Chairman of the Board is on LabGab

[Quistess Alpha]

33 minutes ago, Drayke Newall said:

to build that tree you made in Second Life not outside of it.

Unless some significant changes are made in the viewer to allow temporary build caching or something like that, one of the main points @ChinRey seems to be making is that Most other non-prim building techniques would take too much bandwidth or server computation. The limited capacity of SL (limited because other things are generally going on besides a lone avatar in a region building something) is less like running an app on your phone, and more like trying to run an app on a remote server. It would be funny to see how well Blender works (if at all) through SSH.

[ChinRey]

1 hour ago, Mollymews said:

I would urge you to keep on making stuff on the OpenSim platform

That's what I'm doing but I can't honestly say I believe in Second Life anymore.

I know I've told this story here before but I once met a seasoned professional 3D modeller on the beta grid. He had just joined SL because he wanted to be creative and not only make the models his customers ordered. He came up to me and asked, before even saying hello: "Why is everything in Second Life so high poly?"

I think that says it all. SL has gone too far and there's no turning back now. Opensim is in some ways even worse although it varies a lot.

The only realistic solution is to start anew but not the Sansar way - Sansar is even heavier to run than SL. Whether there is a market for well optimized virtual world is anybody's guess though and as far as I know, nobody has tried yet.

Oh btw, Molly, I sent you a copy of that tree. It turns out it was one of those I had uploaded but not listed for sale.

Edited June 25, 2021 by ChinRey

[Drayke Newall]

43 minutes ago, ChinRey said:

But it's the method that is the problem. CSG meshes will nearly always be that "dirty" and often even worse so it is not a suitable tool for inworld building.

However, with that being said, one of the main ideas behind the prim system was to emulate CSG but with much simpler algorithms and it actually does that amazingly well. Back in 2002/2003 they had to keep it very simple since computers weren't nearly as powerful as they are today. But it is possible to expand on the basic principles with more profiles, more paths and more "prim torturing" parameters. Even a few simple additions - like the hexagonal profile you hinted at - would greatly increase the possibilities of prims with only a fairly small increase in the computing load and I'm convinced that would be the way to go to create a modern, flexible, user friendly and high performance inworld buiding tool.

As I said it is where LL need to head to. How they do that, whether it is using such distortion methods or introducing another method that is suitable for the prim who knows. In world creation needs to come back fully to SL as it is something that even now people new to second life expect and it reduces the learning curve.

It is the defining factor of why Second Life became popular.

43 minutes ago, ChinRey said:

It wasn't a good example because neither mesh, scultps nor prims are really suitable for making plants. You really need a special algorithm and asset class for that. This is why LL added the system plants. Unfortunately this is something that has been even more neglected than prims and what was cutting edge in virtual plants in 2003 is sadly outdated today.

Which is another downfall of their lack of updating. As I mentioned earlier. What LL achieved in 2003 was revolutionary. It is those revolutionary features however, that have been neglected and is now a matter of too little too late.

42 minutes ago, Quistessa said:

Unless some significant changes are made in the viewer to allow temporary build caching or something like that, one of the main points @ChinRey seems to be making is that Most other non-prim building techniques would take too much bandwidth or server computation. The limited capacity of SL (limited because other things are generally going on besides a lone avatar in a region building something) is less like running an app on your phone, and more like trying to run an app on a remote server. It would be funny to see how well Blender works (if at all) through SSH.

Yes, LL need to introduce new caching systems. It was at one stage on the road map but that was years ago and it got no where.

[Drayke Newall]

5 hours ago, Mollymews said:

also too I notice that Patch Linden mentioned at SL18B that Linden are going to make a Avatar 2.0. A plain all-in-one mesh body that is Bento/BoM compatible as the starter avatar for new people.  As a way for new people to easier transition to more advanced user-made bodies/heads/etc. Which I think is a good thing  

Wow only took them 10 years of people asking for it to become a reality with LL excuses of they will never do it because they dont create content. Wonder what has changed, has the new board members pushed for this? Would be interesting to know who made the change. The trouble now is that I have a feeling it is to late. If it was done when requested in 2010 it would have been good. Now though... I suppose only time will tell.

I would say also this is a good thing but only if they make the body equivalent to what is out there already. Doesn't need the features or huds etc of the user creator bodies just the mesh, bento and BoM but needs to look as good. If that doesn't happen it will be DOA.

[Mollymews]

1 hour ago, Drayke Newall said:

Would be interesting to know who made the change

 

if i had to guess then I would guess Patch Linden as Vice President of Product Operations has been driving it. Is a guess as I dunno actually

this is not to say that other Lindens haven't been wanting to do this, I think they have for quite a long time. Just that they needed a senior person to drive it forward, it being just one important thing to be driven amongst all the other important things still to do

also as well I think the timing is right. The outcome of Belli is that there is a known/measurable demand from paying SL customers for turnkey products. A mesh compliant starter avatar being a turnkey product. And it also fits with the current Linden initiative to grow the paying new resident/consumer space

 

 

Edited June 25, 2021 by Mollymews
emasure

[Mollymews]

1 hour ago, Drayke Newall said:

I would say also this is a good thing but only if they make the body equivalent to what is out there already. Doesn't need the features or huds etc of the user creator bodies just the mesh, bento and BoM but needs to look as good. If that doesn't happen it will be DOA.

time to speculate. Reading between the tea leaves

i think this starter avatar is going to be about new residents. The most asked question by new residents since 18 years is: How do I change my avatar ?

and today that is not how do I change this classic avatar I am wearing to another classic avatar. The question is how do I change to a mesh avatar same as everyone else

so I think the starter avatar will be all in one (head/body as one piece). Bento/BoM/Rigged and Modify (Shape slider reactive). It will also be PG I think. No meshed nipple bumps or crevices with only two feet positions - flat and high. Basically a much nicer sculpted version of the classic avatar

new person wearing starter avatar when asking the perennial question: How do I change ? will get told about how to change using BoM layers and rigged clothes/hair/attachments/etc. Which will make their transition to other bodies and heads a whole lot simpler. As this latter takes the same viewer interactions to do. Interactions they learned with the starter avatar

 

Edited June 25, 2021 by Mollymews
typs

[Qie Niangao]

3 hours ago, Mollymews said:

so I think the starter avatar will be all in one (head/body as one piece). Bento/BoM/Rigged and Modify (Shape slider reactive). 

I would like to see it 100% open source / public domain, and then see Slink or another reputable Mesh body adopt the geometry completely for their next generation of avatars. Sure, there's an argument to be made that the diversity of body models creates consumer choice, etc., but it also creates Stupid Incompatibilities. I would gladly sacrifice any marginal advantage of any of the avatars for a way to use the identical clothing models on all of them.

Give the clothing creators some incentive to perfect their models instead of spending all their time rigging to each pointlessly tweaked body geometry.

[Phil Deakins]

6 hours ago, Drayke Newall said:

In world creation needs to come back fully to SL as it is something that even now people new to second life expect and it reduces the learning curve.

It is the defining factor of why Second Life became popular.

I haven't read this thread, but I've said for years that LL destroyed the fundamental nature of SL with user-mesh. So I couldn't agree more with you. I'm not against mesh at all. I'm against it being necessary to make good stuff outside SL.

[Innula Zenovka]

While I don't know about creating things with mesh, I do have some experience in scripting.  I started out back in 2007 using the viewer since that was all that's available, but for the last ten years, at least, I've done almost entirely in external editors (Sublime Text for most of them).

That's partly because external editors are generally easier to use than the internal editor because of their additional features (code completion and so on) but also because when I'm writing a script of any complexity, that and the wiki are I'm concentrating on, and I don't want to be distracted by what's happening around me in world.

Then when my script is written and I start testing and revising it in world, I keep on copy pasting back and forth between my text editor and the viewer's script editor, because the same applies.    When I'm working, I'm working, and don't want to be distracted (that's why I work on my laptop at home rather than from a coffeeshop or bar, because when I'm in one of those, I'm not there to work).

I would imagine that serious mesh modellers, too, want to use/learn to use powerful tools like Blender and Maya,  and, like me, need to concentrate on the detail of what they're doing without being bothered with IMs and declining unwanted offers of tps or promos.

So I am, I fear, somewhat sceptical about calls for better in-world tools, since I'm really not sure that they're needed.   People aren't put off scripting by the necessity of learning  complex UIs, as they may be with 3D modelling software.     They don't learn because that's not what they're interested in doing, which is why I've never bothered to learn Blender or animation software, despite the fact I'm very interested in animating both avatars and objects -- I don't have time to learn the software because I'm more interested in doing other things in SL, so I prefer to work with people who have learned animation and 3D modelling but don't have time to learn how to make their creations interactive.

LL don't need to worry about recruiting would-be content creators -- there's always going to be plenty of modellers, scripters, animators, texture artists and other creators to supply residents with all their needs, to my mind.  

To my mind, LL should be concentrating on attracting people who aren't currently drawn to SL, but might enjoy what the platform (and our creations) have to offer and it's not at all clear to me that they'd be attracted by better in-world building or scripting tools.     

 

Edited June 25, 2021 by Innula Zenovka

[Coffee Pancake]

Body 2.0 MUST be 100% open source and come with documentation and development kits.

There MUST be an iterative open development process, not "hey everyone what do you think of this? - don't bother responding, we've actually finished the project / run out of time / money".

It MUST be developed in and around tools that actual SL creators have access to, aka blender. Not Maya (which costs $1700 a year).

It MUST not be a simplistic / gateway body that gets junked in the trash just as fast as the current body / starter avatars.

It MUST work with viewer UI. Preferably be rolled in the viewer with the same level of integration as Body 1.0

[Quistess Alpha]

19 minutes ago, Coffee Pancake said:

It MUST work with viewer UI. Preferably be rolled in the viewer with the same level of integration as Body 1.0

IMO just being able to swap the default body between different mesh modes (perhaps some viewer option or config wearable(a new addition to shapes perhaps?) to have your default body look like avi 1.0 or 2.0, with 2.0 default for new residents) and being able to apply materials to the default body would be enough. (assuming most of Coffee's other stipulations)

[Sid Nagy]

6 hours ago, Innula Zenovka said:

To my mind, LL should be concentrating on attracting people who aren't currently drawn to SL, but might enjoy what the platform (and our creations) have to offer and it's not at all clear to me that they'd be attracted by better in-world building or scripting tools.     

 

Is it really an all clear that they do come to pay, pay and pay again to get established in SL and that there is not much more to do than rinse and repeat the paying and a bit of the adult stuff?

The best years in SL for me were when there was tons of building action in the sandboxes. Where you talked to each other, learned from each other. All had the same materials at hand to build, amateur or pro.

Nowadays if your not a graphic artist or a coder you better bring a filled wallet to enjoy SL.
25 USD per month is nothing, It gets you a small parcel, some furniture and some clothing, hair and a body (within a few months)
Is that really so appealing for the masses to come and explore mainlands dessert in all its ugliness (apart from a few lovely places) and all of SL's notorious flaws?

 

 

Edited June 25, 2021 by Sid Nagy

[Drayke Newall]

3 hours ago, Innula Zenovka said:

I would imagine that serious mesh modellers, too, want to use/learn to use powerful tools like Blender and Maya,  and, like me, need to concentrate on the detail of what they're doing without being bothered with IMs and declining unwanted offers of tps or promos.

It is not about catering for serious mesh modellers. It is about allowing average joe mesh modellers the ability to compete with those 'pros'. No one would argue for LL to stop allowing third party program creation all we are suggesting is to have a competitive in world creation system for those that are just starting out.

There are many a story where new people used the inworld prim creation system as a stepping stone or their introduction to mesh modelling as it is simple and gave them an eye opening that they can easily make lindens to play the game without a learning curve a mile high.

Quote

To my mind, LL should be concentrating on attracting people who aren't currently drawn to SL, but might enjoy what the platform (and our creations) have to offer and it's not at all clear to me that they'd be attracted by better in-world building or scripting tools.     

And how well has that been working out for them? They have been trying that for years and have seen no growth at all unless you believe the LL propaganda machine. Concurrency is near where it was pre pandemic and dropping.

There is also an uncanny oddity when you look at user concurrency and the introduction of external mesh creation and the redundancy of in world creation. Concurrency was at hype levels of 60-80k pre mesh. As soon as mesh took a hold and started to dominate sales concurrency dropped to 50k and below. Now that might be a coincidence but it certainly is odd.

Edited June 25, 2021 by Drayke Newall

[Innula Zenovka]

11 minutes ago, Sid Nagy said:

Is it really an all clear that they do come to pay, pay and pay again to get established in SL and that there is not much more to do than rinse and repeat the paying and a bit of the adult stuff?

The best years in SL for me were when there was tons of building action in the sandboxes. Where you talked to each other, learned from each other. All had the same materials at hand to build, amateur or pro.

Nowadays if your not a graphic artist or a coder you better bring a filled wallet to enjoy SL.
25 USD per month is nothing, It gets you a small parcel, some furniture and some clothing, hair and a body (within a few months)
Is that really so appealing for the masses to come and explore mainlands dessert in all its ugliness (apart from a few lovely places) and all its notorious flaws?

 

 

Most of the masses who put US$ into the in-world economy by buying L$ aren't here to explore, though.    They're here to enjoy and create their own little niche in SL, by renting or buying land, or Linden Home, and make their own environment.   Or buy mesh bodies and clothes and design their own avatars,  or hanging out with friends in clubs and so on. 

[Lucia Nightfire]

8 hours ago, Qie Niangao said:

I would gladly sacrifice any marginal advantage of any of the avatars for a way to use the identical clothing models on all of them.

Give the clothing creators some incentive to perfect their models instead of spending all their time rigging to each pointlessly tweaked body geometry.

LL needs to invest in an in-world feature similar to Marvelous Designer's Auto Fitting like they did in Sansar which, IMO, was one of the better features of that platform.

That would cut out the need to rig & weight a shirt/dress/pants/jacket/etc. to one body, upload, then have to repeat the process for 4, 5 or 8 of the most popular bodies on the market.

When a clothing maker only needs to just design the clothing on a template and upload and let the end-user's viewer auto-fit the clothing to the body they're wearing, it cuts out a tremendous amount of production time and lets them focus on outputting more clothing as well as allows newly aspiring or less experienced clothing makers to enter the market.

[Silent Mistwalker]

1 hour ago, Drayke Newall said:

competitive in world creation system for those that are just starting out.

Would be very helpful to those of us who have little to no spare cash lying around, too. Sure there's Blender and  GIMP but they just don't quite reach the same level. Take Photoshop for instance. You can no longer just purchase the software, install it, activate it and run with it. Now you have to pay for a subscription. Yet, by having PS, you have cheaper access to other companion Adobe software like Lightroom. You don't get that with GIMP (that I'm aware of).

LL needs to "level the playing field" if they want to retain more people. 

[ChinRey]

6 hours ago, Innula Zenovka said:

I would imagine that serious mesh modellers, too, want to use/learn to use powerful tools like Blender and Maya,  and, like me, need to concentrate on the detail of what they're doing without being bothered with IMs and declining unwanted offers of tps or promos.

I'm not going to repeat what Sid, Drayke and Silent said but there are several other good reasons for an upgraded in-world procedural building system too.

---

My scripter alt usually works the way you do, writing in an offworld editor copying and pasting back and forth. It's not nearly that easy with mesh. Every time you make the slightest change you have to reupload. That's quite a lot more cumbersome than a simple copy/paste operation and unless you are one of the lucky ones who has access to the beta grid you have to pay the upload fee too.

---

There are no mesh modelling equivalent to external script editors. Yes, there are tools like Maya and Blender that can do the job but it's not what they were made for. It's like using a spanner for a hammer: you can drive in a nail with it but it's far from ideal.

---

One of the biggest issues with mesh is LoD. Even professional builders with tons of experience from other platforms have problems dealing with SL's peculiar LoD system and the way it affects land impact. The result of this are butchered LoD models, ridiculously over-strengthened LoD models, obesely inflated LoD factors and heavy custom textures for details that would both have looked better and caused less load if done as geometry. With procedural building materials like prims it's fairly easy to create procedures for automatic LoD model generation. With polylist mesh, it's very difficult. Look at the picture below these are low LoD models for the same twisted torus generated different ways. Left to right: the prim's autogenerated one, the uploader's default, the uploader with settings changed to match the triangle count of the prim, Blender's decimate modifier set to match the triangle count of the prim and Blender's limited dissolve function set to match the triangle count of the prim:

This isn't even a particularly bad case, we've all seen much worse.

I've already explained the reason for this in a different thread, so I'll just link to that post rather than repeat myself.

(Edit: Maybe I should mention land impact in this context too. I didn't try but I don't think even I could make a spiral like this as a 1 LI mesh without butchering the LoD models. As a prim it's not even difficult.)

---

Then there is the eternal issue of lag and load time. See if you can still find a complex all prims build in SL (Vintage Village at Verdigris may be a good one even though it has some sculpts and meshes too these days). Then check out a modern all mesh build you don't have cached. Notice how much faster the prim build loads and how much higher the fps is there.

This is because prims require a much smaller and easier to handle dataset than the same shapes would need as polylist meshes. This more than compensates for the inevitable superfluous geometry (it would even if SL mesh makers tended to bother culling superfluous geometry). According to Avi Bar-Zeev himself, this is the reason why LL developed prims rather than implement mesh in the first place and the argument still applies today.

---

The problem with prims is of coruse that they're not flexible enough for all the things we want to build today. That can be fixed though, as I've already said earlier in the thread. There will always be objects that can only be done effectively as polylist meshes but there is a lot to gain if we could shift mesh from being the main building material to being a supplement used only for those special cases.

Edited June 25, 2021 by ChinRey

[ChinRey]

6 hours ago, Innula Zenovka said:

LL don't need to worry about recruiting would-be content creators, to my mind -- there's always going to be plenty of modellers, scripters, animators, texture artists and other creators to supply residents with all their needs, to my mind, because SL already has plenty of offer us. 

There is one comment I have to add to this even though it's a bit off topic.

New content creators do not have access to the beta grid. That means they will have to pay for each test upload they do. How is this going to affect SL in the long run? WIll they give up on testing their builds, will they give up on buiding for SL or are they willing to pay that extra (and completely unnecessary) expense? Only time will show.

[Janet Voxel]

I think you can log into the beta grid again, you just have to log into the jigglypuff (hurr hurr pokemon) region.

AFAIK, they're still working on it, but you can get in there with the jigglypuff workaround.

[LibGwen]

2 hours ago, Drayke Newall said:

There is also an uncanny oddity when you look at user concurrency and the introduction of external mesh creation and the redundancy of in world creation. Concurrency was at hype levels of 60-80k pre mesh. As soon as mesh took a hold and started to dominate sales concurrency dropped to 50k and below. Now that might be a coincidence but it certainly is odd.

That's because the makers of goods have decreed that we can no longer wear each other's clothes.  Pre-mesh we could sculpt our bodies to look like we wanted and wear any clothes we wanted if we were able to put the work into tweaking a copy to fit however we wanted to look today.  Now we have to accept a narrow band of stereotyped clothing and stereotyped body proportions or hunt the old stores for what we really want--or be shut out.  And often be shut out *because* we insist on wearing what we really want.  A person has to be pretty artistic or pretty stubborn to continue to do things their way in the face of such exclusivity.

 

Edited June 25, 2021 by LibGwen

[ChinRey]

1 hour ago, Janet Voxel said:

I think you can log into the beta grid again, you just have to log into the jigglypuff (hurr hurr pokemon) region.

AFAIK, they're still working on it, but you can get in there with the jigglypuff workaround.

You can log on with difficulty if you already had a beta grid account before LL broke it. I always log on to Morris there. Usually I have to try four or five times before I get through.

But I was talking about new content creators. It hasn't been possible to create a new account on beta for more than half a year now so if you didn't have one already, you're out of luck.

As for LL working on it, I believe it when I see it. Maybe they'll come up with a fix eventually but probably not this decade. If Linden Lab had actually cared about content creators, they would simply have created a small separate mesh testing grid with only three or four sandbox regions. It would have taken one of them a day or two to set up though and the hosting cost might well have amounted to a hundred dollars a year, maybe even close to 200, so it's obviously way too much to ask for from them.

Edited June 25, 2021 by ChinRey

[Janet Voxel]

12 minutes ago, ChinRey said:

As for LL working on it, I believe it when I see it. Maybe they'll come up with a fix eventually but probably not this decade. If Linden Lab had actually cared about content creators, they would simply have set up a small mesh testing grid with only three or four sandbox regions. It would have taken a day or two to set up though and the hosting cost might well have amounted to a hundred dollars a year, maybe even close to 200, so it's obviously way too much to ask for from them.

Ok...just breathe.

[Silent Mistwalker]

8 minutes ago, Janet Voxel said:

Ok...just breathe.

 

She did. You just missed it. It was right before that last paragraph. 🤭

[Extrude Ragu]

On 6/22/2021 at 4:14 AM, Jaylinbridges said:

 

 

23:50

on AWS

Quote

...AWS is more expensive to run than the way we used to...

Quote

... That decision was not made for financial reasons. That decision was made to have a better experience for the residents...

I seem to recall being told the move to AWS would reduce running costs for LL (And in turn, bring down the cost of tier for residents)

I don't remember my experience improving during this time either. I do know that between then and now my group chats have ceased to be usable, the beta grid becoming near-impossible to log into, and scripted objects incurring a new 3 second delay before they work increasing general 'clunk'.

What part of the experience has supposedly improved since the move to AWS?

 

[Mollymews]

12 hours ago, Qie Niangao said:

I would like to see it 100% open source / public domain

6 hours ago, Coffee Pancake said:

Body 2.0 MUST be 100% open source and come with documentation and development kits

 

It MUST be developed in and around tools that actual SL creators have access to, aka blender. Not Maya (which costs $1700 a year)

agree that Body 2.0 should be public domain, and Linden should enable the viewer to:

1) right click on the Library version. Press Create New. Which copies the Library body to Body Parts folder with the resident as Creator with full permissions

2) The viewer allows export (resident being the creator).  Export to .dae

resident can then take the weighted model into their mesh tool of choice. Rig clothes, hair, etc for it. modify the body itself: wingies, anthros, etc. sexy times body parts as well

then upload their work back into SL as creator legitimately

agree that Linden should commission a Blender Import and Export add-on to handle this. I would also suggest leaving professional tools like Maya, etc to the professionals to make/secure their own import/export add-ons for those professional products