AO vs Built-in Sit anim

I'm working on a fixed AO, that when you attach it plays 1 single animation (1 frame pose that loops, priority 4, twisted all joints) all the time, no matter what default animation is going under. I placed a timer event (0.25) only for checking out when Avatar is Sitting, so I can stop the default sit anim (also priority 4) which otherwise would override my animation.

The problem is that, still at those tiny time intervals, it starts playing the default sit when I sit, then stops, and plays mine. Only plays the sit for a bit, although not always the same duration.

I believe what is happening could be the correct behavior: play a priority 4 anim (mine), play default sit anim (also 4, which overrides mine), stops sit (which shows again mine). I understand that.

But why I still can see the sit anim for a bit (a sec/half sec, always varies) when I'm polling to stop it at 0.25 sec?? Shouldn't be almost imperceptive?

Thanks in advance.

I think your problem would disappear if you'd upload your animation with longer lead-in and lead-out times. Half a second each should do it.

I'm afraid longer eases won't help here, you'll just see a different glitch during the transition!

Instead, most AO's use a llStopAnimation to stop the default sits, and that way you also don't need your animation to be priority 4 (so it will work with others) nor a laggy timer.

If you look at the freebie AO's, (search "sit" you'll find how they go about it.

Even with a 1/4 second timer, the defaults can leak through. There are two things fighting you. First, scripts are only told about animation changes after the fact, so the default will always have started before the default can be stopped and/or overridden. Second is that these messages are being sent over the network, and throughput is variable.

This problem isn't always apparent with most AOs but it is always there. It doesn't usually show because most sit animations bend the body anyway, so "starting to sit" just turns into a different version of "starting to sit" instead of "suddenly standing back up".

Yes, I'm stopping the default sit, but as Viktoria said it always starts first before I can stop it.

And the glitch is more visible in my case, as my anim poses the avatar far away from my agent. So when I sit the Av quickly tries to sit where I am (agent), changes its mind along the way, and then moves back at the anim position.

So the 1/4 sec is relative.. that explains it. I don't see any workaround with this. Any ideas will be appreciated.

Yep. There really isn't any event you can rely on to make things faster, so if the sit happens right after you check, you get the full timer wait plus whatever lag goes into the viewers seeing the changes.

Does your animation allow for a bit of cheating? In other words, could you get away with some kind of " I meant to do that, ha!" kind of bending action that could be played at sit time to kind of cover for it? Yes that's cheesy, but cheesy often works in SL =)

Apparently not. Is not much the bending that bothers me, is the abrupt movement, remember that the Av is some meters away from the agent. Is more like it bumps against an invisible wall in the air.

Besides that, I was checking the wiki and the moving_start event.



The attachment is changing position when I sit, also (as I mentioned before) is moving (I wish it were not, but it is), or at least is what you can see. But isn't triggering at all. So I guess it's not considered a movement, just a position change?

Sitting on a sit target won't get picked up. Sitting on a prim without a target (the sort of sitting that kicks in the autopilot movement) *should* get picked up but it doesn't happen reliably, because moving_start is broken in general