Last digits & other first observations

[bhesterman-disa]

Mona,

I have started testing various DOE functions and haven't found any problems yet. The testing is going slowly because Spirit is so different.

Here are some issues that I found that are unrelated to DOE:

I had previously mentioned that the last digit of pi is wrong. I checked a few more things like e, ln(2), sqrt(2), gamma that have well-known values that I could check. All were off in the last decimal place by one digit. Sometimes high, sometimes low. If I remember correctly, the HP calculators that I used to use calculated two more digits than what was displayed in order to ensure that the displayed digits were more likely to be correct.

Unlike Philip, I think the grid line darkness is fine. I do think, however, that the plot axes should be solid black, not gray.

I tried printing files with plots on two computers and Spirit crashes.

The values in matrices print gray, and are unreadable with a 600 dpi laser printer.

Here is my biggest complaint, because it reflects a fundamental disagreement with a design philosophy in which alignment of plots is based on alignment of the bounding box instead of aligning the axes. When I have more than one plot on a page, I like to line up the axes. If the plots are side by side, then I align the horizontal axes. If one is above the other, then I try to align the vertical axes. If I have two plots with the same horizontal axis values, then I try to make the plots the same size horizontally.

These goals are not always exactly achievable in Mathcad, but Mathcad Spirit is much worse because the grid is coarser. The fundamental problem is that alignment is based on the boundaries of the bounding box, and not the plot axes. Thus, if argument names of two plots have very different lengths, alignment of axes can be impossible. It would be much better to have the anchor point that snaps to the grid be the crossing point of the two axes. I did notice, however, that the plots in Spirit can be finely adjusted by stretching. Maybe we are early enough in the design process that this could be fixed. The bottom line for me is that if the plots don't look good, I will tend to avoid using Mathcad to present data to others, but instead use other programs for presentation purposes, even if the calculations were done in Mathcad.

Regards,

Bryce Hesterman

[PhilipLeitch]

On 1/28/2009 1:22:59 AM, bhesterman wrote:
>Unlike Philip, I think the
>grid line darkness is fine.

Just to be clear on this one - on my laptop the grid lines litterally not visible with "normal" tilt. I have to tilt my screen at an angle (about 20 degrees) to see the grid. On my desktop monitor (also LCD), it is a little better (and I have more control over display colour/contrast/brightness). But - it is still not clear. So what I was asking for was the ability to adjust colour of the grid. That is, user options for clolour/brightness.

> I
>do think, however, that the
>plot axes should be solid
>black, not gray.

Again - I think more customisation in general would be good. I would like to be able to choose green and blue axis if I wanted to put the plot on a poster (my partner is a teacher and colours are helpful in education).


Without quoting your last bit - I'm in agreement but for a slightly different reason. Basically, I would like the option (default) to align to grid, with an alternate option of dragging and dropping wherever I want (unbound to the grid). I feel that I should have the ability to put things wherever I want, depending on what I have to present and how I have to present it. The grid is great for "working" but too constraining for presentation.

Philip
___________________
Correct answers don't require correct spelling.