Thanks Tracy,

I wrestled with this as well. The options I explored were:

1) to show the action applied simultaneously, which is probably the most 
accurate but makes it harder for a user to see that the arguments correspond 
atom by atom 
2) to show the actions left to right and top to bottom (western bias in 
reading/writing style) which is less random, but also suggests a particular 
order that may not be present
3) the approach that I finally took, which still could suggest a pattern which 
does not actually exist.

Of those approaches, the first is actually my preferred since it is also the 
quickest (but then I already know how the operation works on matrices, a novice 
might not understand how to decode the simultaneous result).

During this testing phase, I'll try these different approaches (as well as any 
other suggestions that arise) and we can see which is preferred. As you suggest 
there are some situations such as Insert where order is very important.

Thanks so much for your input.

Cheers, bob

On -Feb23-2010, at -Feb23-20107:11 AM, Tracy Harms wrote:

> In the second section of the video the addition is shown by pairs of atoms.
> The sequencing of these pairings seems randomized. I find that aspect
> unappealing.
> 
> My intuition is that I'd like to see them applied simultaneously, as there
> is nothing about this verb that suggests ordering. I recognize that this
> might run against the intent to make things clear by showing them in a
> piecewise fashion.
> 
> If there is to be a temporal ordering, I'd prefer right-to-left,
> top-to-bottom. That would help emphasize the sort of evaluation that will
> need to be envisioned for things like Insert.
> 
> --
> Tracy
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm

----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to