[Viewer 2.3 Beta Available Today With Display Names and More]

<obligatory meh>

 

It works.

[[VIDEO] Align planar faces for easier texture mapping in Viewer 2.2]

I've had a Scripted tool to do this I've used for a couple years. I have to say it's very nice to have the functionality built into the viewer now. I'll have to play around with it some more, the scripted tool can operate on cut faces, not sure if this does.

[Viewer 2.2 Available Today!]

I have been using a third party flavor of 2.2 recently, and I have to say I like where it's going. Great work.  It's good to see the Snowstorm Team pushing forward with improving the issues closest to the hearts of the residents, and embracing some of the more ingenious 3rd party mods. 

Keep it up.

[Back to Basics Technology Improvements]

Thanks for your comments on the JIRA. Now if only we could get some vote for it. You know LL doesn't implement things without thousands of votes behind it.

[Back to Basics Technology Improvements]

Unfortunately the code you wrote won't always work, as the auto-play parameter can be overridden in the user settings. The problem is this setting is global, while more likley people would prefer to selectively have some items auto-play, and others (or most) not.  There is no way for a developer to be sure people don't have auto-play off, and that their interface, data is viewable by the people who want to see it. My proposal makes auto-play a per-object setting in essense.

[Back to Basics Technology Improvements]

My point still stands - we really have not seen it exploited because few want to subject themselves to the 2.x viewer.

I'm afraid that's not he only reason. The other major reason is how it was implemented.  Because of the way LL designed MOAP to 'turn off' under so many conditions, it can't be considered a reliable user interface. It serves nobody to be 'off' and unviewable in 70% of the cases it's encountered. And the fact it's not persistent, makes it very inconvenient to use for customers/residents.

I posted a proposal to correct some of these issues in the JIRA: https://jira.secondlife.com/browse/VWR-20428  as I would like to see MOAP be all it could be.

[Back to Basics Technology Improvements]

I have to say the general consensus from users I've talked to, is that lag is worse. Specifically in the realm of inventory loading, lost inventory due to lag, and issues related to this. So while you may have fixed one small cause of lag (avatars teleporting into the sim), I wouldn't unfurl the "Mission Accomplished" banner just yet.

As for the group chat fix forthcoming, is that going to have any positive impact on open chat performance? Many of the same issues are seen with open chat (long chat delays, out of order chat, chat never arriving, etc)  I hope these fixes will also address open chat failures as well.

[Mesh Import Open Beta Starts Today]

Did you actually read the comments? Only one referred to the economy that I saw and Jack responded to it. Everyone else is just sharing about mesh or expressing frustration that things don't work. I'm just not seeing the mass hysteria you are suggesting is happening.

I think his name explains all. That anyone would even suggest there should be cost associated with anything is hysteria, as if he had his way, there would be no costs at all, with anarchy and consequences be damned.  You know, the Open Source way.

Shame on me for being the adult here and wanting to tackle the hard questions associated with Mesh, which is preserving a balance of cost/benefit for consumers/creators, so SL still has a reason to be. I'm sure LL as a company wants to preserve that too.

[Mesh Import Open Beta Starts Today]

Thanks. We definitely need to get some clarification about what these various cost numbers stand for.

I know the one on the left is the rendering cost. But the 4 different values on the right, bear some explaination.

[Mesh Import Open Beta Starts Today]

Looks nice. How many prims is it counting as?

Not parts, but actual prim count on the parcel.

[Mesh Import Open Beta Starts Today]

I  understand. It's just my opinion that this detail should have been ironed out earlier rather than later.

Mesh will work, there's no doubts about that. But more attention should be paid to the implications of it's introduction.

For those who are interested in examining LL's proposed cost formula closer, I have created a command-line tool to try the formula out under different conditions. Link can be found on the Talk page in the Wiki for the Streaming Cost formula. I know I for one will be giving it a hard look.

[Mesh Import Open Beta Starts Today]

It would be nice to see some people post some numbers about the complexity of their build, how many LODs they included, and the corresponding prim count the said object came out to be.

It seems LL has waffled again and changed the way prim counts are calculated. It would be good to know what they are aiming for now.

IMHO, this is the single most important factor about mesh, as if they get this wrong, it will destroy the economy.

[SL Marketplace is Open and Ready for Business]

Looks great, works great. Good work.  Not perfect, but time and effort solves everything.

[Mesh Open Beta Begins on October 13th]

One of the nice features of the coming mesh design is you can make your own custom template for clothing

It's a double edged sword. The downside being that without a clear standard, the market is going to fragment badly. Every creator will define their own UV mapping, and clothes from one creator will not match any other creator. You will then be tied to using only clothes from that given creator with your mesh avatar. I've seen this all too well from my years at Renderosity, and various creators making their own mesh figures (Victoria, Aiko, LaRoo,Stephanie, Poser std models, etc. ) LL's mesh may not be the best, but at least it sets a hard standard for clothing interoperability between creators. 

I fear that  standard will be lost in the coming months/years. (at least until creators learn how bad for business it ultimately is)

[Mesh Open Beta Begins on October 13th]

Not sure, but as a follow up question, can multiple items use the same bones?  Or are we going to have conflicts where a jacked wants to use the same bones as some other item?  Inquiring minds want to know.

[Farewell, Avatars United]

It's in the Facebook TOS. They may not have evicted you yet, but if anyone ever alerts them to your avatar account, it will be deleted.

[Mesh Open Beta Begins on October 13th]

When this beta starts, is LL going to provide people with specific information about Mesh and their implementation?

It's all fine and good to say:

"Rigging is is attachment based; if you import a mesh with compatible bone weight information, you will be able to use it as an attachment that will animate along with your standard avatar."

But without the documentation to explain how it's done, it doesn't help. The documentation should come well in advance of the beta, so people have time to read it, understand it, and begin implementing it.

[Farewell, Avatars United]

From first-hand experience I can say this is a big understatement. I've been on three sides of this fence. I think what most people don't realize is that often the most profound improvements get no reaction whatsoever from anyone outside the company.

 

It takes someone gifted to successfully complain about things going right. : )

[Farewell, Avatars United]

Classic example of why not to sell your company. Maybe something LL should think about themselves.

[Viewer 2.2 Beta has Arrived!]

I find it hard to get excited, as https://jira.secondlife.com/browse/VWR-19964 still isn't fixed. Come on now, bring thousands of users joy and fix this bug.

[Next Steps for Mesh Import]

 

Mesh support was announced pretty much exactly one year ago. Plenty of time to choose a 3D modeling application and get familiar with it. At least two of those applications have been available free of charge: Blender and Wings 3D. Maybe they slipped below your radar because they are made by those open-source hippies that you so despise.

Anyway, there is no excuse for creators who chose to twiddle their thumbs for the past 12 months. You either want to make money or you don't.

There is one big issue that makes your statement very fallacious.  LL put anyone and everyone involved with mesh under a NDA. So no creator was able to make any decisions about what to buy or what to learn. People haven't had 1 year. A few months maybe, and only because the NDA was dropped for 1 day by accident, and some info leaked out.

The NDA is probably the single biggest problem I have with the Mesh project. Nothing regarding Mesh should have ever been under an NDA.

Then your statement would be truthful.

[Next Steps for Mesh Import]

I see. Thanks for the additional Info. A triangle still has to have 3 vertices, so there's still a rough analogy to number of vertices as compared to number of triangles.

Personally I'm satisfied enough with the current numbers to say the world isn't ending. Barring LL making a huge change to the 4k = 1 prim ratio that is.

[Next Steps for Mesh Import]

Well, this is what i've gleaned from the people talking about it. Which considering the NDA is lifted, i'm not seeing many talk so far.  There are two key factors,  how many K of data =1 prim, and how many k of data = 1 vertice.

LL will set the first, and I think it's clear 4k per prim is not set in stone yet.

The second is completely under the control of the creator. I can think of a lot of ways to optimize the upload file to reduce the number of bytes per vertice.

So milage may vary. I'm just doing rough math to get perspective. Now if LL would give us hard numbers, we could do some very hard calculations.  : )

[Next Steps for Mesh Import]

Actually my numbers are way off.  I misunderstood about what they meant as a limit. it's actually vertices they are limiting, not upload size.

 

it's a 64k Vertice limit  (65535 for those keeping count)

but 4k of upload size per prim  (4096 to be exact)

if you make a broad assumption that 1 vertice uses 30 bytes in the file, the file size for a 65535 vertice item would be  30x65535= 1966050 bytes / 1024 = 1920k of data.

1920k of data / 4k per prim = 480 prims.

of course the actual size is probably greater, especially with UV maps.  Someone reported a 64k vertice file would be around 8.75 megs.  (8750k / 4k per prim = 2188 prims )

I feel much better now.

[Next Steps for Mesh Import]

Well, if what people are saying in the SL Beta group is true, meshes will give you 200 prims worth of vertices for the cost of 15 prims. Basically, prim limits just went out the window. New Era indeed.  I'm kind of disappointed they set the limit so low.  (4k upload = 1 prim,  60k upload limit = 15 prims).  The limit should be based on vertices, and be quite a bit higher. Otherwise using meshes will be the way to get around prim limits in a big way. And that's only going to increase lag.

Prim limits are really just a way to limit vertices, after all.