ORC Slays ARC

Nalates Urriah has an extremely interesting update from the Mesh user group meeting in which changes to prim equivalence on mesh and changes to Avatar Rendering Cost (ARC) are discussed. There’s also information about a funky new build panel with confusing looking information in it as part of the details of Object Rendering Cost, or ORC …. I’m not making this up!

There’s a picture on Nalates blog of the new build panel, which you can view here. I’m rather amused to see there’s a “What is all this?” link on that image because a hell of a lot of people are going to be wondering what the bloody hell it all is!

Basically this is all about trying to make things easier to understand, prim equivalence is confusing, Gwynneth Llewellyn recently blogged about how confusing it is and how unfair it seems, you can read that here. The comments are worth reading on that blog post too as people try and get their heads around everything.

Prim Equivalence is going to be changed to Land Impact, which may or may not make things easier to understand, but it should avoid issues with people thinking a seven prim mesh linkset is the same as a seven prim normal linkset. Actually last night whilst messing around with basic mesh I managed to link two mesh prims and come out with a prim equivalence of 1 but that just goes to show how confusing things can get. However, unless we go to a new standard for all prims, meshes and sculpts, we’re always going to have confusion, the prim system has always been easy to understand, the problem has been that it’s not always a fair reflection of resource usage. This will run and run I’d imagine because I don’t see any easy fix for this.

Avatar Rendering Cost is about how much load an avatar in your view places on your viewer, this is a viewer side statistic and different people will see different costs on different machines, clear as mud hey? This is one of the reasons why this has always been a controversial feature, although the green, amber and red guides help you personally to see what’s taxing your graphics card, they are not the same for all people, one man’s amber is another man’s red, for example.

Avatar Rendering Cost is going to be recalculated and given a new label of “Show Draw Weight For Avatars” if you want a simple explanation of how these costs are calculated, I can’t help you, if you want a detailed, complicated, lots of figures and formulae explanation, then read Nyx’s post on the wiki, which you can read here.

As with all user group meeting discussions, the information relayed could very well change before being made official, although these changes are well on their way, something could happen to prevent them from being fully implemented, so keep an eye out for more official announcements.

Mesh right now is at a very early stage so it’s not surprising that issues are still being ironed out, Mesh is not as easy to use as a prim and probably never will be, even texturing inworld isn’t as straight forward, which may give creators pause for thought on making items modifiable, although eventually I’m sure this will all be sorted out to a satisfactory conclusion, early Mesh clothing is already running into the issue of people having trouble resizing it, experience will help here and the more people use Mesh the more solutions we’ll see.

Where this all leads is still open to debate, I’m not seeing many Mesh buildings, personally I’ve only just got my head around the concept of being able to upload a Mesh cube that I can texture each face of inworld, and that was by accident, on a post over at SL Universe Dain Shan corrected me when I said it wasn’t possible, the trick is to assign an individual material to each face of the cube in Blender, then when you upload it, those faces with materials assigned to them act like standards faces and you can texture them individually, although texturing them still isn’t like texturing a standard prim, the Maths are all messed up for a start, anyway I haven’t seen many Mesh buildings, possibly because prims are far easier to use still for building, partially because of prim equivalence issues that Gwynneth raised in her blog post too.

Mesh will take a while to reach huge adoption, so there’s plenty of time to figure things out but I do feel Linden Lab have got it right here in rolling out Mesh before everything was perfect, this way they get a much wider test case and these discussions will be very helpful in making more people desire Mesh.

 

2 Replies to “ORC Slays ARC”

  1. Hmm. So will parcels start using “Land Impact” instead of # of prims on the About Box? And prims/groups of prims will be labeled with LI as well (thus taking into account that torii and sculpties use more resources than simple, untortured prims)? If that’s the case, I see how this is going to be more “fair” (in the sense that people will be billed exactly for what resources they consume), but a huge mess in trying to figure out how much land one needs to rez everything — specially because the algorithm for PE/LI will change with the size of the items!

    My two biggest issues with that approach are:
    1) This makes it pretty much impossible to predict how much your parcel can actually hold, taking a PhD in maths to figure out beforehand how many objects you might be allowed on that parcel.
    2) The value for PE actually changes with the size of the objects. So it will mostly mean making all (or many) objects for sale non-mod, to prevent buyers from complaining about the amount of resources they take after resizing them.

    This is a huge change, which reminds me of the old question that customers would ask hosting providers about their Web pages: “My website has 10 pages. How many MBytes will I need, and how much bandwidth should I buy?” This is a question that has no answer (since it depends on a lot of variables!), so hosting providers very quickly moved to a “flat fee” model (or at least one where you could define a ceiling of resources consumed). Linden Lab has this “flat fee” model which is easy to understand and explain. It also makes the same sense as the one used by hosting providers: for the vast majority of SL residents, their land will be empty of visitors, so the actual drain on the grid’s resources is zero. But a few will obviously have their sim crammed full with 24/7 usage. In that scenario, LL overcharges the empty sims in order to support the resource consumption of the always-full sims — trading off complexity of resource calculation for simplicity of the pricing model. Web hosting providers do pretty much the same.

    Moving to a resource-consumption-oriented model is not necessarily bad. After all, most cloud services do pretty much the same: you will never know how much you’ll pay at the end of the month. Still, even on cloud computing, some keen companies have been trying very hard to develop a pricing model as closely possible to a flat fee. I would say that this is a “natural evolution” of all pricing structures on the Internet, and as such, LL’s own pricing scheme, which has always been a flat fee for rezzing an absolute amount of content, should not be replaced by a dynamic pricing model.

    But of course others might disagree — for example, someone who logs in to their home once per day for a few hours but their parcel remains free of avatars in the remaining hours of the day might prefer a model where they pay only for those few hours of actual resource consumption. So perhaps a mixed model as an option is not a bad idea to cover all cases — and a non-flat-fee tier model might actually be cheaper in many cases, thus potentially encouraging more residents to buy more land that they use little, and pay just for that use.

    It’s an interesting idea, and I can only hope that LL plans the change very carefully.

    1. If the beta grid changes come to the main grid, then land impact will be the new figure to watch but it seems as if sculpts will be legacy items and not be subject to the same weights as mesh. Prims are legacy items too, even twisted ones.

      You do raise an excellent point regarding what will work on your parcel, sold as seen will work, start modding mesh, land impact can go up, no mod all of a sudden may become popular. Not just because of the land impact issues, but because creators will get frustrated answering questions from people who have modded a mesh item and now don’t have enough land resources to keep it, as you point out, I’ve always sold items mod, sometimes I’ve done a no mod option, but mod is the popular choice today, will it be when this rolls out?

      Consumers like easy choices, 1 prim is 1 prim, it’s easy to work out what your parcel can support.

      There is one huge barrier to resource based usage in Second Life, Linden Lab have been looking at this for years, but the barrier is how long it takes a sim to come to life.

      Kitely have a resource based model using amazon cloud services and for a couple of people or friends, it is cheaper, but a store? It is extremely expensive, stores want to be available 24/7 the costs mount up. I do like the concept of resource based modelling but LL cannot get sims online fast enough for it to work, long term though, this may well be a solution to the expensive land barrier.

      I agree with you about options, but I want to see more options for premiums too, there are plenty of premiums for whom the 512M land tier is not used, give them a different option.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Follow

Get the latest posts delivered to your mailbox: