Drayke Newall

1. Your name has been added to the Goblet of Fire 2. A very competitive game played between forum users of who can 'say aah' the most 3. Let me get my monocle to make your post look smarter 4. My more contagious form of pink eye 'forum blue eye' is playing up again.. thought I would let your post touch it 5. A time killer game of Like goes on, like goes off... Like goes on, like goes off... Like goes on, like goes off...💡

The problem is that it adds a layer of complexity to an already complex avatar dressing/creation system. Whilst yes, it is a matter of just wearing the included alpha cut it isn't just that simple. For example, some times people made clothing layers that crossed over the bounds of the LL allowed zones. A shirt that went longer over the pants meant that if you bought some pants with a no-mod alpha layer, the shirt lower section was alpha'ed as well. Not very new user friendly. Then add to that the further complexity of avatar modding - eg. A furry avatar, the alpha cuts provided for human avatars will not necessarily work for them. Then take bodies that have their own uv mapping system such as Regalia etc, those alpha cut layers wont work for them either. This is where segmented alpha cuts on bodies work better. It is also why people still use this system over alpha layers. Alpha cut layers aren't a be all and end all fix and the issue itself comes down to how LL decided to do BoM. Rather than tweaking the old baking system, they instead used the same system of large BoM layer segments (Upper Body, Lower Body, Left Arm, etc.). They also didn't allow for auto-hiding when mesh is worn within the BoM system. Making it harder and not user friendly for everyone. What LL should have done is segmented the BoM layers into smaller parts similar to mesh bodies and then have them under a grouping. So for example, you would have BAKED_LEFTARM as the group heading and then under that heading, smaller BoM segments such as BAKED_LA_NAILS, BAKED_LA_FINGERS, BAKED_LA_HAND, BAKED_LA_WRIST etc. By doing this it would allow creators the same customisability as the mesh alpha segments, colour systems etc, already used but with BoM and without the need for people to create their own alpha layers by uploading a custom alpha unless in unique situations. With smaller segments all the creator would need to do is the following: Inventory>Create> New Clothes>New Alpha and then simply select from a list of tick boxes next to the parts you want to hide. Save as alpha and done. No estimating where the alpha should stop on a texture in phtotoshop and no need to upload the texture. Just a simple inbuilt menu system that uses a default 8x8 masked alpha texture from the library to alpha each segment selected. Then they should have also made it so that a mesh object can be linked to BoM alpha segments so that when the mesh clothing/object is worn it auto-hides like current bodies do with their auto-hide scripts. Then no need for individual alpha layers in every folder of your inventory. You also have further customise options with such a system like removing another part of the mesh body hud, finger nail colour options. Just select BAKED_LA_NAILS and change the colour in the viewer without needing to go to photoshop, load the uv map, find where the nails are and then start painting. The lack of customisability of BoM etc, and all the extra steps needed is what makes optimisation for a creator and user impossible. If LL gave better inbuilt creation systems the older systems would die very quick.

Young Henry enters the room with a bucket of popcorn...

I'm not going to get into the whole why a community focused on a race was first as it will just add more fuel to this dumpster fire of a thread. As far as I can see from their announcement, there is literally no purpose to these community pages other than to give themselves a pat on the back for making Second Life *HIDDEN* communities. Before everyone jumps up and down or throws stones etc., and says its because of PR or retention or whatever and has a purpose, think about this. They have announced a new initiative in advertising different communities in Second Life be it RL or SL yet, no person (user or non user) can actually access that community page from their website. In their announcement they have stated that these community pages "live directly on their website"... um... where exactly? No link on the home page, no link in any sub-page. Hell, there isn't even a link if you login to the website and go to the community tab. Literally what is the point of such a thing if it isn't highlighted outside of a featured news article (no one reads and lost in a day or so due to other news articles) or an obscure and specific search engine search (no one will do)? Before people say "its new and they will probably make a link". I'm sorry that ship has sailed. No person or company is that inept (apart from LL it seems) to make an orphan webpage hidden on a website and make a huge fuss about it as a new feature. I think @Paul Hexem was incorrect in his post ("even when it's not LL- who is almost certain to screw it up.") implying that LL will screw it up. They already have by making it not accessible anywhere on their website. Come on Linden Lab. If you are going to make a new webpage at least send your web designers to a school to learn that your new webpage 'community initiative' actually needs to be accessible SOMEWHERE on the main site upon release. 🙄

I know SL can never run at a AAA game level, it is fundamentally different with user generated content, but it should at least try to ensure optimisation takes place since they are not going to rewrite their engine anytime soon. LL focus on one aspect and not the overall picture. I also get Coffee's point and agree that a lot of the restrictions need to be improved as far as content creation goes, but they have to be done within reason so that it stops the abuse that is bound to happen. Sure as Coffee says, increase the texture size allowance to 2048 - I agree with that. What I dont agree with is increasing it over all with no extra limitations put in place as all this will do is increase the abuse overall. A render engine can only do so much without a rewrite. Bethesda found this out with Skyrim when they saw that they could not introduce 2k or 4k textures without causing lag. They saw the effects of this with modders creating content that literally induces lag or crashes the game. This is why they are rewriting their engine all over again for their next elder scrolls release. SL is the same, you cant arbitrarily put in new higher graphically demanding content systems without affecting performance overall as people will take it to extremes just like they did with the mods in Skyrim. Impossible as it is, what really needs to be implemented in such a situation is a dynamic scaling texture system. One where if someone uploads a 2048 texture, that texture scales with the size of the object. So if it is small it renders at 128 and if it is large it renders at 2048. Well not sure I would call it cute however, there are serious flaws in the viewer fps wise that should have been fixed years ago. You have proven what I said about windows and fps graphically so thank you. I dont know whether it is because they are running on OpenGL or whether it is something else like HD images for those viewer components but surely LL are aware of the windows taking up fps? That said, looking at the firestorm and SL reduce lag wiki pages they dont talk about closing all viewer windows to increase fps so perhaps they dont. Even odder is that if the chat window is closed and not on the screen and someone talks, your chat bubble button on the bar blinks. Even this minute little detail reduces fps. Nice viewer btw.

Please quote me where I said that adding more faces to a mesh was not a good thing? In fact I even said it would be a good thing for larger mesh or more complex mesh. The problem is that whilst yes avatars are a huge dynamic in the performance side of things the avatar body themselves isn't the only mesh component and hence why creator optimisation is equally if not more important. Even if avatar mesh bodies where to have only one mesh layer allowing the alpha to work on just the one layer it is still a small fraction of the fix to the problem as far as optimisation and performance goes when avatars are additive in nature. You will fix the body with the additional faces but when people add on highly complex unoptimised clothing, hair, jewellery, glow, shine etc then the problem isn't fixed. Add onto this people wanting even more attachment points and even more animesh attachment points for premium plus and it all adds up making the improvement on the body being null and void when content that doesn't require additional faces are unoptimised and added to the avatar. I have a mesh hair piece that reduces my fps by 30 frames. How is your body mesh fix going to improve on that mesh hair? Your gain of 5 frames from your mesh body improvement is negligible in comparison. You are missing the point entirely I am making as you are fixated on avatars and nothing else. As I mentioned an unoptimised mesh built sim can reduce your FPS by 30 frames over an optimised mesh sim (including textures). For example, how does fps per avatar have anything to do with a person going to Belli and lagging around despite them being the only avatar in view or even on the region? It doesn't help those people at all and the only thing that will help is content creation optimisation. How you are missing this point I am unsure. I get over 100fps (draw distance 100 and no shadows) whether I have my Jake body (one of the worst offenders atm) on or not. If however I go to an unoptimised sim full of crap content with no avatars around my fps plummets to under 50. But as you say fix the avatar bodies or add more texture faces as that is going to solve everything...🙄 A increase of 5fps you are talking about is not massive at all when a sim can reduce your fps by 30-50fps. I agree 100% with you, but it irks me when people say 'fix the avatars and it will fix lag' or 'content optimisation plays no role in lag'. I mean this is just basic stuff that a lot can be fixed by simply having LL give tutorials on their wiki site or in viewer expandable hover tips. The issue doesn't just stop with avatars or content optimisation or even the render engine. Sure as Coffee said you can gain 5fps by having extra faces on mesh, or you can improve fps by updating the render engine, but all that pales in terms of viewer fps loss. Go to options in firestorm, add the fps counter to firestorm and see how many frames each of your viewer windows take up. In no AAA game does inbuilt menu's or chat screens that are part of the interface reduce FPS when they are opened. It is unheard of in final released versions. Here in SL though they do. Open the chat window, loss of 5-10fps, open the Inventory window, loss of 5-10fps, open the Outfit window, loss of 5-10fps. The same fps loss is also shown by fraps so it isn't a inbuilt fps counter issue.

Did I say that there isn't lag from other areas? No. Of course there is going to be lag from the old code or improvements than can be made to the engine systems to reduce lag. I do not disagree with you, but stating that content optimisation is not the main issue is just silly. Yes SL isn't a game... but, it still has a render engine though and whilst yes that render engine is limited such as users needing creative ways to overcome some of those issues like for instance darkness in a tunnel, there is no difference as far as content is concerned or the restrictions to a game. Even in AAA games, content is made based on the render engine and its limitations. If the render engine can't do x, then they ensure to optimise content to do it. Have you ever looked at the assets of a AAA game? Do you think a AAA game maker uses more than a few texture faces per object? Most games even modern dont exceed 8 faces per mesh and even then 80% of content uses 1 texture face across the entire object unless that object exceeds a certain size. That content over a certain size such as a house is made from multiple meshes just like SL content. i.e. outer shell, inner shell, roof and interior details such as fireplace or objects. The difference in SL is, that content is created by amateurs that think that there is no issue in uploading highpoly content (like you do) such as the phone example mentioned earlier in the thread and use 1024 textures on it. In a AAA game that phone will have 12 polygons (triangles) and a 128x128 (if that) texture and a normal map to take care of the finer details. Conversely in SL that phone will have 11000 polygons and a 1024x1024 texture to all 6 sides because the mesh allows that many textures and their is no poly count to size ratio restriction. LL gave such freedoms to the detriment of performance. Sure having more texture faces will be good for large objects but it is still a catch 22. At the moment 8 texture faces limits the amount of individual textures on an object. It is a restriction that is there to ensure that texture use is limited. It is designed to force UV mapped textures and even now people abuse that amount of texture faces and that isn't my words it is Oz's. Do you really think LL adding more faces to a mesh is going to stop people making high poly and unoptimised content? All that is going to happen is people are going to create the same high poly content and use more textures to make the item look 'even better' and because they can. Thinking other wise is naive. You are thinking only in terms of the avatar body and how more texture faces will improve that. In SL such thinking is bad as it leaves open unforeseen consequences in other content creation areas. I think Oz Linden says it all in this quote of his: "What kind of challenges are there in second life that prevent it from having the high end graphics and speed of other modern games like shadow of the Tomb Raider and others. Well the real answer is that it’s they don’t have the same challenges we do; we have user-generated content when you’re looking at a at a triple-a game the developers have built a rendering engine for that game. ... And they’ll give the artists who are creating the content who are you know building out the dungeon, or whatever it is that’s in the game, they’ll give them very strict rules; ‘you can only have so many triangles in a scene period’. You have a budget. You can only use transparency or for translucency on you know X percent of the scene. You are not allowed to have anything transparent in front of anything else that’s transparent. That kind of thing. And the results is that they can then get extremely good performance. ... We don’t have that luxury because all of the residents are creating the world and we want them to have as much flexibility as we can afford. But the result is that they do things like having partially transparent lab gab things floating up in the middle of the sky above the presenter, and translucent windows that are looking out onto the things outside… I’m just I’m marvelling at and I do this every single day. I marvel at what people build I mean it’s just incredible - it’s beautiful but, it’s not predictable and it’s often not very well optimized. …. it gets used and it gets used in ways that we didn’t think could even work but it do work and it’s just fun to watch but it’s a challenge to build something Optimal… Will we improve graphics performance over time? Absolutely we will. We have in the last year hired a couple of really good graphics people and they’re doing that’s what they’re spending time on and I expect good results. Will we get to the level of performance of a triple-a game? Definitely not. It is just not possible; not and have the creative scope that we that we want our residents to have." –Oz Linden, Second Life’s Lab Gab Episode 20, 4/3/2020 emphasis mine. Even Oz agrees that content is the issue in why performance isn't equal to AAA games per the OP. There has to be restrictions and those restrictions in place even now, according to Oz are not enough to help as they want creative freedom for residents but are trying to balance that freedom with optimisation. Sure they can optimise the engine and the rendering pipeline which Oz stated they are, but in Oz's own words "Will we get to the level of performance of a triple-a game? Definitely not. It is just not possible". It isn't possible because users do not optimise their content, they do not think about how an alpha over alpha reduces performance, how multiple textures drain FPS, etc. Sure SL is different as everything is user created and LL want flexibility in design but thinking that Lag or performance issues are not related to content creation is ridiculous and thankfully Linden Lab dont agree with you!

The problem isn't that we disagree which one is worse. The problem is that of late Coffee has been suggesting not that both are to blame but giving the opinion to content creators that not optimising their content is perfectly fine. That to me is a not a good thing especially when such words come from a TPV creator. Words such as this: "but no ones SL is being made worse because some odd bod bit of tat has too many triangles, or because someone pushed the mesh uploader too hard and the thing looks bad." That entire statement is false as it is affecting performance and users enjoyment. That is all in this thread I have argued against. I have not in this thread or any other thread suggested that either one doesn't matter or that LL don't share a lot of the blame. On the contrary, people here by now should know where I stand with LL and their lack there of. That said Coffee's suggestion that content creation optimisation is merely a trope I do have a problem with. Let's not forget that whilst yes, the rendering process does cause lag, optimisation of content is the only fix we have at the moment and by optimisation of content it makes that rendering process run quicker and smoother. But the mere suggestion that polycount is arbitrary or not important is absurd no matter which way you look at it. The latest beta viewer is on its way to fix a lot of that rendering process lag, but even then I can guarantee that there will still be lag and the usual 20-50fps drops we have now from poorly made content. I'll just therefore finish with I agree to disagree.

If that was the case then there would be no reason for any optimisations at all for mesh. Basically what you are saying is that LL's efforts in reducing polygons through the mesh upload process, LOD, ARC, occlusion, etc, etc is pointless as 'who cares what the polygon count is as a PC's can render it no problem at all' so dont blame the creator. This all despite evidence of users experience with high poly or large amounts of unoptimised mesh reducing FPS saying otherwise. Whilst yes there is some lag caused by the render pipeline as well as the badly made cache, lack of hardware support, multithread, etc, that does not excuse creators of not optimising their creations as it does have an impact just like multiple textures can have an impact rather than minimal textures. The excuse you have made does not explain how an optimised mesh build on a sim can have hardly any effect on a users FPS whereas an unoptimised mesh build reduces FPS to in cases single digits even if it is all in the cache and loaded. It also does not explain how a single high poly small object, at a specific point on the sim, can reduce FPS when a person looks at it and increase FPS when a person doesn't look at it even if it is all in the cache and loaded. The only explanation for such things is due to content creator's bad optimisation of their builds.

Whilst I agree that modern computers and engines can render more and SL needs to make use of modern hardware, there is no situation in both SL or any modern game that a phone model requires 11000 polygons. Trying to argue otherwise is just ludicrous. Even today most character models in modern games are between 10,000 to 100,000 triangles and nothing more yet, you are saying that a content creator in SL isn't to blame for lag due to wearing on their avatar or having in their home a 11000 tri phone (literally a cube with tiny buttons) more than what some games allow for an entire avatar? Game modelling is not just about 'x' graphics card can render 'y' polygons therefore complexity can be whatever and therefore content creators dont share blame of fps loss. Any suggestion of such shows an underlying ignorance of game modelling and design especially when it comes to multiplayer games such as Second Life. Based on what you have said, why don't LL just remove normal maps entirely? Their primary purpose is to allow for low poly models to render as higher poly models, which according to you, creator optimisation doesn't matter and is probably why content creators hardly use normal maps as they dont understand their purpose. Additionally, optimisation isn't always about a computer being able to render 'x' amount of polygons. Optimisation of models both polygon and texture wise is also about reducing the load those models have on the PC systems so that it has computer power left to render other things like shading, terrain, animation, movement, scripting etc or even multi-tasking whilst playing a game like SL. There is a reason why people in Second Life hardly turn on shadows. Turn on shadows in a fully optimised (model wise) sim and you will loose hardly any fps. Turn it on in a full sim with largely unoptimised content and your fps will plummet due to the PC needing to process everything else as well. It will also not come to a hard stop as you suggest but slowly reduce your fps incrementally due to loss of computing power for those other process not just rendering of models. That is unless Second Life is different to other games and I highly doubt it is. There is also the whole dynamic of the less polygon models have the more you can render on screen and in SL's case the more you can decorate your sim environment with. So suggesting content creators are not to blame for lag or slowing down of the viewer based on their unoptimised content is absurd and not helpful to those of us wanting a smoother SL. ETA: Also I agree with you that Linden Lab need to add systems to allow for better optimisation (such as like the did with normal maps), however what is the point of them doing that if Content Creators dont use those systems or they feel based on reading things like your post that they dont need to optimise the content as they are not to blame when they are. Its all well and good to shift the blame squarely on Linden Lab (they are to blame for a lot) however Content Creators as far as pure content created lag goes are to blame here far more than LL.

Trouble with ready player me is that it will never work with Second Life without way to much coding.

oops double post. refer below.

I wouldn't put it past Facebook as they have idiotically made the same kind of dumb decisions in the past based on legislation. Shockwaves worldwide as Facebook turns off the news for Australia (techxplore.com) Whilst not as big a market as EU, Facebook did actually act on their threat when Australia decided to change the rules regarding news on Facebook, with Facebook turning off news on their platform for Australia. This backfired and they eventually decided to play ball, but at the loss of subscribers, businesses and people jumping ship to other platforms. The news that they lost subscribers over 2021 doesn't surprise me at all as at least in Australia, many people stopped using Facebook when they simply acted on a threat to try and not pay their fair share.

I know they have been on AWS for ages but their physical location hasn't been changed nor the server stuff. I was simply stating what Ebbe had said about having the assets/server infrastructure located on AWS servers in each country rather than all on American AWS servers. Perhaps it has been done. I was under the assumption from Oz before he left that the following hasn't been done and was on hold: "Today we are located in the US, which means that people from Australia or Asia or Europe have to travel quite a ways, which is hundreds of extra milliseconds of latency. So if you want to have a very dedicated community in Australia or somewhere, we could maybe start to distribute our server infrastructure to be closer to where the actual customers of those regions are, which would make things more performant." Ebbe Altberg, VWBPE, 15th, March 2018

As far as I know Oz never said that JUST moving to AWS would reduce cost. It was always stated that moving to AWS would reduce costs ONLY if they use a method similar to Kitely with off-loading non populated servers. This still hasn't taken place. There were three reasons stated by Ebbe why they were moving to the cloud. Ability to expand indefinitely (scalability) without the need to purchase a new data centre. The physical data centre was a reason why Homesteads where tied to full regions as due to how cheap they were in comparison they were worried too many people would buy a homestead and they would need more server space in their physical data centre. Was to have the regions similar to Kitely where populated sims were loaded on regular always on servers and when they were not populated they would be off-loaded to inactive servers ready to spool up when a person is present on the sim again (reduce land tier and save LL money overall) To have the asset server data easily accessible across the world in different countries to reduce the download time. (improve performance) The roadmap was that they wanted to get everything on AWS first then they will start to transition the servers to off-loading and the assets onto Cloudfront (servers in different countries and unlimited bandwidth like Kitely). This hasn't happened and the last thing I heard was for the foreseeable future they were not looking into either due to cost and time. This whole 'we are not looking into it due to cost at the moment' all happened when Linden Lab was sold.

The part you quoted had nothing to do with lowering tier of mainland. I was simply saying that LL not instantly reselling abandoned land is costing them millions per year in lost tier. Talking about actual land tier. LL earn about $50million/year dollars in tier at the moment across all forms of land. Whilst yes offering tier too low they could loose money but if done right that isn't necessarily the case. The actual size of land isn't the issue. What costs LL is the resources used on the land and the avatars on the land (AKA bandwidth and server processing). This is why homesteads are cheaper as they offer less prims and less avatars than a full region. A quick and easy way for LL to earn more is simply to remove the requirement of owning a full region to buy a homestead. The homestead is already priced to allow for LL running costs as well as profit therefore offering that land not tied to a full region has no downsides for LL due to everything being on AWS now. The restriction was in place due to space requirements and maintenance in their physical data centre. Another way LL could increase their profit and potentially bring in more users is to offer a region with say 10 avatars and 3500LI (old homesteads). They could offer these for say $70 whilst probably still making a profit making tier cheaper. It isn't a case of offering more land is going to break them it is the bandwidth and server processing power that makes LL loose money. This is also why Kitely's pricing takes into consideration, avatar limits, prim limits as well as how much ram a server has despite being able to offer 4 regions for $20.

Fun fact. Roughly 20% of mainland (not including belli as this was a stat from 2017 that never changed year on year) is abandoned land. This equates to just over 1600 full regions not earning LL money. To put it another way that is over $3,360,000/year of potential income to LL lost due to them not processing abandoned land for sale quick enough if at all. This loss is probably even more now that Belli has plenty of houses still for sale. They are still paying those hosting fees though with Amazon despite loosing all that money... Not to mention all that sea being lost revenue. Maybe you are correct and LL are compensating for such things. Goes to show how bad a business model mainland is.

No probs. I think it is important that actual figures are shown where a real case scenario exists to compare with SL as many dont know the cost of AWS. If I search for them again I could possibly even find the quotes (if they still exist) from the CEO of Kitely talking about the cost savings and how he recommended Second Life to run on the same system years ago to reduce land fees. Many people just argue "but LL said x,y,z" and dont realise that LL are a business that want to earn money for their other ventures. It would be different if SL was public and we could see where there expenditures go and what costs they have, but seeing as they are a private company and stopped that long ago we can only compare apples with apples and with what evidence we have. In this case we have an identical platform (as far as region system's go) using AWS with lower land fees successfully for 12 years and another platform using the same or near similar region hosting for 2 years charging vastly more. To me when I see that, it is a case of one of those platforms (kitely or SL) is lying. Considering both are still available I lean to the larger company pretending that cost saving cant be there. Now sure there are other factors to consider like SL having a larger population and more regions needing to be regular servers than not. But there is no way that all 23k+ regions have a person on them all the time. There is also mainland to consider with that needing to be on regular servers but LL are also paying for vast tracks of mainland to stay abandoned instead of putting them up for sale through their land auction system. It should be if someone abandons land it is automatically added to the land auction so that LL are always getting the hosting fee or cull mainland in size so that it feels less 'empty'. That said my figures were hypothetical as I calculated it based on each region being on its own server, which it isn't. In reality as well as with Kitely, there are usually 3 of 4 regions per server in Second Life. So the actually dollar amount in all cases is less but the cost savings with such a system are still there just with lower dollar amounts. The hosting prices though come from the article I linked so they were correct in 2011.

Of course it does, because LL are running the servers ALL THE TIME as a regular dedicated server. Why can you not understand this. You have a real world example of how it can be done but, I take it your response will be "that's opensim it is dramatically different to Second Life". The fact is, it is not. Just that LL aren't doing it right or as they originally stated they were going to. AWS is EXPENSIVE for dedicated servers. It always has been and I am not arguing that it isn't. That is why any person using AWS tries to limit the amount of always on dedicated servers. LL however are not doing that and so it is costing them far more. Kitely has been using AWS since 2009ish (keep in mind their profit margin is also dramatically less than Linden Lab's). That is 12 YEARS prior to SL introducing it. Here is an old article from 2011 about the costs associated with AWS and how Kitely manages to make their servers so cheap. Now yes, the article is old but even if costs have increased by a lot (or reduced), the costs and cost savings are still valid as Kitely still offer far cheaper servers. They have increased in price since 2011 but $15 for a region is still far cheaper than LL charging $229 per month for every server. From the old 2011 article I linked earlier here are a few figures for you and others on Kitely's server management: Hosted on Amazon EC2 No set-up fee (LL still charge $350 despite it literally being automated now). High traffic dedicated server of approx. 20 avatars on the region at all times (i.e. running non stop - with more avatars costing even more) would cost Kitely $2,976/MONTH Regions that aren't accessed at all over a month will cost just $0.10/MONTH to keep (off-loaded) up ready to spool up to a regular server when accessed low pop or no pop servers are off-loaded and then if they become active with even just 1 user they are loaded back to a regular server then off loaded if they have no activity again for a period of time. there is a slight delay as they are reloaded back up when a person teleports to them the reason Kitely now offer multi-region servers like 4 regions for $20 is to solve the issue they had where an adjoining region would not be loaded. Now lets hypothetically use those old figures for SL, to work out how LL could pass on cost savings to the user like they said they were planning: LL have 27343 servers at the moment according to gridsurvey. That means that always on is costing them using the above figure a total of ($2976/month) $81,372,768/MONTH. This is the current hypothetical price LL are paying using 2011 figures above as they still have all regions as regular dedicated servers. Now lets say hypothetically 1/2 of the regions at any given time have no activity on them and LL off-load them until they are needed. That means 13672 servers would cost LL a total of($0.10/month as above) $1,367.20/MONTH That leaves LL paying $40,687,872/MONTH for the remaining always on regions if they use a hybrid AWS hosting like Kitely does. Due to having to pay HALF what they would pay for always on Linden Lab can then pass on those savings by reducing tier just like Kitely does. This is how Kitely reduce the tier costs so, please explain again why LL are saying they cant lower tier again? :EDIT: This is also the system Ebbe originally announced LL were going to use but still haven't. As to Solar's laugh emoji. You must be an insanely fast reader Solar with abnormal comprehension. As you made that laugh emoji just 45 seconds after I posted my message. So basically, you had to click the thread and read both my response and the source I posted in under 45 seconds. I say BS.

I know, they are far superior in some region things compared to LL but what I meant was that a dedicated server is far more expensive than their spin up-servers. $15 (comparable to LL 1 region server) compared to $150. Also keep in mind that their servers are slightly different, for example their cheapest option which has the same land, prims, etc. as LL's regions only allow for 10 avatars and their dedicated allows for 80 avatars over the entire 64 regions not just 1 region.

No, the creation engine is a modified version of the Morrowind NetImmerse engine which was then rebranded as the Gamebryo engine when the creators of NetImmerse merged with another company around 2004. Bethesda continued to modify the NetImmerse (now Gamebryo engine) for Oblivion as per their license. When Skyrim was started in 2008, due to the amount of modifications to the NetImmerse engine they had done, most of the code was Bethesda's and therefore broke their license with Gamebryo allowing them to rename their forked engine to the Creation Engine. Its core architecture is the same though and dates back to 2002 the same period as second life's code. Due to it being the same core engine it also has the same issues as the original NetImmerse engine from 2002 such as being cell based loading with limits on how many NPC's can be in a cell at a time as well as other things. This is due to NPC's being treated by the engine the same as a player. Does this sound familiar to another engine? It should... Second Life suffers the same problem in that it to is based on cell (region) based loading and the way that each avatars meta-data is processed affects framerate as well as other key systems hence why it can only have a certain amount of avatars on the region or suffer bad framerate. For example, an avatar as well as its inventory in SL is reloaded every time an avatar crosses a region boundary, hence why you 'lag' across a sim border and sometimes appear as a cloud. This cannot be removed as it is hard coded into the cell based loading that most MMO games ran on in 2002/3. (SL is a little more complex in that whilst some regions are on the same server others are on completely other servers, but the system is very similar to how cell based loading works). Likewise both engines (Creation/NetImmerse and Second Life Engine) have certain things tied to the framerate such as physics calculations. Likewise the terrain is a heightmap-based terrain generator meaning only one axis of terrain modification (up, down, left, right) so caves and cliffs are very difficult to make well and look good without assets. These are the limitations of the creation engine as well as SL's engine that cannot be removed without a complete rewrite as they are hardcoded. This is why many of us wanted a SL2, an exact copy of SL but on a newer engine. We got Sansar instead, a complete opposite of SL. That all said it is irrelevant to your point. Both the creation engine and second life run the same way even down to their core architecture as explained above. The difference being LL refusal to update (modify) the systems they can to be better like Bethesda did due to LL not wanting to break content.

As someone who has done large multi cell mods for skyrim, oblivion and morrowind which all use the creation engine, I can tell you that you are wrong. Bethesda's creation engine is no different to Second Life and if you have used their creation kit to do a mod you would realise this. Take Skyrim for instance. The main world is similar to mainland having cells instead of sims joined to each other whereby in the .ini file you can set cell range. Everything else is a separate cell not linked (like a region). The more cell range (draw distance) you have the more lag you have just like SL. Likewise, the more assets, larger textures and complex models that are placed in a non linked or linked cell the more lag there is and potential to crash the game. This is no different to Second Life. Using the default assets in Skyrim you can make a fast loading mod as the default BSA files compact many thousands of textures and meshes into one file therefore it makes the engine able to search and render those files instantly as they are loaded at game start-up. Any mod in Skyrim that adds additional textures/models that are not part of the BSA files increase the load time of cells while those assets are loaded. The only difference between the creation engine and SL other than streaming assets and the obvious things like script sizes, environment, etc, is that the creation engine allows the use of occlusion portals and planes which break up cells into smaller parts. If this was allowed in SL load times would improve INSANELY. The LOD system is also far superior where trees in the distance are rendered as billboards (2D planes) and those closer being full 3D. SL's LOD is still based on collapsing 3D objects rather than billboards. Take for instance a shopping mall that has multiple individual shop rooms. In the creation engine you can place a portal around each shop with a door plane at the entrance of each portal zone. Doing this means that everything in that portal isn't rendered until you enter the shop or you look through the door plane. Second Life's occlusion system is pathetic to the point it doesn't work and it runs on a proximity system. The closer you are to the object the quicker it is supposed to load even if that object (or avatar) is behind a 64x64m megaprim blocking your view. The viewer though is still downloading and rendering, just loading things closer quicker. The problem with this is that eventually EVERYTHING in the region is loaded, meaning your computer is rendering many, many thousands of individual complex meshes all at the same time, with many animating and running scripts, which impacts FPS. If Linden Lab were to, as part of the sim editing tools, make it possible for sim owners to use occlusion portals and planes (like creation engine and other engines like UE5) you would find your fps increase hugely. To put it simply, the room you're in would be the only room that has anything loaded and when you move to another room the objects are rendered in the new room your in and the objects in the room you left are derendered until you go back in that room or can see them. Same would apply to avatars. All other assets are gradually download in the background and added to your cache just not rendered until they are visible. It is these systems that make other games less laggy. As Coffee stated, it is the simplicity of LL's coding and allowances that are impacting detrimentally on Second Life. No, costs are higher because LL are running an always on system and refuse to change it. When Ebbe and his team announced the possible cost savings that could happen it was stated that it would only be possible if they have a hybrid system of some regions always on and others not used as much to spin up when visited similar to kitely, opensim, etc. If this system was introduced the cost saving would be there as Ebbe, et al .stated and intended. AWS runs on a $ per use system. If the servers are not on until they are used then LL dont pay any money. This is how Kitely manages to run their sims. They have cheaper options for spin up servers and a more expensive option (near same price as LL) for dedicated always on servers.

Well another positive of the system is that you can continue to use your current inventory system and not bother with the outfit window with newer categories, just like I assume you do now with the outfit window. At least it gets a modern graphical interface in SL quickly and easily for those that want it and allows newer users to see it as a modern viewer. But, you'll be happy because LL have already said no to the idea so it wont happen anyway

It would actually be very, very easy to create a graphical yet completely user controlled appearance editor taking into account SL's unique systems. Just need to do it within a system that already exists in SL. Trouble is from my discussions with Lindens even this is too hard for them, despite the system already existing and needing very little modifying. For example: Take the outfit window add categories to it (such as hair, eyes, shirt, pants, etc) in addition to the outfit category and when clicked change the window to its selected category with items listed in picture form (like current outfit category) make inventory system folders (cant delete) named same as outfit categories (such as hair, eyes, shirt, pants, etc) link outfit categories with new outfit system folders in the inventory make uploads to the outfit window FREE (this is key) i.e if outfit take picture button = true do not charge upload fee uses same method employed for each category as outfit category now So inventory name stays the same but each hair gets treated like the outfit folder in the inventory now, so when it is saved it gets placed in the user selected named folder with the user pic saved to that hair item button on the outfit window. Positives of this are: makes graphical interface appearance editor everyone wants and needs without rewriting permissions or item coding. allows the user to decide which hair (or whatever) they want in the category (downside of IMVU is that hair you no longer want to use makes the list longer where this wouldn't as you still manage it) user controls the pictures not the creator - removing need to try and attach an image to the item. This is also important in SL as things are modifiable. In SL allowing the user to take the picture means they can take a pic of a hair that fits x body or a white created hair the user modified themselves to blue etc. Allows user to keep items where they want to in inventory and just uses a linked item in the outfit, hair, shirt etc folders. Allows the user to name the item to what they want (whilst not renaming it physically) and find it easier within the appearance window as opposed to opening 20,000 folders/items named randomly in the convoluted inventory. All of the above done within a system already in place in the viewer, fully customisable and manageable by the user (keeping in line with SL's modifiable and creator spirit). Yet... all to hard apparently.

True, however as I have said, when people type in "Second Life Reviews" when they are thinking of playing second life and want to find out about it, they dont have to look for it as it is there in plain sight in negative user reviews. The difference is that when you flag things to LL even with undeniable evidence they do nothing about it. Take MP for instance I have flagged numerous items in the past, such as the ones Scylla mentioned earlier and they are still up. They were not false reports, they are plain to see what they are for, yet nothing happens. The more worrying thing is that Linden Lab know full well there are problems and monitor them (with such issues increasing every year) but are actively choosing to ignore them by saying things like you have "its against our ToS and if we catch them we will act against them". I may as well copy this from my still hidden post as I dont know when it will be unhidden and if this was the case then can someone explain this quote from the court case (Second Life Is Plagued by Security Flaws, Ex-Employee Says | WIRED) mentioned in my still hidden post in response to @Innula Zenovka: "According to the lawsuit, in 2018 the manager of Linden Lab’s fraud team “presented information to Linden board members in quarterly fraud reports that acknowledged a high number of such Age pl*y [sic] violations were actually occurring on a regular basis each quarter.” The suit says Pearlman “was concerned that Linden Lab was apparently allowing the users to violate age pl*y rules, by not implementing appropriate procedures to prevent violations from repeating at the same levels each quarter.” The lawsuit claims that Scott Butler, Linden Lab’s former chief compliance officer, wrote a memo to other executives in June 2018 “urging compliance with cybersecurity laws consistent with Pearlman’s repeated concerns” … A former high-level Linden Lab employee confirmed the contents of the memo. The former employee said the memo “indicated that there should be more scrutiny on the ‘skill gaming program,’” and recommended Linden Lab adopt a suggestion from Pearlman to determine why it “had not been able to prevent the seedy population of ‘age-pl**ers’ from returning to Second Life, time and again.” The former high-level Linden Lab employee I bolded, I have no idea who it is but from the article it seems like he was contacted by the journalist outside of the court case and not part of the case (I may be wrong though). The article and court case also suggests that LL were not complying with anti-money-laundering rules, which after she was fired LL introduced such complying measures 1 year later. Suggesting in this case she was correct, so makes one wonder what else she was correct about. Other things to highlight is this: "Pearlman urged Linden Lab to review its age verification and consent review process, as she was worried the company could be erroneously collecting data on minors and enabling children to use the platform without the consent of a parent or guardian". Now whilst they dismissed this notion of enabling children to use the platform with their usual ToS spiel and we dont need to comply with x,y,z, it is disturbing to see that kids under 13 and under 16 have reviewed SL Here claiming to have played and seen content they shouldn't have, and in some cases encountering borderline stalking. Most are old but there are some from 2017-18 when Pearlman was employed and some even newer from 2019-2021 These children reviews of kids well underage would (if to be believe and I see no reason why they shouldn't) appear to confirm what Pearlman said to be factual with her concerns regarding age verification despite LL winning the case upon grounds that she was toxic and inept. :EDIT: This also shows that AR's are not effective (if at all acted on by LL and despite what some users have said otherwise in this thread) as LL literally are conducting quarterly reports showing no change in the number of violations. Meaning that more than likely they get banned, then create an alt and start again. I also dont think it is a matter of it going underground, it is more a matter of lack of procedures to prevent it like Pearlman suggests in the court case.