Drayke Newall

I agree, so long as they look outside the box and see the potential possibilities in these new releases and how their userbase may use them. They have to be aware or at least somewhat aware of how people (not just those that go to their meetings) are going to use these new systems so as they can stay ahead of it and ensure optimisation and a reduction in lag is at the fore. I still am dumbfounded by them not realising that people will make mesh avatars over the default avatars if they didn't update the default ones, especially given that an update to the default avatar to a new quality mesh was requested for years before mesh was released. They even had time after the release of mesh and perhaps when the first users started to look at mesh bodies to say to the userbase "oh hey, don't go to far in creating mesh bodies as we are planning to update the old bodies". Sure technically it would have been a lie from them making people think they already thought of this (which they should have), but in doing so at least they would have stagnated mesh body releases enough to update their default bodies to something better and in the meantime ensured that at least the new user that comes in doesn't get stuck with a default body that would turn any perspective modern user away. Sure there would have been other bodies still released to use over the default one, however due to labs oversight, not only did they instigate and make obsolete the default body (I know some people still use it) and contribute further to the a lag by onion layer avatars they also now have had to push through baked on mesh when the default avatar already had all this in place. (BTW I am not saying mesh body diversity we have now is a bad thing or that baked on mesh shouldn't eventually of been released) When and if labs release new features, it is these kind of QA things that labs should be ensuring never happen again as their development schedules to fix other issues within sl get pushed back as they realise these shortcomings and try to fix the optimisation and lag issues they in themselves create by not fully exploring how people in world are going to use said 'improvements'.

I believe this is how Niran did it in black dragon viewer when I requested the feature to be looked at, by using the obsolete calling cards as the link with the favourites.

I'm not missing your point. I fully understand it but you are not taking into account that people are lazy. I know plenty of people who make things in an external program like blender or max and upload it into sl without reducing poly count. I've even bought mesh objects where unseen faces are not removed, not because they were missed but clearly because the person making them doesn't know. So clearly your perception of "in order to make anything at all in blender..." is somewhat flawed if we are even having these discussions. I also know how hard uv mapping is but once again its these things that make the lag less. Yes, and my rl contracts for my business show clients the project turn around time. Does that mean that if I clearly show that something will take 4 weeks im not going to get a phone call from the client 1 week later asking where there project is? No. People only see what they want to see or if they don't understand something skip over it and see that not optimising objects still upload fine into sl so continue doing it. And yet here we are arguing as to why people aren't uploading custom LOD files or optimising content. There is always an excuse. The wiki is to complicated, I don't understand the program enough, what's a LOD... Once again in perfect world it would work like you want where people read a wiki or spend a full week watching youtube tutorials on 3d modeling and optimisation but its not a perfect world so we get left with a mess like mesh is at the moment. And yet the lindens response to many things is sorry we cant do that as the coding for the program is so complicated due to multiple updates over old systems that a rewrite would be needed. And yet unoptimized crap mesh is within sl and does sell and has brought performance down so clearly you do need tool tips or other methods to reinforce all this. Additionally the tool tips idea is more about first time builders in sl. Not serious creators. Just have a tool tip option turn off in the options if serious creators know what they are doing and don't want to see it. There is a reason why Autodesk put in place hover tool tips in there programs that expand showing sometimes small video gifs on what things do. Because they realised no one read the manual on there site. Whilst I understand your acknowledgement and point expecting me post on every 3rd party viewer site asking for a feature is a little much. Perhaps if labs had a 3rd party viewer discussion forum here these kind of things could be picked up easier as its the one place most developers look at all the time. But never the two shall meet as far as labs sees. I also realise that 3rd party viewer discussion takes place in the second life viewer section but having a dedicated feature request forum section and not a jira would be better as then it is picked up and commented on by all users not just those few that look at the jira and provide labs some form of idea on the popularity of the feature. I believe niran has looked into it and put it in his viewer as I requested it. Not sure if it is still in though. As to cache is there anyway to have it so that you can set a specific sim you regularly visit to not have its cache files overwritten so as you don't have to keep reloading them because they were overwritten?

Of course if your a content maker you could make varying LOD levels. That's not my argument. My argument is how many creators know you can make the LOD levels or even at what point each LOD mesh needs to be i.e. what is the difference between low and lowest. You are assuming amateur creators know all this (or want to know/research this) when most only know how to do basic meshing and only want to know basic meshing to earn L$ on marketplace. It's the same as why people stick 1024 textures on small objects when an 8x8 texture would do. They don't know the difference. They just see 'oh I can build in second life', upload a 2048 texture, gets converted automatically to 1024 and use that, as the defaults labs have placed in the software allow for the systems to be abused. A lot don't even uv map their stuff. That's why I still believe a higher texture size should cost more to upload than a lower one as this would help show people that higher textures need to be used sparingly. By taking out in world creation and even not updating it they brought all this to the fore. Heck, you can even create a mesh box that has a lower render cost than the default box prim. If I remember correctly each face of a box prim has over 18 triangles and gives a display weight of about 400 as opposed to 2 triangles from a mesh box with a display weight of half that. Additionally when you provide creators with a secondary avenue that is flawed such as the LOD input method that allows them to abuse the numbers to get the lowest possible LI at the lowest L$ upload cost, why would they go to the trouble of spending extra time to create their own LOD models when labs have given them an easier way. If they updated their build/prim tools to create mesh inworld as I mentioned before all those people that want to do basic mesh creation would have all the tools they need and then labs can govern the LOD automatically without relying on creators to know how. Then labs could also allow 3rd party creation tools to be used but make it so that if you want to upload from an external program you have to upload proper LOD levels. Its a win-win situation. Labs gets more users staying in sl longer 'cause people can competitively build again and also ensures that only people that know how to properly do LOD create objects externally. That's why it's important, as halebore Aeon said, to look at the main problem and fix that first instead of using a band-aid. Its also why I mentioned the avatar render cost as an example. They tried to adjust the arc levels couldn't get it right 'cause they didn't think of anything else other than adjust arc values and so stuck a massive band-aid on it that resulted in something that only degrades the user experience and sees people leave because the experience has degraded. This means if I want a lag free environment I can have that at the cost of ruining my visuals. Did they even look at issues that users have raised for years regarding texture size on avatars, glow, full bright, particles etc and limiting these first? I assume no. Did they try looking at tying the arc values to each attachment point so each point can only have a set amount of an arc value? I assume no. They gave up and gave the jelly doll solution. Wiped their hands of it and said "there fixed". Sure somethings cant be fixed, after all the program is 15 years old and no one denies this, but there are still varying solutions that can help and allow more users to log in and stay in. It has always been an issue trying to teach people how to build things as well as making the first user experience and ui easier, but labs still fail to find the simple solution in things. Why are there no in viewer tutorials? Its been 15 years... Open up the build menu and a window slides from the side of the screen (like how their old viewer used to do with the people window etc.) and it has a written tutorial with inworld playable video on how to build in sl. Change the tab on the build menu and the slide window changes to context relating to that build tab. All modern games have in game openable help tutorials because it is easier and doesn't break the experience like opening a wiki in a web browser does. Why on the build tools don't they have a dynamically changing "build tip" on each of the build menu's? For example go to the upload texture panel and you see "Build Tip: did you know, the smaller your object the less texture size you need." Add varying texture size upload cost and it could read even better "Build Tip: did you know, the smaller your object the less texture size you need and it costs less too." A simple solution to help a larger problem, but after 15 years they still scratch their heads and wonder why no one is building properly or having their arc levels for the avatar high. Why does the favourites bar on the top only allow for landmarks instead of also people from your friends list? Having the option where I can add the people I hang out with the most up on that bar and being able to right click and teleport, im, pay etc. them directly from there rather than opening the people menu (clogging my screen with windows), finding their name in a large list and tping them from that menu. I could go on but the post is already big enough so sorry for the long post.

Don't want to come across as negative, but whilst i'm sure they will look into it, the very fact they aren't looking to alter the existing flawed core methods and have just come up with oh lets just increase the calculations doesn't inspire hope. They have a flawed method in place already as far as not fixing lod levels on upload and instead of fixing that first and trialing that to see if it helps to resolve the issue they instantly use a sledge hammer and increase the LI cost. Case in point, take avatar rendering cost. How many times have they attempted to 'fix' the issue by recalculations to the arc only to result in people still abusing it. So their final solution lets just make it so people turn into jelly dolls and let the user determine the value. So what happens some new person logs into sl for the first time it auto sets the value initially and if low you get their first impression as ugly jelly dolls or they don't know what the setting is leave it high and lag and leave. Labs need to first look at the core systems and get them right before further hampering sim LI costs it should be a last resort. As to CoffeeDujur's point regarding custom lods.. in a perfect world yes, but one again you are relying in amateur creators to know better which they don't and trying to teach them different is near impossible. It took almost a year of arguing with people just to convince them that mesh is far better than sculpts. Whilst yes I do agree custom LOD's would be better, by not having automatically calculated LOD values for those that don't do custom lod's is part of the issue that needs to be fixed. If labs are too lazy to do this then remove the option to enter lod values entirely and just make it so that you have to upload custom lod's otherwise you cant upload. You have to force the content creators hand. If that doesn't work increase the L$ value if you don't upload custom lod's. Just don't give them a secondary option that provides them with a flawed lod calculation that can be abused.

So basically instead of working on other more beneficial methods to reduce lag and optimisation issues labs response is this? Bad move I believe. The whole issue at the moment why people use hack methods to reduce the LI at the cost of optimisation of an object is to get more use out of the land LI restriction. If labs boost land capacity and increase the rezzed LI numbers instead of giving extra LI usage to the sims to use and leaving the calcs for LI where they are it will just result in different hacks to reduce LI rates again. This would mean back to square one and an never ending cycle of increase and reduction. I haven't uploaded mesh in a while so may have changed but, why when loading a mesh into SL is there not a default cap on the upper and lower end of the LOD calculations. Why can people still set all levels of the LOD as the same value or even down to 0? They need to look at this first. If you have an uncapped and non default LOD input method of coarse it will be abused. Would it really be that hard to have whatever the higher LOD is the lower LOD's are calculated automatically at 1/3 (or whatever) of the higher figure and making them fixed? Next thing to look at is increase the L$ upload figure of mesh to compensate for the optimisation issues. Make it more expensive to upload unoptimized stuff would also help. Whilst probably not possible without a lot of recoding make textures resize to their designated surface area. 1024 for large, 512 mid and going down to 8 for tiny. Why do I have to endure lag based on some creator making a tiny diamond texture on a ring stone 1024? If this isn't possible vary the upload cost of textures rather than 10L$ for all levels, Increase higher texture L$ upload figures and reduce lower. Any form of optimisation that involves decreasing potential gains in sim LI as a whole is counter productive and unnecessary. Changes like the above will ensure that nothing gets broken or becomes unusable whilst also ensuring future creations start becoming more optimised which will eventually, like what happened to sculpts get phased out of use due to better stuff. Additionally it would also open up further potential L$ revenue for labs (if the texture upload increase is put in place) to offset land prices whilst benefiting the community with better thought from creators into creation optimisation.

Bit of a late reply regarding prim object creation improvement, but plenty. For starters allow prims to be joined together to make one (not linkset) object like a chair similar to how mesh works (i.e. base, back and arms become one single mesh and not 4 individual objects linked together), then create another identical object that is designated as a physics shape in the build menu that you can join to the object and hey presto basic competitive mesh development in sl without 3rd party tools. Sure still basic solid shapes sl has currently, but at least it allows sl to go back to some basic form of "your world your imagination" and not "your world your imagination - as long as you create it somewhere else". From there they could improve shape manipulation, allow subtraction of 2 objects against each other, object face removal, texture face designation, improve the texture mapping tools etc. All tools that allow optimisation within sl, decrease lag and allow the optimisation of objects to be somewhat controlled and governed. There are plenty of changes they can make to improve it and all possible to do it in small gradual steps if needed. Same thing goes for animation, having the tools within sl to actually create basic animations would be far more beneficial to not only the user but also retention of users. Having people being able to build in the world and interact with it is what kept a lot of people in sl. Plenty of people leave because they say there is nothing to do in this game. We've all heard it and it's technically true. Sure you can shop, meet people etc., but it was the ability to create and interact with the physical world that made sl the unique revolutionary software of its time. (Sculpts removed a lot of the inworld creation and mesh just killed it outright). Plain and simply, the inworld creation tools should have been updated when mesh was released but for some reason labs maybe didn't think of it or took it as to hard? Other than prim and inworld creation updates there are so many things that seem to be put on the backburner and forgotten that if updated in a reasonable timeframe would help improve user retention, experience etc. I mean what is it 2 maybe 3 years when they did that survey on how to improve the marketplace? What was the result of the feedback diddly squat. No new category designations (i.e. gatchas get their own out of the way zone), no new advanced search features, nothing. Then 3 years later we get a 15th birthday blog post on them improving marketplace this year. Happy it's been finally looked at but seriously it should have been improved years ago not 3 years later.

I think a lot of the inventory management becomes a nightmare due to just the ui of it all. While yes a windows explorer system would be nice (if screen space is an issue have the tree part of the window collapsible, leaving the folder area), even simple things can make management easier if the ui is changed somewhat. For example: A long object name in your inventory means expanding the inv window just to read (no copy), (no modify), (no transfer). Why can't these be expressed as simple icons with green C M T icons or red C M T icons for no copy etc. It would certainly help for quick glancing through things. Not sure if this is possible but the ability to right click an object and change its icon from the default cube to something more relevant. For example a furniture object - rather than a box you can right click the object and select from a list of icons and pick for example a chair icon. This will eliminate a lot of the mess, at least I have, in the objects folder and enable a quick glance approach to the inv. rather than having to read everything. Point 2 could also be applied to folders where if a new folder is created you can right click and select a variety of folder icons (or as Penny mentioned colors) not just the old yellow folder. have an icon WIP or V1,V2 etc. icons for version numbers. The inv. window as it is now has no quick glance function to find things quicker which icons could help with. Also this may not be possible but have search by icon making search even easier.

Yes, I figured that would be the case as mentioned at the end of my post however, whilst it may take more programming changes and resources now; at least it would secure a more viable, profitable and simpler username system and sign in process for the future that would far outweigh the overall initial set-up time/cost. If we are stuck with the current version of Second Life without the prospect or viability of another iteration, then it is these changes that need to be made to ensure that the already dated program continues to have a future. Not what seems, are just half measures or a Band-Aid. Half measures have always plagued LL's ideas and implementations. These result in further complicated, costly and/or irreversible decisions like the user/last name issue we face at the moment. Another (slightly off topic) prime example being the mesh body issues we face now (some may disagree). By not having the foresight of updating the default body to a better mesh one (despite users telling them to at the time) as part of the new mesh system, it has not only resulted in a more complicated mesh body system that not only hinders and perhaps deters new users but also, has resulted in the necessity of trying to implement further changes to other older program systems. If done correct the first time it would have resulted in a far better system garnering existing users, increasing user retention, simplifying and expanding user creation and all the while not needing additional resources to re-invent the wheel in an already complicated system. These decisions resulted in baked on mesh which now suffers the same half measures taken with what seems to have the issue of not being able to utilise the materials system - a backward decision. It is this kind of backward decision making that really just shows the lack of foresight and dare I say it, incompetence of Linden Lab to not be able, at the first instance, implement such simple systems. Instead we are now left with a complicated system that, with the reintroduction of last names is bound to conflict with the now widely used display name resulting in further confusion for users. It also creates unnecessary options as well as in all good business sense an unprofitable solution. I'm sure there are some that want a last name and will pay the double fee of premium and name change, but why anyone would pay for a last name when a person can simply change their display name for free and turn off the option to show legacy/user names, I do not understand.

I must be the only one that seems to think the whole last name idea to be a stupid one that just complicates everything for the new user experience and a step backwards, especially given that display names are a thing. Why cant the whole process be like any other gaming (yes we all know SL isn't a game) account creation process. For example: You have an account/login name which must be unique (User453* or whatever) and is chosen at account creation. The account name is hidden in all aspects of SL, so wont show in profile, search, MP, forums etc (makes the account more secure as people can't see any aspect of your login details at all anywhere). You also have a Display name which works as it does now. You also choose your display name at account creation for your avatar with a clause under stating that this Display Name is what your avatar will be shown as in all aspects of SL, however, can be changed at a minimal fee of L$20 (or whatever) at any time or make it so that you have to be in the first tier of premium to change your display name. Any legacy name either last name or resident, upon login when the system is implemented, is given a dialog requiring the user to enter a new display name (for free) explaining the above as a reason. Their last name or resident name is automatically defaulted as their account/login name and then hidden in all areas of SL. The above allows LL a general income of X (similar to how partnerships work now), makes the naming system simpler, allows for people to see only one name to an avatar (instead of having to tick boxes in options to display both user and display names) and makes the account a little more secure. There are only a few downsides to this method that I can see. Those serial name change people that seemingly like to spam people in chat every week with X is now known as Y. If this is an issue make second tier of premium to give the user 1 or 2 free name changes per month. Another will be the older script issues unless the script calls for the avatar UUID (which I would assume to be the better method anyway) but in all honesty half the problem with SL now is the fear of breaking old legacy scripts/Objects (most of which are sold by people who haven't been in SL for years with no update to current/better standards) which has impacted on SL moving forward. Additionally (and I think better) using the above method could also allow for a handle instead of a username and provide the ability for multiple avatars per account (with the cap of how many avatars per account being tied to which premium tier you choose). This once again simplifying the new user experience as firstly they are used to it through other programs and games and secondly means not having to remember multiple passwords and usernames for one game/program. Considering the whole last name thing is in the process of being implemented anyway, it is probably to much hassle to rework it.Also given how the resident naming was implemented years ago perhaps the current code just doesn't allow for drastic changes..

I would say depends on the hud, but how much detail do you really need on a hud apart from small icons etc? an emoji is less than 8x8 texture in some instances and still renders fine on a screen on a website. Hud should be no difference. Generally the smaller the space the smaller you need texture wise. Most games only use 1024 textures for large objects where you will notice pixilation. Well made UV Mapped textures being the unique in that a 1024x1024 UV mapped texture with normal and specular will be enough to completely texture an entire body.. head, body, feet, hands, nails, eyes, beard, skin and some times clothing clothing. Why should SL be so different? Why cant hud makers create their hud in mesh and apply a single uv mapped texture on it?

Or here's a crazy thought.... Why don't labs just keep their prim system, optimise the prim mesh (yes, I have a mesh box created externally that is identical to the prim box yet still is more optimised than the prim box) and simply add commands like Boolean etc and when done the relevant unused faces are removed. That way people can still create in world and create, rather than a linkset object, a fully meshed object that is a genuine 1 object. Sure it wont be as good as a external program but at least is brings back somewhat optimised in world building. I bet you the response would be no... to hard...no one wants that. The fact of the matter is the program simply is outdated and to complicated to make it more user friendly for those that want to build in world. Not to mention sl is still catering to extremely low end hardware, something no other gaming company would do and this negatively effects second life moving forward. Since the topic was about in part long load times of things etc. Lets look at textures even. Why cant labs make it so that texture size is relative to the texture face size. Have a small finger ring... sl downsamples the texture to a max 8x8 texture on it. Have a large 64x64 background scene sl allows max 1024 texture on it.

I think the problem with Second Life is that in many ways Linden Lab seem very stuck in their ways. A lot of things aren't fixed and consequently these things negatively affect the user retention. I also think the whole SL is dying or SL sims are reducing are also a by product of the way Linden Lab have set up things. For instance, since the creation of Private Regions for a person to be able to buy a homestead you are required to have a full region. Now while I assume there was a valid reason for this, surely that has lapsed as they have upgraded their servers etc. and I think a lot of the regions closing would stop if this policy was removed to allow anyone to purchase the cheaper Homestead Region, especially now with the increased prim allowance. It would certainly benefit the smaller RP communities and even help the 'hangout' communities that seem to die almost every week. It would have certainly stopped the closing of a Hangout/rp sim I and another person created and had for a number of years. If the idea above has its issues i.e. influx of server usage and stress (which I assume was the original reason for the stupid policy) then perhaps (at the risk of being hounded by the people that seem to have issue with premium members) to increase the premium account usage, allow those that own a premium account the ability to buy a homestead region without the need of purchasing a full region or something similar. Whilst I am a premium member I do believe the current benefits are lacking and only find the attempts by labs to increase the premium usage laughable... I mean seriously gifts, sandbox's (unpopulated) etc. Oh and I do wish they would make the ui a little bit more user friendly. Like the ability to add people to the favourite bar or even side bars so we can right click tp/im people without opening the people tab etc.