG]

One of the features advertised for Prime 4.0 was going to be plot enhancements.  However, at some point that was pushed out to Prime 5.0.  Therefore, I certainly hope Prime 5.0 will be capable of producing a plot with gridlines.

[Fred_Kohlhepp]

Given:

• The track record of promised capability vs provided capability
• The change in billing/licensing that Richard has highlighted

I'm not going to hold my breath for Prime 5.0 or gridlines, or Axis labels, or legends, or . .  .

I like breathing! 

[RichardJ]

I'm never going to pay them a subscription to keep access to my own work, so at this point I don't care what will be in Prime 5.0. It looks like Prime 4.0 will be the last version of Mathcad I will ever have, and therefore Mathcad 15 will be what I use until I can no longer install or license it for some reason. After that, I will have to move to a new platform.

[JamesCahill]

Is anyone actually using Prime? It seems that few of the old Mathcad Mathcad customers are doing so. To whom are they selling Prime? Are the users of PTC's programs using it? I haven't come across any analyses done in Prime in my work although I still see some Mathcad results occasionally. Does PTC have anyone actively working on it?

The whole thing is a shame.

[jeffreyianni]

I'm unfortunately using prime.

Coming from Mathcad 15, the only advantage I've come across is Excel input/outputs were made a little easier.

[Fred_Kohlhepp]

If you're using prime, then you still have access to 15.

[VladimirN]

Here is what was promised earlier - "Beyond Mathcad Prime 4.0" (in 2015 year):