Jump to content

Penny Patton

Advisor
  • Content Count

    1,591
  • Joined

  • Last visited

Community Reputation

960 Excellent

4 Followers

About Penny Patton

  • Rank
    Advanced Member

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Here's another article I wrote that I hope is helpful for sim builders: https://pennycow.blogspot.com/2018/08/simple-performance-tricks-anyone-can-do.html
  2. Just in case anyone doesn't remember Nimble, here's a video of it in action. I brought this up to Oz a couple of years ago and he had no idea what I was talking about. He didn't seem to know what Windward Mark is either.
  3. 14 years of being in SL. 😜 Seriously, tho, I think one of the reasons LL is reticent to publish their plans for SL is that those plans are constantly changing. They'll be talking up a feature that's "definitely coming soon" one day, then the next none of the Lindens can remember that ever being the plan. Remember Avatar Puppeteering? Nimble volumetric clouds that were going to be added to Windlight? ARCtan? Remember in 2007 LL promised we'd be able to trade windlight settings as inventory by the end of the year? There seems to be a lack of direction and leadership in the Lab. Features only seem to get worked on so long as there is a Linden that feels on working on that feature. If they're fired, quit, or get distracted by a laser pointer the feature gets dropped. That's why I say "LL has been promising a lot of good things" rather than saying "LL is working on a lot of good things". Given their history, the chance that even half of what LL promised two years ago will actually see the light of day isn't exactly good.
  4. You're probably experiencing more of a performance hit because of your avatar than you realize. SL tends to run so poorly for anyone, regardless of their computer's specs, that people tend to overlook performance issues that would be unacceptable anywhere but SL, such as texture thrashing (where textures keep going blurry after rezzing because your hardware cannot keep them in memory due to the sheer amount of textures being streamed all at once) or stuttering (where your computer freezes for a moment or two every so often), let alone the fact that for SL's visual quality, anyone with a computer with a dedicated videocard made in the past 10 years should be experiencing a steady 60fps at all times, at max settings. When SL itself is such a burden on your hardware, being able to judge just how much any individual part of SL, such as an avatar, is affecting you can be difficult. That said, if you're noticing that most people have a lower draw weight than you, you probably want to reign it in a bit. It's difficult for anyone to offer any exact numbers to shoot for because avatar complexity is calculated client side, based on your own hardware. (Hey, Linden Lab, you really should fix that. ARC should be a consistent metric for it to have any meaningful use.) And it gets worse. Avatar Complexity isn't a very accurate method of determining an avatar's actual rendering complexity. ARC fails to take textures into consideration in any meaningful way, and textures these days tend to be the undisputed number one cause of performance issues in second life. Nothing else even comes close. I can tell you that you should always aim to have less than 100MB of textures on your avatar. Even 100MB is excessive for an avatar, but it's downright reasonable compared to what most people have going on these days. 200-300MB is is common, and avatars with 500MB to 1GB are not as uncommon as you'd think. (And it will only continue to get worse as long as LL continues to twiddle their thumbs and hope that the problem will just go away by itself.*) That said, all that means is that not all avatars with low complexity are actually easier to render, but avatars with high complexity are always an excessive burden on your videocard. It's not about computers having limited specs. No one is running SL as well as they should be. You could have a computer that runs most modern games just fine, but struggles with SL because almost no SL content creators bother to optimize their work to make it appropriate for realtime 3D rendering. Some even just snatch high-poly models that were never intended for realtime rendering and upload them to SL. We're talking models created for advertising or pre-rendered footage purposes. At the same time, if you take an older/less powerful computer into a sim that was made by someone who understands the need for optimization, that computer will get better performance in SL than people with much more powerful computers will ever experience in Second Life. *Linden Lab's CEO Ebbe Altberg has more or less said that he just hopes the problem will go away by itself, despite the fact that it has gotten exponentially worse over the years and there have been no sign of that changing. Dare to dream, Ebbe. Dare to dream. You might want to start looking into a plan B, though.
  5. Although, while we're on the topic, I do happen to believe that LL should invest some development in the in-world building tools. Not to compete with content made in Blender or other 3rd party tools, but to allow people more freedom to easily create simple things. Don't underestimate how much of the simple joy of creating was a factor in SL's early success. Let me use another SL feature as an example. If you know how to script you can do amazing stuff in SL. But to do even very simple things, like create a chair or a teleporter, SL requires you to learn how to script. If SL had a visual interface for simple scripted features, like adjusting sit targets, apply texture animations, and creating experience based teleporters, more people would be able to easily create such items on their own. Sure, if you relied entirely on the visual scripting interface you'd never be able to achieve quite as much as those who had a strong understanding of actual scripting, but surely you can see how the SL experience would be enriched. Just imagine if anyone, regardless of their skill level, could easily set up an animesh NPC to walk around their yard, stop to water some flowers for a few minutes, then continue around the house to sit on the front porch for a half an hour or so, then get up to check the mail. Applying the waypoints, the animations, everything via a simple to use interface. Pathfinding that anyone could use, not just people who've spent hundreds of hours eyeballs deep in the LSL wiki.
  6. Who said anything about LL making anything like Blender or Maya? You're making some crazy leaps in logic here. I'm not talking about LL making their own version of Blender in SL, I'm talking about all new features. Stuff that isn't currently possible, or at least isn't possible to do well, in SL with any tools, third party or otherwise. You're on the completely wrong track here.
  7. You've got that backwards. The question you should be asking is "what percentage of SL residents would use content created by these tools?" And the answer is somewhere in the neighborhood of 100%. Let's say someone invents a car with square wheels. Clearly it does not work. Do you assume: A. The car failed because the design was flawed. or B. Nobody likes cars. LL's attempts at new user orientation activities have always been flawed at best, completely wrongheaded most of the time. The new user experience you're describing illustrated the shortcomings of the SL content creation tool set. Plenty of us told LL that when they introduced it. They didn't listen and they failed to learn from the mistake. Instead of learning that square wheels are a flawed design, they decided that nobody likes cars. You're absolutely right on this, and I don't believe LL should allocate 100% of their resources to large projects that will take years to complete. I've always felt LL should have multiple teams. Have your A team working on those big, ambitious projects that take years to complete, while your B and C teams knock out many of the smaller features and fixes. LL has to take the long view regarding how they allocate resources, but this also means keeping a forward momentum that is tangible to the userbase. There's a lot of low hanging fruit LL ignores because too often they allocate all of their resources towards the big projects. They could have a small team knocking out a weekly or at least monthly stream of fixes and features and I think the userbase would feel as though much more progress was being made. I wish I felt the same way. I feel like I bang my head against the Linden brick wall and once every few years it moves an inch forward and half an inch backwards. In the meantime, I continue to have terrible headaches. I think they've made some promising statements in the past couple of years. Animesh is long overdue and I'm already seeing it get used to great effect. There's still a lot LL needs to deliver on, and much, much more that they need to work on but have yet to even acknowledge. On a very much related note, a year and a half ago Oz told me they were introducing new camera controls and were including my camera settings as a new standard preset....but then it never happened and the last time I asked Oz about it he seemed to have developed amnesia. I has to ask some friends I knew were at the same office hours meeting to make sure I hadn't imagined it. They remember Oz bringing it up, too. I swear there must be something in the water over there. Someone should get the EPA to test LL's pipes and watercoolers.
  8. Because the same features that would let us make games in SL would also allow us to create more engaging non-game environments. Not to mention those same "gamey" features will allow LL to make more effective tutorials, mitigating SL's steep learning curve. SL used to have a vibrant educational community, and that community would benefit greatly from such features. Not to mention the ever popular sailing and airplane communities in Second Life. For example, in one of the clubs I sometimes go to they added NPC animesh characters around the club. These characters were scripted so you could interact with them. In addition to adding to the atmosphere of the club, they'd point out club features and how to use them. Things that would otherwise be easy to overlook. "Gamey" features can also be instrumental in getting people to socialize. If you go to a club where there's nothing to do but stand around, play dance animations, and try to find someone else who's active and wants to chat you're going to be far less successful than if you go to a club and there's a working, and engaging, bocce ball game on the deck outside. (This is part of the reason bars in real life tend to have pool tables, dart boards and other games.) Also remember that the better LL's business is, the better SL's future is. So getting elitist about who LL should be marketing SL towards is self destructive. The Linden builds were my entire point. The new user experience, apotential new SL resident's first exposure to SL should draw them in, not push them away. I'm always highly skeptical of people who make claims like this. Every single time it turns out they're ignoring some pretty glaring performance issues (like texture thrashing, low fps, or graphical stutter), that they're just so used to seeing in SL that it's become background noise to them. From the computer you described, if you are not getting a consistent 60fps at max settings and never need to turn your graphics down at all for busy areas, then SL is running fine on your 5 year old PC. For how SL looks, you should never experience texture thrashing, severe fps drops, or any other graphics issues. Lowered expectations tend to be the kool-aid of choice in SL. And regarding "potato PCs", most of them should run SL fine, too. Or at least respectably, depending on just how spud-like the computer in question is. Which brings us back to that self-destructive elitism. "SL runs fine for me, so who cares how it runs for anyone else?" How are you enjoying those fee hikes LL introduced because they need to shore up their profits in the face of a dwindling userbase? You got to have a broader view of these things. Yes. Undoubtedly. How is this even a question? I agree completely that LL should be seeking multiple opinions and I also agree with some of the suggestions in this thread on LL reaching out to active and inactive users for feedback, and that LL should be making better use of their ability to collect relevant metrics. I have also read Longevity in Second Life (ok, so I only read it just now, but I have read it), it's a very interesting study and I believe it correlates with a lot of my views on how LL should be developing SL's social tools, (the study seems to show that people stick around more when they have a broader range of friends in SL with whom they interact regularly) improving the marketplace, the creation tools, and generally working to reduce confusion with regards to SL and products sold within SL (as the study also shows that as people spend money in SL they become more invested. Moreso when it's money paid to other SL users rather than LL directly. ). The emphasis being on the social engagement angle. I see the study was done in 2009, I think it would have been interesting to see what additional data they'd have gained by continuing the study through the SL bust and first year or two of the SL recession. I don't usually toot my own horn here on the forums when it comes to my credentials, especially considering that I subscribe to the "what happens in Vegas stays in Vegas" view of SL social interactions, but I majored in computer animation, multi-media and game design. I have nearly 20 years of design experience and I also happen to have been in SL since 2005. In addition I've been an art instructor, teaching all ages between pre-schoolers to adults, and in both the public and private sector. While I do agree LL shouldn't act on my opinions alone, I do hope that my background, as well as my body of SL work, adds some perspective and weight to those opinions.
  9. If you want to improve SL's ability to draw in and retain new users, you have to identify what keeps people from checking SL out, and what drives away those that do before they get "hooked". Here are the major complaints you'll hear from people who've checked out SL but didn't stick around: There is nothing to do. It's ugly. Despite looking ugly, it also runs poorly. It's confusing. Number 1 "There's nothing to do" is the hardest for LL to tackle. It requires new features and new content creation tools. I'd also argue it would mean improving the existing tools which could also fall under number 4, but I'll get into it here as it applies to giving people more to do in SL. We finally got animesh, and so the ability to create believable NPCs and that's great, but there's still a lot LL needs to do. Improve SL's physics and give us the ability to create more immersive vehicles. Sailing and flying are already popular SL activities, imagine how much moreso if we could make boats, planes and cars that all felt more responsive, more like actual vehicles. Sure, upgrading the physics to the point they'd need to achieve this might break existing vehicles, but it would totally be worth it if it meant a big enough improvement for new vehicles. Fix Search. It's very difficult to find relevant content and experiences in SL. Shop and club owners tend to toss in any popular keyword they can, whether or not it applies to their shop/club/event. How can LL curb this problem? By improving and expanding the "interest tags" feature they seem to have forgotten about. We should be able to apply a limited number of keywords (I'm thinking three, five at the very most) to land, marketplace listings and events. The limited number would encourage people to only use tags that really do apply to them. Then interest tags could be an additional factor in search results, or a complete filter so that people could search exclusively by tags Create an SL calendar. The calendar could be integrated with profiles, event listings, groups and land. When at a location in SL you could open the calendar to see upcoming events listed at that location. With the calendar open you could see all events for all of the groups you were a part of, or filter to specific groups. For events you were interested in, you could set reminders, so SL would remind you when an event was coming up. You could see account creation dates for your friends so you could wish them a happy SL birthday. Group anniversaries. Partnership anniversaries. Etc. The big thing is that we need the tools to create more immersive, more game-like experiences. We already have more such tools than I suspect most SL content creators are aware of, but LL doesn't make these tools obvious or very user friendly. The fact that one needs to learn how to use scripts to create something as simple as a chair or a teleporter one can simply walk into to use, or even an animated texture illustrates the problem. If people could create such simple scripted elements more easily, they'd be even more common than they are now. Many game engines and other software for creating interactive experiences have a "visual scripting interface". For example, there is absolutely no reason one could not build a prim chair, select it, and from a simple menu set the chair as an item avatars can sit on, and through that interface adjust the sit target and even apply multiple animations that could be selected through a dialog menu, all created and applied to the chair without the resident ever looking at a line of code. Such menu/UI driven interfaces people should be able to set up NPCs with interactions, pathfinding and scripted events. Interactive elements like a control panel in a spaceship that opens a door, including all of the lights and sound effects. You'd find much more engaging content springing up all over SL. Here is an article I wrote illustrating that last point with a comparison to both professional and end user tools in other software. Number 2 "It's ugly" is another tough one, but not impossible. The main take away should be that presentation matters. Mainly LL needs to expand it's art department to create, and guide the moles in creating, more impressive public builds, like the new user orientation sims, the welcome areas, and other public Linden managed areas. The moles are a talented bunch, but they're largely talented SL enthusiasts. With a little more guidance from professional design leads you could get a whole lot more out of them. More ambitious projects, more engaging experiences. And a much better looking SL experience. SL also desperately needs sound and music people. Why is there no SL sountrack? Why are most of the public builds almost entirely lacking in sound design? People notice that kind of thing and it detracts from their experience. And LL needs to lead by example. When LL shows something can be done, residents are more likely to pick up on that and use similar features in their own creations. More "professional polish" in all aspects of SL's design. Better UI sounds, a better looking interface. Better looking starter avatars (they've improved dramatically over the years, but they're still frequently 7' tall and often suffer poor proportions, or just looking kind of generic). A new system avatar. Brand new. (For those already chomping at the bit to complain that replacing the existing system avatar would break a lot of legacy content, I'd like to preemptively point out that SL could support multiple system avatars and allow users to toggle between them. In fact it already does. The male and female avatars use different mesh models and have different slider weighting.) For the new system avatars they should look as good as any of the mesh bodies sold currently. They should support materials. LL could even introduce a new type of system clothing, simple wearable models that system clothing textures could be applied to, similar to how the clothing that comes with the popular Aesthetic mesh body is made. And the new system avatars should have their slider weighting improved. Currently, we can make avatars between 4 and 9 feet tall, but you can only have proper proportions if you stick between 5-6.5' for women, or 5-7' for men. Go outside that range and even with maxed out sliders arms will be too short, heads will be too small, etc. And women are not allowed to be as tall as men can be. In a world of our own imagination, this is rigidly enforced by the appearance editor. LL could fix all of these issues. Number 3 "it runs poorly" is another difficult issue, but not insurmountable. People tend to misplace the blame for this, believing SL runs poorly due to an aging engine or LL's bungling programmers. This is not entirely the case. The fact is, SL runs pretty impressively well when the content rezzed inside of it is moderately optimized. The main problem is that avatars are frequently using over 200MB worth of textures apiece. Sims regularly exceed 8-10GB worth of textures. Avatar attachments, unbound from Land Impact limits are frequently excessively high-poly. Add to all that content creators who don't understand how LOD works create content that only renders properly at the highest LOD, then tell their customers to go into SL's debug settings and set Object Detail far higher than you can through the regular graphics options. There are solutions to all of these problems, I'm not going to get into them here and instead I'm going to provide a couple of links. Click here to see just how much you are being affected by unoptimized content. There's also some links in the article to tutorials on better optimizing content for SL. And here I explain how Linden Lab could hitch "new content rules" to new features, preserving legacy content while nudging the content creators and the SL resident population at large onto a trajectory that will see old, unoptimized content slowly being phased out in favour of new, optimized content over time. The "slow and steady" approach that creates the least amount of disruption for the userbase, but guaranteeing major improvements over time. In addition to all of that I'd argue an LL run content creation blog is an absolute must. LL needs to engage with the users on this topic. There's a lot of misinformation on the forums about what causes lag and other performance issues, with no real official line from the Lab itself. LL could provide layman explanations on the impact of high poly attachments and excessive texture use on framerates, rez times. Explain why "texture thrashing" occurs and how sim builders can reduce, or even entirely remove that particular problem from their own sims. LL can't just lay out new rules, or even provide new tools, and expect the userbase to just go along with it if the Lab doesn't explain why these changes are made and how the userbase benefits. Such a blog could also be used to demonstrate new features and share good content creation habits in ways that every SL user could understand, not just those who spend time reading technical documentation on the wiki. Number 4 "it's confusing" requires a major attitude change within the Lab itself. I already pointed out how LL could put more engaging content creation into the hands of a much broader range of SL users with a "visual scripting" interface. But LL would also need to make everything about the SL experience more intuitive, and do it without sacrificing the freedom and creativity we currently enjoy. A better designed tutorial that users could revisit as much as they need. I've long argued LL needs to create tools for interacting with the viewer UI, so much like a videogame a tutorial could open menus and point people directly at buttons and features, this would also open up a lot of possibilities for experience designers, but in the short term this could be done with YouTube videos LL could trigger to pop up during scripted tutorial experiences or as a part of the viewer's own help system. LL needs to also better identify what trips up SL users, new and old and figure out how to fix it. I regularly run into people who don't know about "alt+camming", some of them are people who've been in SL for years. Pushing people towards mesh bodies just to look as good as long time SL residents is a barrier both financially and in terms of easily figuring out "how to SL", so here again a new system avatar would help. The reason "display names" turned into such a fiasco is how confusingly LL made that change. I still see people using their display name like a titler, so more a lot of other SL users they have no name, just their title. These are all things LL needs to work out before rushing into a new feature. There's much, much more to this I could get into, like how the Jira is an atrocious method of getting feedback from the broader userbase and that it's crazy how the SL viewer, even projects viewers, don't have a feedback feature. I agree with Zed that SL has astounding resiliency despite its many flaws, and that with a little work LL really could breath all new life into the platform. In fact, I'd argue that until LL can figure out how to make SL more of a success, until they learn the lessons SL should have taught them, any attempts like Sansar at a new platform are doomed to failure.
  10. For what it's worth, I do agree with Coffee that expecting the average SL resident to understand why some content is bad for SL is unreasonable. However, I disagree that it is 100% on the heads of content creators alone, because most SL content creators are themselves simply SL enthusiasts who got into creating content for SL and were never exposed to proper content creation habits like you'd find in a game design course or game modding communities. This is why LL needed to step in from the beginning, and why the only way real change will happen will be if LL steps in now. LL needs to better communicate to both creators and non-creators alike how unoptimized content impacts performance. A blog laying it out in layman's terms with examples would be more effective than a wiki page in this regard. LL needs to provide content creators and consumers alike with better tools for recognizing badly made content. In the stock SL client you cannot easily see how much VRAM or triangles your avatar is using. It is especially difficult to track down individual problem attachments. It is entirely impossible to know how content sold on the marketeplace fares in this regard, unless the creator volunteers that information. LL should figure out how to encourage that in more product listings.
  11. Here again, you're confused. I'm not talking about distorting textures to maximize the amount of pixels used and I'm not saying people should be using more tiny textures. I'm talking about the fact that too many textures in SL look like this: When they should look more like this: Combined with textures from other parts of the model that are also only using a fraction of their actual texture. In SL content creators often give each part of the model it's own texture, that texture using only a tiny fraction of the pixels, when they could easily be combined into one large texture without sacrificing any detail, because they'd just be combining their textures in a way where they're not eating up 16MB of memory on four 1024x1024 textures where 3/4 of each are blank empty space and instead using a single 4MB of memory on one texture where the whole texture is being used.
  12. Let me try and reword what Da5id tried to explain to you. Let's say I rip a truck model from Supreme Taco Truck Racing XII. As everyone knows, STTRXII uses the latest in graphics so the models don't use the same normal and specular maps that Second Life uses. Because I'm a lazy Second Life content ripper, I either upload the maps that the model does use, which look terrible in SL because they're not the same sort of maps SL is looking for, or I just skip the materials altogether and upload the model without it. Because I'm a lazy SL content ripper who doesn't know what they're doing I've probably also messed up the alpha channels, too. So now this truck, which looks amazing in Supreme Taco Truck Racing XII looks kinda rubbish in Second Life. If I were a smarter, less lazy content ripper what I might do is convert the material maps from the STTRXII formats into the formats that SL looks for in normal and specular maps, and I'd be sure not to screw up the alpha channels in the original textures, this would lead to a truly Supreme Taco Truck in Second Life, comparable to a lot of other content made specifically for SL. You do have a point about screen real estate when it comes to smaller details, but you're forgetting (or ignoring) something. SL users will often upload horrendously large textures for tiny details. Like buttons on a jacket. However, unless you zoom in far closer than any SL user, even a photographer, is ever likely to zoom in, you will never notice if that 1024x1024 jacket button was replaced with a resized 64x64 version of the same texture, so it really does not result in a worse looking jacket to do exactly that. And you keep forgetting (or ignoring) that one of the most common content creation faux-pas in SL is that people auto generate UV maps that only use a portion of the full texture, meaning all that extra pixel density isn't even being used by the model, it's just sitting there, hogging up memory, while adding nothing to the model. Another thing you keep forgetting (or ignoring) is that I'm not suggesting LL needs to impose rules that are nearly as strict as what game devs hold themselves to. As I've always said, LL only needs to curb the worst habits to see major performance gains.
  13. And it's not surprising, look at how many people on this forum will argue that the textures don't impact performance at all.
  14. It is my small hope that by sharing accurate information, and pointing out when someone is spreading misinformation, I can help at least some people improve their own SL experience and drive up awareness. Possibly even enough awareness to nudge LL out of their hands-off approach to the problem as people realize just how much they're being affected. People tend to be shocked at just how much of an improvement they experience when they arrive in an optimized sim. The more such sims that are out there, the more people will see the difference firsthand. There is too much misinformation regarding this topic in Second Life. I've frequently seen this scenario: Someone has a flat normal map for multiple parts of an object. Each part using a different flat normal map. And each of these unnecessary normal maps has an alpha channel masking out around the UV portions of the texture driving the memory cost up even more.
  15. It's true, though. I've told her these same things in multiple threads before now. As recently as YESTERDAY.
×
×
  • Create New...