weird behavior

Alex Tweedly alex at tweedly.net
Sat Jul 17 21:10:06 EDT 2004


At 23:24 17/07/2004 +0200, Wouter wrote:

>Hi all,
>
>Some weird behavior.
>
>If  the style of a created graphic is set  to polygon and the points of it 
>are set to a list like:
>
>-----------   start of list
><point1>
><point2>
>
><point3>
><point4>
>
><point5>
><point6>
>------------- end of list
>
>then revolution adds another point to the list.
>it becomes:
>
>-----------   start of list
><point1>
><point2>
>
><point3>
><point4>
>
><point5>
><point6>
><point1>
>------------- end of list
>
>whereas in metacard  no point is added
>
>If  the style of the graphic is set  to curve and then reverted to polygon 
>no point is added no more.
>????  Puzzling...

I saw something like this when I was trying my "speedy graphics" examples, 
but thought it was just something I had done wrong. Or rather, was still 
doing wrong - I kept seeing odd behaviour that I don't understand, and was 
kind of stuck in trying to complete the project.
So I'm really glad you sent this email :-)

Looking into it some more, with this hint in mind, I think I have a clearer 
picture.

I think you have a malformed polygon. From the Transcipt dictionary (points 
entry):

>A blank line in the points indicates that the previous and next vertexes 
>are not connected by a line--that is, the line, curve, or polygon is 
>broken into two (or more) pieces. If the last line of the points of a 
>polygon is blank, the polygon is not closed. A closed polygon's start 
>point (the first line of its points property) is the same as its end point 
>(the last line of its points property).

NB:
An open polygon has a blank last line
A closed polygon has the last line == first.

Your example, like many of mine, had neither of those conditions true. So I 
guess it is, strictly speaking, malformed - and hence the behaviour could 
be undefined.

I've been unable to see any failure with any example where I had a properly 
formed polygon according to specification.

I still think it can be considered a bug - but at least it can be avoided / 
worked around.

-- Alex.
-------------- next part --------------

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.719 / Virus Database: 475 - Release Date: 12/07/2004


More information about the use-livecode mailing list