Set Scale limits or recent Bugs?

Ok, first - thanks for the recent help, I've got everything running how I want it. Now I have a very weird problem I've never had before...
llSetScale isn't responding properly anymore.

Linked object is 21 prims

I've tried both normal listens and linked messages (currently using)

I always have one prim that does not respond, *AND* it changes from place to place. I thought perhaps I had forgotton some root-prim property since that was the one not responding, but then... I went to a different sim, rezzed the object, and a *different* prim stopped responding.

Here are code snippets:

Child Prims:
vector startScale = <6.66744, 4.14140, 0.19629>;
vector endScale = <0.010,0.010,0.010>;
default
{
state_entry()
{
llSetScale(endScale);
}
link_message(integer sender, integer channel, string msg, key id)
{
if (msg == "open"{llSetScale(startScale);}
if (msg == "close"{llSetScale(endScale);}
}
}



Root Prim:
vector startScale = <0.03409, 6.37608, 4.00900>;
vector endScale = <0.010,0.010,0.010>;

default
{

state_entry()
{
llSetScale(endScale);
llSetAlpha(0,ALL_SIDES);
llListen(8,"","","";
}

on_rez(integer num)
{
llResetScript();
}

listen(integer channel, string name, key id, string msg)
{
if (msg == "open"
{
llMessageLinked(LINK_SET,25,msg,NULL_KEY);
llSleep(.5);
llSetAlpha(1,ALL_SIDES);
llSetScale(startScale);
}
if (msg == "close"
{
llMessageLinked(LINK_SET,25,msg,NULL_KEY);
llSetScale(endScale);
llSetAlpha(0,ALL_SIDES);
}
}
}

Note that I tried using the same child script in the root prim and using a separate listen-only script. Same thing occurrs.

Also - the SetAlphas were thrown in there to simply form a temp solution to the problem. But since it is shifting which prim has the problem.. it is no longer sufficient even as a band-aid.

Any thoughts, ideas, or experiences?

I've already shift-copied after deleteing and scrubbing all scripts/prims. I don't think it's left over script issues or anything. I've relogged a dozen times, tried inventory swapping. Nothing seems to help. Every once in a blue moon - the whole thing works like it should, maybe once or twice, then a prim will 'stick' and not scale along with the rest.

(Should I offset them all by a slight amount with a timer event? Too many scale calls at once?)

Thanks,
- Hap

The only thing I can think of is that your changes are violating the distance rules...

Agreed. However this is basically a preview screen, with option buttons along each side. (Once I throw in the completed script - which I've refrained from doing until I get this sorted out - it is nothing more than a testure switcher to create a slideshow to preview stores for people - the buttons on the side select the store)

I'm using setscale for nothing more than a nice effect. You click on the "Previewer" object, and out springs a nice large framed screen with streamlined buttons and interface to use.

After further testing throughout the day - it seems consistent that the root prim is the culprit. I've replaced it completely, rebuilding it only from the setscale vectors alone. I even re-wrote the script from "New Script" By hand from contents, as a last ditch effort afraid somehow a cut and paste had gotten a bug.

Wow... any other thoughts would be appreciated. Thanks for input so far. It's all link messages and it's all within maybe 10-15 meters total spread with a depth of perhaps .5m at most.

Take care
- Hap

how many prims are changing size at once? can you target the link messages?

there have been ocassional random failures of message sends in LARGE sets of prims using LINK_SET, but 21 is well under what I've seen (more like 120+). but those never occured repeatedly in a single prim.

I'll assume there isn't a difference in the linking between the differing failure tests (in otherwords not the same link number failing in all cases)

are you sure the message or call is failing? maybe get some echo code in there, to make sure the message is going through... then if it is try making sure it's not the viewer that's freaking out, by tpin'g away and then back.... it could just be that it's not updating on your screen.

Indeed you were right... SL having a bad hair day. it works just fine now without any changes to the script at all. (I did add some says in there to make sure all of them were firing, and they were)

So I had nothing more than a couple of days worth of weird frustration.. probably just due to a client update or something.

Thanks for all the input. Pretty much sews it up now. Once again I appreciate all the help all.

- Hap