View Full Version : A couple of errors in LScript

09-21-2003, 02:25 PM
Hi guys!

This is not a help demanding post. The only purpose is describing a couple of fixes needed for LScript

First of all, the new indexOf method doesn't seem to work well when trying to find a point along an array of points (obtained with point.polygon(), for example). I'm pretty sure of this being an LScript issue since I writted my own indexOf function and all worked fine. To be more specificall I used it to get the different positions of a point along a poligon (if there's more than one, of course). Something like this:

index = pPoints.indexOf(point);
index = pPoints.indexOf(index + 1, point);

When used, it returned something like "bad argument type provided to indexOf)

However, when using my own function (more or less wassomething like this

indexOf: offset, value, array
for(i=offset; i<=array.size(); i++)
if(array[i] == value)

...all worked fine

On the other hand, when updating to 7.5c I've experienced behavioral changes in requester control active / visible toggle properties, having to rewrite a part of my code for this. As an example, one of my controls had to be invisible when a ctlchoice had value 1. However, if that control was set to one by the own LScript, no matter how I tried to refresh it, since it continued being visible until the user switched it manually. This didn't happen on 7.5b. As a proof of this, try my LScript jettobevel on both versions (this happened when you selected the profile option and there was no profile; in that case the own code switched to the last selected option and all controls changed their state, which is not done in last version)

For the interested, jettobevel has been updated (more than updated it has been 50% rewritten). Along with solving the commented issues, there are new features such as:

-Support for patches and VMaps in fake-realtime
-Support for profile curves, with interactive interpolation options
-Internal improvements in all routines, now special cases such as polygon with points repeated (usually polygons with holes) are calculated.
-Improvements in the refreshing routine. Now, a 100 steps undo will undo in a single step, there's no undo limitation for it to work.
-You can save and load your current multiple-steps bevel on disk. Interesting for repeating a 20 steps bevel on other models
-Another minor improvements
-100% stable. 0 crashes on my computer in all kind of ill-mind-comming-from tests, working as supossed on all them.

I shouldn't say this, by I think is one of the most powerful and solid modeling tools you can get free. (I have just attached it to the 'b' key on my computer). The only restrictions are those impossed for the LScript interface.

I'm now writting the documentation. I will release on my future web space as soon as possible.

Bye and excuses for my poor english and for this long self-promotional post.