Changes to upload fees in our future?

From the release notes for 1.23.0 RC:

Changed: Make asset upload fees variable at the viewer

Sure sounds like they're preparing for changes to the price of uploading, but what will the changes be? Different prices based on size of texture, perhaps? Higher charges for animations? Bueller?

I hope not.

jesus, what the hell.

LL you can't keep squeezing us!!!

I think this is one of the carrots to making people go premium, lower upload fees.

Oh that'd be nice.

If it's to do with premium accounts, it's more likely to be that premiums will keep the same upload fee while the fee for non-premiums is higher.

Different upload fees based on texture size does make sense, if you think about it.

Yes.

Mesh support may be added soon. They're big!

then expect others like me who have no need for perm to start in lol

The actual upload fees were already variable and provided by the server.

The real change is that there is a new environment variable LL_FAKE_UPLOAD_PRICE you can set in the viewer to make a custom price appear in the menus. It's probably a testing and documentation hook for the upcoming behind-the-firewall product.

(ETA: if anyone wants to play with this, it's only enabled on Linux, you would need to tweak llinventory/lleconomy.cpp to make it do anything on other platforms.

There are still old hooks in there for energy use and teleport fees and other old things, so maybe there is some end-of-the-world rumor material that can be generated from those.)

OMG if we had to pay to teleport... *shudder*

If that happens my prices go up too.

We used to.

My daily bill was around $700-1000L.

But, we could also fly across the grid if we were really feeling lazy.

Good luck doing that today.

Now this would be awesome! do you have any additional info on what makes you bring this up?
I know there is a Jira on it.

http://jira.secondlife.com/browse/MISC-1494

The first time I looked (perhaps it was a different one) I didnt remember seeing it being a choice of mesh import or improvement of existing modeling tools.

The advantage of meshes is that you can take them with you. That is you create them offline
so they are already yours. As opposed to prims which mostly stay in SL.
In addition you will be able to create exactly accurate shapes you need.
There are ways to "create" prims in blender and then import them into SL but it is a non trivial process.
Plus meshes can be easily converted to other object formats one could use in other programs, worlds etc.

I think it would be a great idea for LL to bring in mesh support especially with the coming of Blue Mars.

Ideally we could get mesh support And improved prim modeling tools. One thing of note is that mesh support will make sculpties obsolete. But in most cases they are sim resource hogs
anyhow.

Sorry, Infiniview. I was just guessing. I know nothing.

I think Viktoria's explanation for the variable upload fee is more likely.

only if a 1024^2 image costs 4 times what a 512^2 image does

Please.... stop give em new ideas.

Actually, I'd rather see a sliding scale, to encourage texture sheeting. If a 1024 with 4 embedded 512's in it were to cost a little less than 4 separate 1024's, people would be more likely to combine textures. Needless to say, this would cut down dramatically on asset requests, network traffic, and load times.

Of course, there would be a downside as well, which might negate the benefits. People could end up combining textures that have no business being combined. The concept only works for textures that are all going onto the same build. If any of those 512's are to be individually placed on separate builds, then having them all mounted on the same 1024 would be a serious detriment. So there's no real way to win.

The only really workable solution remains education. For better or worse, you can't successfully implement a purely technological answer to what is ultimately a human behavioral issue. Plug the current holes, and people will just find newer ways to be equally stupid.

Haha, no that was just me taking off on the mesh possibility. Wasnt trying to accurately guess
the reason for the detail change.

Does it matter? I thought 64x64 or any permutation matching that size was the limit with sculpt meshes?


Wouldn't creators be inclined to think twice, if only for a moment, about uploading excessively heavy textures? Perhaps some creators might be forced to learn from this?

33x33 is the limit on a sculpty's polygons. The actual sculptmap texture is under the same restrictions as regular textures - 1024x1024?. This is assuming you don't use the lossless upload option, which has a greater restriction.

If Linden Lab adds proper mesh support then we'll be uploading model files and not textures. The file sizes will be quite large in comparison to compressed textures.

LOL! If you think its laggy NOW.... if they add Mesh Support, wait till someone uploads a badly constructed Mesh.

Cant wait till someone tries uploading a 100000 face count mesh of a chair or something


They'd have to limit the poly count on meshes or the Grid would grind to a halt.

It's 33x33 vertices. That is 32x32 = 1024 quads (2048 triangles) at the highest level-of-detail. Oblongs are the same number of quads. Using 33x33 vertices requires a 64x64 map texture. Nothing is gained by using larger maps now that lossless upload is available - the extra pixels are just ignored, and compression of larger maps may cause distortions. Viewer 1.23 will (supposedly) introduce smaller maps with less triagles.

Sculpties are essentially a very restricted subset of possible meshes, stored rather ineffficiently. The resource consumption of meshes with similar numbers of triangles will be very similar to that of sculpties. They will be more versatile, but will have to be limited just as sculpties are to avoid lag. Handling LOD changes and collisions will, if anything, be more complex.

Meshes will make it much easier to make a greater range of shapes (for those who can use 3D software) but they will not dramatically increase rendered detail without cost. If anything, they will tempt builders into even more excessive triangle counts and increased lag.

I don't think we'll have to wait that long for hyperlag .... flexi sculpty forests are coming first.

Yeah, I knew this. I was just describing them loosely for the sake of getting to the point. I should've known somebody would correct me.



Larger maps are still necessary because the lossless sculptmaps can take too long to rez. and the extra "ignored" pixels help the relevant pixels to withstand the compression on upload.

Try comparing how long it takes to rez a 128x128 lossy sculptmap with a 64x64 lossless sculptmap. You'll be surprised!.

Obviously you do lose some quality with a lossy sculptmap. But if you're working with organic forms then it's barely noticable. Now if you're making a sharp edged object then I would recommend continuing to use lossless 64x64 sculptmaps.

ps - I've worked a long time with sculpties. I'm not making this stuff up