First impressions

[Ninetrees]

Well, I admit that I have just begun to test the alpha version, but it doesn't seem to me that much has been fixed since the 'beta'.

Rich
http://www.downeastengineering.com/

[PhilipLeitch]

I disagree. I don't think it's "much" different - but I can definatly see differenes.

There are many more functions and features. They all are like easter eggs - hidden away for you to find.

PolyfitC, PCAeigenvals, Anova and similar are nice little inclusions. The Design of Experiments might be handy... I would have to hold off until I saw some more "fleshed out" examples.

The programming was a little difficult to get the hang of but seems to perform very well.

It is also far more stable - though why they don't throw application wide error trapping around it is beyond me.

The plotting seems far more complete, but I can only do so much with 2D.

I thought symbolics was going to be excluded. Am I wrong in thinking that the attached image is symbolics?

Philip
___________________
Nobody can hear you scream in Euclidean space.

[PhilipLeitch]

One other point - while it is a little faster than the "beta" - I'm still finding it slugish when clicking around the page or entering/altering values.

For instance, if I point and click near the top of the page, pull the mouse down to the bottom of the page and click again, then go back to the top and click again. I can easily do that second click at the the top before Prime has moved the cursor to the bottom.

If I repeat that over and over it ends up just bouncing the pointer around the top clicks, ignoring the bottom ones (or visa versa).

Clicks into formula feel like "Click... and... done". I think it's about a 1/2 second delay to get into a formula.

Other times it is faster - but most of the time it is about that.

And the ribbon is much the same. I press on a drop down and it almost instantly shows me a ghost image of options... then pauses... then shows the options. I'm guessing the "Fade in" stuff is just too slow. Don't bother fading, just show the silly thing instantly.

I've attached examples of this - I don't feel that it should pause long enough for me to take a screen shot of it doing this!

Philip
___________________
Nobody can hear you scream in Euclidean space.

[ptc-1517601]

Philip,

You said you thought that MP1A1 was much more stable than the previous beta -- but in my recent experience, this is not the case.

I installed MP1A1 on a machine with WinXP Home and typed in the essentials of an existing Mathcad 11 worksheet that has a simple example of programming.

This worksheet reads two matrices of comma-separated values via READCSV, then defines and executes a function. The worksheet saves O.K. and loads O.K. But with it I am getting a lot of system freezes that remind me of the behavior of my old Performa Mac with OS 8.1. That is, the cursor simply freezes and ignores cursor movement or keyboard stimuli. With all of these freezes, it is possible to recover with Ctrl-Alt-Delete. But Ctrl-Alt-Delete causes the machine to reboot after a few seconds of dark screen, without presenting the Task Manager window.

Is anyone else experiencing these system freezes with MP1A1?

Roger Mansfield

[PhilipLeitch]

On 4/27/2010 10:10:44 PM, astroger wrote:
>Philip,
>
>You said you thought that
>MP1A1 was much more stable
>than the previous beta -- but
>in my recent experience, this
>is not the case.

>
>Is anyone else experiencing
>these system freezes with
>MP1A1?
>
>Roger Mansfield

I had big problems with the previous Beta failing on me, while on my first couple of days testing this new Alpha only crashed about once to twice a day.

However - yesterday the reverse happened so I'm happy to eat my words. I got a whole bunch of crashes (maybe 10), including the freezing you outlined a couple of times (needing to terminate the process). This is with restarts. I'm also using XP, but XP Professional.

I've especially had issues with resizing, such as with function polyfitstat where I get errors (crash and burn) when I view the nested matricies and then trying to resize things so you can read them. Somtimes it actualy shrinks them, sometimes it gives me sliders that look like they should go up/down but actually go left/right (attached - first image) and scrolling on a nested matrix doesn't seem to work at all (Second image - I had selected to scroll one of the nested matricies).

Also - I think it's a little weird that I have to select "Collapse Nested Matrices" if I want to "Expand" the nested matricies. It's like my Windows XP where I have to go to the Start menu if I want to shut down my computer.


Philip
___________________
Nobody can hear you scream in Euclidean space.

[mzeftel]

Philip,

The resizing of nested matrices is a feature that went in just before the Alpha, so it may still be a bit unstable. Can you send an email to the mathcad-alpha box, detailing the steps when it freezes on you? If you've already done so, we just haven't gotten to it yet.

The button is a toggle button for collapsing and expanding nested matrices. Would you prefer that it was named Expand Nested Matrices?

Thanks,

Mona

[PhilipLeitch]

I'm making sure that I email anything that I add here - so that's done.

As for the name - it should be dynamic. If I click on a nested matrix that is already collapsed, it should ask if I want to expand and visa versa.

If MathCad is to follow the Microsoft logic then it must make the ribbon context sensitive and adapt/adjust items based on what is selected. And that includes the naming of buttons.


Please don't take this as me supporting Ribbon. I think it might make it easier for new users, but it is a productivity hit for me.

Philip
___________________
Nobody can hear you scream in Euclidean space.

[GuyB]

On 4/25/2010 9:12:57 PM, woodwise wrote:
>... it doesn't seem
>to me that much has been fixed
>since the 'beta'.

If nothing else, the ability to write real programs makes this a different beast.

I've been struggling with the simplest stuff for the past couple hours, so I'm not going to leap to the defense of this product.

But at least I'm not forced to rely on PTC's opinion of what functions I can access.

- Guy

[Ninetrees]

I think that some progress has been made, and I appreciate the effort to include new features. But nearly 30s to start up (M' and Help) v M14 @ maybe 8s or so, the arcane keystrokes to exit a programming block (I had to highlight the last line, including the ||, and then hit Shift-Enter), and the always frustrating return to M', seeing blinking cursor, and beginning to type, only to see nothing. If I am in a math region, editing is actually active, but the M' response is so slow that I am frequently thinking that I am NOT in edit mode, and this is further confused by the fact that if I return to M' after visiting another window, and I am in an 'indeterminate' state in M', then I must click on the page to begin editing. That wouldn't be so annoying during testing (needs to be fixed before release, IMO), but it causes me to wonder, when the M' response is delayed after a keystroke, whether I am entering anything...

Still, I was just expressing my frustration. In order to include new stuff, PTC had to let some of these other items, some of which were addressed in the 'beta', to fall by the wayside. This is, after all, alpha testing; but I wonder what PTC thinks it will deliver for M' 1.0...

Rich
http://www.downeastengineering.com/

[PhilipLeitch]

Apparently the speed issue (which also affects me) isn't affecting everyone.

Also - RE Functions, I'm finding that I'm getting the hang of it and it isn't too difficult. I'm that I'm just pressing "Enter" rather than "Shift+Enter" to add new lines.

I just wish I knew how to add errors (raise errors/exceptions).

Philip
___________________
Nobody can hear you scream in Euclidean space.

[IRstuff]

I was told that the M11 keystrokes work in MP1, so CTRL+' generates the "on error" routine


TTFN,
Eden

[PhilipLeitch]

I see now that sometimes Enter/Return works on its own and somtimes it needs shift - from what I can see if anything is highlighted it needs shift, but if you are just at the end of a line you can press enter.

Also - Ctrl + didn't seem to work for me.

I'm not trying to trap an error - that's easy to do. I'm trying to raise one...

Philip
___________________
Nobody can hear you scream in Euclidean space.