Massive push, an academic obsession

I want to understand how massive push - the kind that occasionally breaks your client and sends you into a screaming fit of obscenities - is scripted.

No, I don't want to script it. I hate that crap.

But I do want to understand it, and it's driving me nuts. I can look at most products and understand - at least academically - how their scripting works. But massive push, I've never been able to get.

I understand that object mass influences push, and that the number of prims an object has can affect its energy repletion. But while I've been able to create objects that push me, on click, about 500m straight up, I seem to cap out at about that.

Would anybody care to explain to me how this works? I'd love to get it, I really would. Examples would be best, of course, but if you don't want to post them here for fear that they'll fall into the hands of script kiddies, I'm fine with talking in general academic tones, or through IM in-world.

Actually, it's quite simple.

There's a limit to how much any one object can push.

So get a bunch to push at the same time.

It's best to not attempt to break Sl's energy model. There are reasons that energy was even introduced. But as a general rule smaller objects gain energy quicker than bigger objects. There are certain ways to "tourture" prims, so that what would normally be a small prim with little mass and great energy replication, will result in small prims with super massive mass and great energy replication. I think that's what you are trying to understand, it's just an exploit of prim manipulation. And there are a "few" other ways to achive greater pushes, but most tax the sim pretty greatly. Good luck.

I have an object I built that consists of 128 very small prims linked together, each with a push script in them. Small prims seem to be more effective than big ones - presumably because they allow you to get the maximum number or pushes as close as possible to your center of mass.

In an empty almost-script-free sim, under Second Life 1.6, I was able to use this to launch me over MAXINT meters. I don't know exactly how high I was, because the altimeter I was using used llFloor() to clean up the output, and llFloor() is actually just an integer conversion and I didn't think to check the debug menu... but I was at least 2 million kilo meters up. That's five times as far as Neil Armstrong ever went.

It was high enough that rounding errors really messed up the appearance of my avatar. and most of my clothes and objects got blown off. My "spotter", Furlop Duport, reported that his map showed me as being below ground level.

I was able to teleport back to the sim, though, and dress myself. My spotter repeated the journey wearing an Extrovirtual "Tiny" as his avatar, since earlier push experiments showed that minimal sized avs made better astronauts (yes, I know, a Tiny isn't really smaller than other minimal sized avs... it's all about style), but we don't know how high he went either. When I thought to try our astronautical experiments later using the camera position debug panel I was never able to even get close to the limits of my altimeter... my best was under 3 million meters... and I usually didn't even lose my clothes.

Under 1.7, I haven't even managed one million meters using the same tool. So the long gone pioneer days of interplanetary space travel will probably never be repeated.

Jeez... I only managed 80,000 ft... with 192 linked and tortured spheres all pushing at 9999... what kind of prims did you use, I wonder?

Not that it really matters - I have no more use for a 1 million meter push than I have for 80k... but it's nice being able to say I can do such a thing - y'know, bragging rights, etc.

Actually, as I type this, I'm plummeting in-world... this is a very pretty blue horizon. Unfortunately, I set all the prims to push owner and didn't bother putting a filter on the touch event, so people keep wandering by and clicking on it just as I reach the ground...

string hindsight = "20/20";

* cough *

In case you missed it.

I wonder, if you go high enough... will you hit the ground mesh from below or something?

I don't know for sure, but looking at the dates on these posts I think I beat the rest of you to MAXINT meters. I'll check when I get home but the timestamp on my snapshots should establish priority.

My pusher was just made of simple cubes, no prim torture involved... seems like it was quite similar to Beatfox'.

PS: I set mine to push whoever touched it... but put a big floating text "DON'T TOUCH THIS" on top of it. Why would someone touch something marked "DON'T TOUCH THIS"?

Teh Funnay!

Excellent implemnentation of hte "Don't stick beans in your ears" principle.

I woud instantly click anything that said "Don't Touch This" in-world.

Okay, now I've become obsessed with hitting the 1 mil m mark.

I've got 128 prims, each with 5 link_messaged push scripts pushing at the max push vector (according to some bloke on the wiki) on the z axis. Each of these scripts calls the llPushObject function 5 times.

The prims are all .01x.01.01 cube prims, linked in a stack formation.

There has GOT to be a more elegant solution than simply adding more linked scripts with more function calls.

On the plus side, I did put a getowner filter on the touch, so people would stop launching me while I was editing.

sorta on topic,
i had a "panic button" set up at my place. DON"T TOUCH in big letters. After you touched it once it would give a chat warning, "please don't touch". The second time you touched it you would orbit the moon. It was touched daily

1. I got better pushing after I quit worrying about "the max push vector" and just pushed at MAXFLOAT. Don't forget the r-cubed falloff is going to pull that value down awfully fast.

2. After the first push that r-cubed falloff is going to make subsequent pushes increasingly less relevant, anyway.

3. Staggering the pushes very slightly also helps.

Beatfox Xevious' super-pusher, which I noticed was 128 prims, sent me 596 million meters upward.

Once I got there, I noticed:

•I couldn't use Mouselook to see down or up;
•I didn't "fall" even with flying turned off;
•I could only see my eyes;
•I was able to actually move laterally, but exceedingly slowly. Crossing sim boundaries took place normally.

Try using 2 billions instead of 9999 maybe ? Also, using multiple small prims containing scripts, all linked to a huge single prim (for mass) might help, though I didn't test it.

I set the push to MAXFLOAT, and staggered the pushes in intervals of llFrand(0.1), and managed to hit 250km. I think that's as high as I'm going to get off a single object that can be re-rezzed (I created one once that sent me high enough to break my altimeter, strip me of some attachments, and show me as a colored blob instead of an av, but when I picked it up and set it back down, I was back in the 80k range.).

Just for the sake of saying I hit 1m mark, I sicced 10 of the damn things on myself and saw what 1.5 million meters looks like. Now I'm wandering over to Beatfox's to ride his launch cube and then I'm going to take a nap, LOL. Thanks all.

Oho! That might be my problem! I'll try building it again, and only using a "fresh" copy. Maybe there's some energy considerations? How about rezzing it as a 10m cube and shrinking it to 1cm from the script? Beatfox could be doing that... having it rez a "fresh" pusher each time...

Also, what my object did was have each script do its delay (based on llGetLinkNumber) and then do a bunch of pushes as fast as it can. I'm not sure (as I said) that pushes after the first one or two have significant effect but it can't hurt.

Almost a year ago the first script/object I made took me to the "end" of SL going straight up ... you end up in like a "corner" of the virtual universe -- it's all black. Imagine being squished in the corner of the top of a giant cube prim, that's what it's like. I had an altimeter, I forget how high it actually is.

I think the only real upper limit to SL space is the highest possible float value. I reached 40 trillion and was still climbing. The sky is just constantly black at those heights.

Well thats only 1 / 236513 of a light year...
9.4605284 × 10^15 Meters = Light Year