Hi

I wrote

x:=1
x:=2
x=2


then I grabbed the second line and moved it to the top:

x:=2
x:=1
x=2

looks like the variable doesn't get redefined. A bug or a feature?

WHOA! then I grabbed the third line and moved it to the top:

x=2
x:=2
x:=1

It shouldn't know what the x value is before its definition, should it? Or am I whongly stuck in Mathcad philosophy?

Konrad

Edited by user 26 October 2010 19:00:09(UTC)  | Reason: Not specified

Hi Radovan
That was at work, auto calc was on, even hitting F9 didn't help. Now at home, it works alright. The difference is windows 7 (64 bit) at work and windows XP (32 bit) at home. Both portable version, unstable. I may investigate it further should anyone be interested.
Konrad

Hello Konrad,

It seems you are right. I can confirm this as well. Moving regions with the arrow keys will produce the situation you explained - the recalculation will not be performed until we press F9, differently than moving with the mouse.

One another thing is problematic. When you move the selected region with the arrow keys up (not with the mouse) above some other regions (not with the mouse), then enter key below this moved region will not produce additional empty lines and will not move all the regions down (only some of them), backspace and delete will not delete empty lines and move the regions up (only some of them). Similar situation would be when you move the regions with the arrow keys down, below some other regions. I think that Andrey is struggling with this and I think this is connected with recently introduced collapsed regions.

Regards,
Radovan

Edited by user 27 October 2010 09:02:03(UTC)  | Reason: Not specified

That's great, Andrey, thank you. Would it be easily done for you to somehow (optionally) mark redefinitions of a variable, like it is in Mathcad? It can underscore a definition with a green wavy line when it is redefined. It helps very much in finding errors in engineering calculations.

All the best
Konrad

Radovan,
I'm in pressure vessels. The document (a stress design report) needs to calculate similar things many times, for example bolts for different flanges. Say, I define diameter of the bolt at the flange A and check stresses, in the next chapter I define diameter of the bolt for the flange B. Because these flanges can vary in size and pressure, bolt diameters will be different. That's alright. But often in the third chapter, I need to use the bolt diameter to calculate something different. Now if I define them with the same name, I have a problem, good if I spot it at all. That's why some signaling of redefinition is very useful in tracking errors. In engineering, consequencies of a stupid error can be dangerous. In fact, the best thing about it is that you can turn it off when you don't need it, but sometimes you really do. Engineers are not the best mathematicians in the world, they need a bit of foolproofness sometimes :-)
Konrad

Radovan,
I don't like it either, but I use it whenever I need it. And switch off before printing- I don't want it on the paper. In a document more than a 100 pages long, it's really hard to find errors of this nature, especially when there are many hundreds or thousands of different variables defined all over the place. Yes, you're right, it's mostly when I copy/paste bits and bobs between pages or different documents.
Konrad

Sure, just expressing my interest :-)
Konrad