Things that drive builders nuts

I read recently another request that LL allow larger prim sizes, and yet again the same response: We don't want to increase prim sizes because of the potential of griefers using large prims with rotators to harass their neighbors.

It is decisions like this that drive builders absolutely insane. Think this through folks:

* Using a 50x50 hollowed cube, an entire building could be built using only 3 PRIMS, including walls, floors and ceiling! To do so using the current methods would require 150 prims!
* With a 50m prim size, we could build an entire floor in ONE PRIM instead of 25.
* One prim would take far less rez time than 25 prims.
* One prim would require far less update time than 25 prims.
* Building would be faster and more efficient.
* There would be NO SEAMS in floors and walls.
* There would be no system overhead used definining repetitious texture placement.
* ... and on and on. The benefits far outweigh anything on the "negative" side.

On the con side... griefers MIGHT use it to harass others. Pretty stupid, considering the proof is pretty easy to spot and deal with. Know what you do in such instances? Kick their slimy tails off the system, like other boards do with users who harass other users! Make it a zero-tolerance rule and bust their butts! That is just so simple a solution! When someone reports the incident, check it out and if it's true... whack-a-griefer. Put a chalkboard on the wall and score points for the Linden who whacks the most griefers. Bet you won't have to play the game very long.

This feature has been needed in the builders community like FOREVER. One would think in a time when system overhead, lag times, rez times etc are at fever-pitch concern, this would be a simple solution to a LOT of problems. And the ONE problem that it MIGHT cause, while worthy of being aware of... surely does not even come close to overriding the benefits to be gained. If there are other concerns in this, I'd like to know them. But personally, I have to believe that the 10m limit was just an arbitrary decision that could be updated with relative ease? Of all the things that could be used to benefit builders, this seems to be #1 on the list... and builders have requested it time after time after time. If you were to put this to a massive SL vote, I would bet you'd get a massively majority vote in favor of increaseing prim size to 50m. Consider it more thoroughly, eh? This is needed.

If you builders out there agree... please speak up here. This change is way long overdue.

As I understand it part of the reason is sim boundries. If you had a prim thats was larger than 10 meters it would mess with the sim transfer overlap of 5 meters and cause quite a few problems.

What problems? As long as the prim does not overlap a sim boundary, how does a 10m prim differ from a 50m prim? And in the case of some who might try to get a prim to overlap a sim boundary, can't that simply be coded so that it doesn't happen (ie, if they try to move a prim over a sim boundary, it simply doesn't go any further).

Thanks for the link EH. Read it. Still stick with my guns on this one. So long as the prims are kept from overlapping sim edges, there should be no problem. I mean, the solution for builders is easy: if for some strange reason you have a build that occupies 2 sims (which is NEVER a good idea, due to sim edge problems)... then put one prim on one sim and an adjoining prim on the other sim... rather than having a prim cross a sim. So simply writing it into code that a prim MAY NOT OVERLAP A SIM EDGE AT ALL... wouldn't that effectively put an end to this objection? Seems to me the benefits of larger prims would far outweigh this one easy-to-overcome negative.

From that thread...

OK, not being acquainted with the concept of "collision events" (at least by that name), my first question would be:

Do "collision events" have any relevance if a prim doesn't cross a sim line? (refer to previous points)

That being asked...

If a prim isn't crossing a simline... how does a 50m prim have any different "collision event" factors than a 10m prim?

The physics thing is a seperate problem, unrelated to the sim crossing prim issue.

Just another nail in the large prim coffin

If you will, explain how this "physics thing" works. Since I have been told-- by Lindens and top-notch builders-- that the 10m prim limit was basiscally an arbitrary setting to prevent people from invading land lines-- what danger does a 20m object present that a 10m object does not? 30m? 50m?

Not meaning to be rude. Truly want to know. All I've ever seen about this subject were ambiguous claims and "griefer" concepts-- but no concrete foundation of reasons. Since this is a function that builders really NEED... would kinda like to know WHY we don't have it... in regular english.

I'm not familar with the internals of the Havok physics enging, but I've done enough geometry to make some guesses. Any non-phantom object participates in physical collisions, even if it's not physical itself. A simple optimization is to compare bounding boxes of objects to see if there is any chance of collision. If the bounding boxes overlap, then you can spend the time to do the harder work to detect collisions based on object geometry. This is called a "trivial reject" test, because it is trivial to do a bounding box comparison, while true geometry comparisons can be computationally expensive. With larger prims, trivial rejects based on bounding boxes become less useful, so you spend more time doing geometry-based comparisons, lagging the sim more.

I don't think this is something we builders NEED. Aside from curved shapes, you can compose large objects from many small prims. If they would give us the ability to create a prim that was a section of a larger piece of geometry, that would let us create giant domes, tori, etc. with pieces that were no more than 10m on a side. After we get Havok 2 and the new rendering engine, we'll probably see an increase in prim limits, so the need for giant prims to reduce prim count will be reduced.

Thanks for the further explanation, but I still dunno as this would make a 50m prim significantly harder to work with than 10m prims. Let me give an example:

If I want to make a BORG CUBE for example (or a large factory building), measuring 50x50m... Currently it requires 150 prims to do so (or thereabouts. There are ways to "cheat"... but not with retaining full texturing options). However, if we had 50m prim sizes, we could make the same cube in 6 prims... or even ONE prim if we don't need insides.

So which does it take the engine longer to handle... 150-10m prims, or 6-50m prims?

Even considering the difference between "trivial reject" and "geometric" object comparison... surely it would take far less processing power to handle 6 prims than 150 prims, yes?

Even with Havoc 2... there are still multitudinous reasons for larger prims. Seamless texturing. Quicker bulding. Solid floors and walls without any prim seam whatsoever. More easily manipulated edge seams. The list goes on and on.

I still have to present the pros seem to far outway the cons. There are so many things in SL that folks have been told and take for granted, but simply aren't the case. It's kinda the Windows Syndrome-- people that have never used anything else think Windows is great and never stop to think that there might be better ways. So I still put forth the concept that one of the best things that can be done to assist builders while significantly reducing lag and prim count problems on land... is to increase the maximum prim size to 50m.

Additional note: as you pointed out "aside from curved shapes" larger items can be made from several smaller prims.

Wouldn't it be nice to use that prim count for details rather than walls?

Wouldn't it be nice to be able to build a large sphere, or a geodesic dome... in one prim (rather than the 37 I recently had to use?).

Wouldn't it be nice to have a 30m curved wall on the outside of your home rather than something that looks like it came out of a Bizarro faceplate?

Wouldn't it be kewl to make a domed-glass home with almost 2000 sq ft of living space... and only 1 prim required for the dome?

Such are the reasons for wanting larger prim sizes. This 10m limit restricts many types of imaginative building.

Wayfinder

Using a 50x50 hollowed cube, an entire building could be built using only 3 PRIMS, including walls, floors and ceiling! To do so using the current methods would require 150 prims!
----------------------------------------------------------------------------------------------------------

I'm a newb builder and wonder how you would put multiple windows in such a cube?

Alexa

Hey Olmy

I already do this every day using a combination of Cadroe Murphy's excellent free tools and an old geodesic machine I bought from Xylor Bakslief (sp? ).

IM me in world and i will send you the free tools or you can pick them up from Cadroe.

I also have a selection of pre-made domes and spheres (Xylospheres!), from small isocehedrons up to geodesic spheres about 60m across if anyone is interested.

These require dozens of hours of hand placement work each however so I would have to sell them not give them away. My house in Achemon is a 57m Geodesic dome if you want to get an idea of the possibilities or what they look like.

Dianne

One way to do so would be to create an alpha texture using Photo Shop or Gimp. Usually though, such builds wouldn't require see-through windows (ie, the Borg Cube concept).
Even if you wanted openable windows, you could accomplish such wth far less than 150 prims. That's why this concept is so important... it would eliminate a lot of hassles builders ordinarily have in SL and greatly reduce prim requirements of things like building foundations.

I understand your arguement for larger prims, but with their current size it is already possible, to grieve neighbors, typically in an attempt to make them move so they can snatch the land. (being done) so the size reduces a grievers power. (dosent eliminate)

with the current system not everyone knows why the prims cant cross so making them bigger and saying "dont do this it might mess things up" is not a worth while solution. there are tons of prims crossing sims daily and they need an exact system in place to control any encounter. currently that isnt an issue because they put that limit.

honestly how hard is it to just use a couple more prims? I know you might have a limit, but you can always buy more land, or build a smaller place. (not your favorite solution I know) in the end you will, (like the new player mentioned) need windows, doors- and any other things to keep your 50X box from being an ugly eye sore. If you just want a 50 X 50 hollow box I never want to see prims size limit be changed even if they fix the sim border issue. if they did, that would only encourage super sized builds and make the grid look like CRAPOLA!

just my point of view and prediction for its future.

you know there is a rezzer for things like that? right? there is a sphere rezzer that can rezz based on pieces (only up to 10X10 of course) into a sphere as large as i think 96m radius

you could cover an entire sim in a dome, make a cool place

Opinion understood. However to answer your questions and comments:

How hard is it to use a couple more prims? Last night I spent 2 hours in a build that I could have done in 5 minutes with larger prims. It's not just "a couple more prims". It's 140 more prims! And even with the instant copy function, the copying, aligning, mass-selecting, linking etc etc ad nauseum of that many prims takes TIME, not to mention prim alottments. So in answer: MUCH harder-- and prim-expensive.

As far as people building large, blocky prims with no purpose, really they already do that, yes? Ugly builds are all over SL. Let's not look on the negative so much and look at the positive: if I use ONE PRIM to build a floor instead of 25 prims... that leaves 24 prims to make the place more detailed and beautiful, to add better furniture, to add that decoration I otherwise might not be able to have.

As for purchasing more land or building smaller... neither of those is really a realistic or desirable solution is it?

If folks will pardon me saying so, whenever I hear anti-large-prim arguments, it always seems to me to take the form of "the sky is falling! the sky is falling!"-- like folks are scared to death of the existance of larger prims and pull reasons out of the air why we shouldn't have them-- and those reasons just don't seem to stand up to scrutiny. In actuality larger prims would become one of the best and most useful builder tools to hit SL in ages... and people who might abuse them (as other things in SL are already abused) could be dealt with as they deserve-- if LL will only step up and do so.

From what I've heard, LL has been a little harder on griefers than they used to be. I think with the thousands of good users they've lost because of griefers (in addition to the Ahern Nuke Incident)... they started paying more attention to just how much griefing was costing them in terms of satisfied, paying customers. So if a griefer was to use a large prim to cause problems-- that problem would be temporary-- as would the griefer's membership.

Larger prims would be very useful even if they were forcibly phantom.
I have to ask myself, with all the superbrains on LL's payroll, why can't they understand the simple concept that implementation details should be transparent to the user, and so all software must be built on multiple abstraction layers.
If havok supports linked sets 30 meters in size, then we can very well have a single, seamless 30x30x30 cube that is represented internally as a linked set of 10x10x10 cubes.

it would be cool if...


but there are too many underlying reasons it isnt yet practicle for SL as it stands today.

If you think some of the builds are huge now, what would you expect to see if they could easily without costing any more prims, be 2X as high ....or 5X? it would be more clustered then it is today.

but it would definently make some things easyers.

I'll go right down the line on why this will not happen. While I know the folks here mean well... I've seen this stuff first-hand.


1) Physics and Space-Related Problems:

To the best of my understanding, there was once a bug that allowed for extremely large prims. Well, some of these were rezzed in Nexus Prim at one point, and I had a chance to analyze them.

The sim hated them. The collision mesh was completely borked, pretending a triangle was a square and whatnot. I'm also guessing if it went physical it would be a nightmare, just spacially.



So right off the bat we have two problems:

* Forcing Havok (2) to work well beyond normal bounds, causing overall slowdown due to greater variation

* Spacial problems - suppose someone attaches that prim to an avatar? Drops one in a sandbox?


I hate to think that the latter concern is due to the possibility of numbskulls abusing the system, but there it is.




2) Allocation concerns:

Suppose I have a 512m^2 plot of land and place a 100 x 100 x 100 prim on the center of it? What now Mr. Magic? You can't return it since it's technically on my land.

If that were fixed, what about a 300m tall pylon into the sky, all on my land? Same problem on a griefing level.



3) Script concerns:

Currently sit targets are bound to a 300ish radius. Using bugs I will not publish (but were reported), it's been proven that breaching this threshhold crashes a freaking sim. How do we address the countless script bounds checks that can be broken in this manner?

What if I send one of these objects across the grid via script?



4) Building feasibility:

Why have this? Link sets, by design, are limited to a 30 x 30 x 30m box from the center of each prim for similar concerns. Wouldn't we have to change both systems? Again, how would this affect spacial abuse?

What recourse do you have against someone building a giant, three-sim frying pan and dropping it on a sandboxen?



5) Viewer Limitations:

Add your own building bug here.

Since we're limited to building everything within Second Life, how would you address this redoubled amount of clutter and new draw statistics?



----

Now, this is not a bad idea on paper. However, the reasons this has not been done are many, and one must remember that we exist in a shared, user-created world with a few people bent on making our lives miserable.

I might warm to this when/if Second Life becomes more focused around user hardware/software, but for now, t'ain't happening.

Thanks for the responses. I'll address them one at a time..

>* Forcing Havok (2) to work well beyond normal bounds, causing overall slowdown due to greater variation<

Am I to understand that Havok by its very nature can only handle prims max size of 10m? That Havok can't handle 20m prims? 30m prims? 50m prims?

>Spacial problems - suppose someone attaches that prim to an avatar? Drops one in a sandbox?<

That's where rules come in. Could it be written into code that prims over "x" size may not be attached to an avatar?

What if someone does drop one in a sandbox? How long would it take for other users to take care of that? Under current conditions I've seen users drop huge builds into a sandbox. It doesn't take long for such builds to be removed, even if a Linden has to be called.

>Suppose I have a 512m^2 plot of land and place a 100 x 100 x 100 prim on the center of it? What now Mr. Magic? You can't return it since it's technically on my land.<

No, but one CAN be banned by LL for doing so, as would happen for ANY significant griefer tactic. This was covered in the first message. Shall we eliminate all the tools in SL that have griefer potential? Oops! There goes CHAT!

>If that were fixed, what about a 300m tall pylon into the sky, all on my land? Same problem on a griefing level.<

There are ALREADY 300m tall builds. I have a castle in ElvenGlen that measures 206m from the ground up. So what if someone wants to build a 300m pylon. If it becomes a hazard to traffic, then users can report it as abuse and it can be handled as such.

>What if I send one of these objects across the grid via script?<

Is it possible to write into the game engine to prevent this from happening? Again, are we being told the Havok engine cannot handle prims larger than 10m? If that's the case, then that's a valid reason. Otherwise, it all boils down to writting the game right in the first place. There are lots of abuses in SL that take place because governors weren't placed on scripting parameters in the first place. That's where MicroSoft messed up-- writing code without thinking about security issues.

>Why have this? Link sets, by design, are limited to a 30 x 30 x 30m box from the center of each prim for similar concerns. Wouldn't we have to change both systems? <

I can't see why that would be the case. We can't link 50m buildings NOW. Why would we have to just because we made the building out of 6 prims rather than 150?

>What recourse do you have against someone building a giant, three-sim frying pan and dropping it on a sandbox?<

You know, I would think simple common sense would handle such problems. What prevents people from dropping NUKE devices in sandboxes now? The fact that they will absolutely be banned from Second Life-- that's what stops them. So my simple answer: if they do, file an abuse report. You'll surely have evidence. But really, looking at this with that same common sense.. so WHAT if they drop a 3-prim frying pan in a sandbox. Who will die? Whose computer will blow up and the monitor shatter? What major nation will suddenly collapse? LOL. It's all a matter of perspective. If someone abuses something, file an abuse report and get on with life. Such potential problems should not stop implementation of a very valuable build feature.

>Since we're limited to building everything within Second Life, how would you address this redoubled amount of clutter and new draw statistics?<

Sorry, I fail to understand. Are you saying that the system will have more difficulty drawing a few 50m prims than hundreds of 10m prims?


>Now, this is not a bad idea on paper. However, the reasons this has not been done are many, and one must remember that we exist in a shared, user-created world with a few people bent on making our lives miserable. <

Yes, and they are going to do so regardless of the prim size builders are allowed to use. They are going to do so whether those prims are 50m, or 10m or 1m. They are going to find SOME way to abuse other users. The answer to this is not to prevent users from having valuable build tools. The answer is to deal with those griefers in a definite manner when they do abuse people. Because no matter what you build into the system.. griefers will still be griefers. They are sick, twisted people who get their jollies out of harming others. Their minds are all messed up and nothing is going to solve that problem except removing them from the board when they abuse the system.

Shall we ban cars from the roads because drunk drivers kill people? Shall we remove kitchen knives from the stores because people are murdered with them each year? Shall we eliminate electricity from our homes because some people short fuseboxes and catch houses on fire? No, we provide the tools, and punish those who abuse them.

Provide the building tools, and deal with the people who abuse them. Isn't that how life works?

A good tool should provide adequate recourse at the system or user level. Using chat as an example, you can simply mute a resident and their objects at will. You cannot make a prim "disappear" like magic here, which has led to several heated debates.

Now then, let's start from the top. First, let's discuss simple math and probabilities. I realize there are several 300m tall pylons in existance. However, you would need to spend several prims to make that happen, and tier accordingly.

By contrast, letting singular prims scale into large numbers would allow those people to buy mini-parcels on the grid and put up the huge farking build of their choice on the smallest of plots. This means there's a better chance the person will grief with it.

Is this worth it so house builders can save a few prims? Maybe, if you're not on the main grid.

The prim system is in existance for two reasons. First, to ease data streaming. Second, to place a finite "value" on each object in the world, as current parameters allow. One could extrapolate from this and ask, "Why don't we just use meshes? It would be a lot simpler." Yeah, it would - but it would be a logistical nightmare. Same problem here.

-----

Next, I will draw on an analogy. Know those debates about malicious teleport scripts? A large part of that problem is people taking the path of least resistance. Because it would be easy to just spit out 100m x 100m x 100m boxen means people would very likely do so... at the expense of everyone else trying to maneuver around it.

So second case in point, spacial concerns again take precident.

-----

And how do you define griefing in this manner? Maybe someone is making something worthwhile with that huge prim in the sandbox and you just don't know it. What litmus test can we use to prove someone is acting maliciously with a prim's scale?

I'll admit this problem is readily visible now, but is less of a present concern because abusing it is difficult. So once again, making it "easier" does not necessarily mean making it better.

Also, you're being very dramatic by saying "well we can just ban everyone that abuses it!" If only it were that simple. These problems are not so black and white as that. The Lindens do not have the resources to play their cards thusly, but in the long run this may be a possibility.

-----

Now, on to physics and rendering. As I understand the current system of collisions, mass, et al, much is based on integrating several pieces of the collision mesh and...

Okay, in English. Basically, more complex prims mean more complex calculations. which ultimately means slower sims and clients. Making extremely large prims might be wise for a cube - simple enough math there.

But how about 300m hoochie hair? Making it look appropriately curved would force the renderer to draw dozens more polygons on the screen, leading to overall slowdown.... back to the same problem.

Making larger prims physical would lead to a similar problem. Suddenly, because the object is so massive, you're having to account for dozens more collisions than you would have otherwise. Collision math hurts sims badly - I've managed to crash a healthy sim by mistake with 28 prims colliding the wrong way.

-----

So at the end of the day, it's all connected. Saying "well, we should just have a better system" is basically telling them to overhaul a huge portion of Second Life. These things don't happen in a day, a week, or even a year.

Instead, what would be more likely if this happened is more of the same. "Oh my God, why does nothing work anymore? I'm leaving."



We will just have to agree to disagree. Given infinite time and resources, this might be a wise idea. Given the current situation... ask for this in about a year or three.

Well, I certainly see your points, and some of them are to an extent valid (I mean really, "300m hoochie hair"? LOL).

But I will say this: if this kind of general attitude existed in the real world... we'd never have had cell phones, automobies, airplanes or made it to the moon-- because people would have been so terrified of "potential" problems they'd have never taken the risks necessary to accomplish great things. Yes, there would surely be SOME abuses of large prims (just as there are abuses of the system now). The widespread benefits would far outweigh such isolated cases.

There are potential problems in everything we do. It's weighing those potentials against the definite gains that allow us to grow.

Well, if you're already building your 300m hoochie hair with more prims, the number of polygons rendered is actually roughly the same, as are the collision calculations. Where you might be onto something is that if it costs me half my alotment of prims to make that curved surface, I have a lot less prims left to make a lot of other calculation intensive objects in my space.

I think the argument about cost/prims is a valid one, however, perhaps this is an option that should be available for adjustment on the private islands/continents? Make it become a governmental decision for those places so that the Lindens don't have to get involved in ajudicating griefer issues.

This would also improve the market for land in those private sims and leave it to the builder to have a choice between the two types of environments. Dreamland, for instance, has some strict zoning rules that would still have to be complied with.