Fluffy Sharkfin

It may have sounded that way but ALM is permanently enabled, is in reference to the fact that you can enable/disable reflections, shadows, etc. when using ALM which does give similar performance/visual results to using a non-ALM viewer but it will still be using the new PBR/ALM renderer.

1,000,000% this! If your mesh has the same polygon density in areas that don't bend or flex as it does on areas like knees, elbows, etc. then you're doing it wrong! When creating rigged mesh it's not the number of polygons that really matters, it's the location of the polygons/edge loops that make the most difference. In fact in a lot of cases having less polygons in the right places will make it easier to get the desired results than having a very dense mesh since having the right topology on joints like knees and elbows makes it a lot easier to ensure that your mesh retains the correct shape when it bends. I'd recommend taking a look at some of the examples on this page Topology - polycount (especially the limb and shoulder topology sections) and noting how the edge loops are arranged in some of the examples.

Sometimes you can get better results by creating a completely different model for your lower LODs rather than trying to reduce your high LOD by a sufficient number of triangles. In this case I'd consider using something like this... The model itself is 18 triangles and as you can see the corresponding UV is neatly squished into a small corner of the texture area. Since you only ever see this model at a distance you don't really need a wood texture for the frame (just sample the average colour for the high LOD texture and use that instead) and the square area in the middle should be large enough to add a lower resolution image of the butterfly with a matching background colour (and perhaps a faint shadow to add a little fake depth). Then all you need to do is include the texture for your new low LOD model in an unused portion of your high LOD texture. ETA: For the lowest LOD you could even replace the above model with a single plane (2 triangles) and place the UVs for it over the same small low LOD texture since at that distance it really wouldn't matter if the model is flat or not.

Understood. The point I was making in reply is that, for those with a disability that prevents them from using either, the potential loss of nuance would most likely be preferable to being excluded from the conversation entirely.

Dyslexia is an umbrella term and there are several subcategories each of which can have a different impact on an individuals ability to comprehend written communication. I'm a huge fan of the written word and am often overly verbose when communicating via text but at no point in time has anyone in this thread suggested that text based communication should be removed in favour of voice. Everyone should be free to choose which they prefer and there should be more emphasis on providing additional support and resources for those unfortunate enough to not have a choice at all.

Aside from the obvious reading disabilities such as dyslexia there are those who have difficulty reading due to a visual impairment of some kind, then there are amputees and those who suffer from chronic arthritis, muscular dystrophy and other conditions that can make typing difficult. I'm sure there are a myriad of other conditions which, while not severe enough to completely rule out using either method of communication, contribute to each individuals personal choice as to which they prefer.

Regardless of how reliable AI assisted systems may be or how much nuance may be lost during translation, I suspect those with a disability which prevents them from sharing in joyful little moments like the one you described would welcome a chance to be included. I keep emphasising words like choosing and choice but I shall make it as plain as I possibly can in case anyone isn't clear on my intention. For most, this entire topic boils down to personal choice and I strongly believe that, rather than yet another discussion about which form of communication is best, it would be far more beneficial if those who do have a choice were discussing how best to help those who don't!

Simply put, conversations on voice tend to flow a little faster/more naturally and it's easier to multi-task while text communication is more useful when attempting to provide detailed instructions or relay precise information. To be clear, I'm not advocating for the use of either. I have some friends who prefer voice and some who refuse to use it at all and I have no qualms about choosing to conform to their preferences. I do however take exception to anyone trying to impose their personal preferences on other people. As has been pointed out this topic has already been discussed in various threads on these forums many times over, unfortunately they usually devolve into the same sweeping generalisations and pigeonholing with little consideration for the individuality of SLs residents which is a shame since a conversation addressing the lack of accessibility in SL and how things could be improved for those who are excluded from using either method of communication due to a disability would be a lot more productive and far less contentious. Still, with the recent (and seemingly exponential) advances in the development of artificial intelligence perhaps it won't be long before there are reliable speech-to-text/text-to-speech apps which will remove the barriers that are in place for those who are currently excluded through no choice of their own.

This! ^^^ If someone chooses to exclude themselves from certain conversations by refusing to use voice that's entirely up to them but that doesn't entitle them to impose their preferences on everybody else just because they don't like feeling left out.

A valid point but the same could be said of people who refuse to use voice, no doubt there are some very talented creators out there that suffer from dyslexia or similar issues affecting their ability to read or write. I honestly don't understand the aversion to any form of communication be it voice or text, each has its benefits and of course everyone has their preference (and for an unfortunate few there simply is no choice due to a disability of some kind or other) but based on the reactions of some people you meet to the mere suggestion of using voice instead of text you'd think they'd just been asked to sacrifice their first-born or perform lewd acts with barnyard animals. That being said I do understand people resenting being asked to "voice verify" their gender since that's just creepy, intrusive and really weird but then since I don't bother with most of the "adult-themed activities" my response to that sort of request is usually just to laugh and inform them that no matter what my RL gender may be they aren't my type and I'm just not interested.

I believe what @animats is suggesting is that rather than sending a request for the entire JPEG2000 file as soon as the object in question is within draw distance it may be preferable to calculate how much of the file is required in order to display the texture at an appropriate texel density for the objects current size on-screen based on the objects distance from the camera, it's relative velocity (in order to ascertain if the object is approaching the camera and is therefore likely to be viewed at a larger size or if its traveling away from the camera and won't require additional texture information) and the direction in which the camera is facing (since objects within the cameras field of view should obviously have higher priority than objects that aren't currently being rendered). As long as the assumptions are correct in regards to the potential performance improvements that could be gained from this approach as a result of the difference in the way in which SL assets are hosted compared to the previous approach it seems like a really good idea (but then I'm not a network engineer nor am I particularly familiar with SLs network architecture so I could be missing some critically important factor). It certainly sounds plausible though and, if it works, could potentially even go some way to solving some of SLs texture loading issues.

If you use scattered planes then you could lower the number of required polygons by grouping leaves together on the texture for each plane, preferably by creating all the leaf groups on a single 512x512 texture (or 1024x1024 if you need the additional resolution) then arranging the UV mapping for the planes so that each plane shows different portions of the texture in order to make the leaf fall look more random and natural. ETA(Part I): Essentially when you're baking textures (whether it be branches on a tree or leaves on the ground) it's always best to try and reuse textures as much as possible rather than trying to bake a unique texture for each part, especially when dealing with hundreds of smaller objects like leaves, since it's unlikely that anyone is going to notice whether each leaf is unique no matter how closely they look. ETA:(Part II & III): I know a lot of people obsess over LI and it's always good to try and optimize both your mesh and textures as much as possible, but I'd suggest not letting land impact play too much of a part in your creative process or decision making. It's better to create something which looks good at any LOD it's likely to be viewed in (I emphasize this because for objects that are in enclosed spaces discarding the lower LOD models in favour of a single triangle is perfectly acceptable) and have it come out at a slightly higher LI than achieve the lowest LI possible and have it crumple into a mangled polygonal mess when viewed from a certain distance. As for texturing, LL are currently working on a new viewer which will enable materials for all residents so although most may not be using them at the moment soon everybody will be able to see them, so consider using materials to create your leaf textures, etc. because they can turn a flat texture like this... (I posted these pics in another thread recently but since they're particularly relevant to the topic I'll post them again here) into seemingly 3D objects like this...

In order to reduce the resolution/number of unique textures that your tree requires try isolating one or two generic looking branches and only baking the textures for those. Then place copies of those new low-poly branches around the tree so you re-use the same branch textures multiple times rather than requiring a unique texture for every branch/leaf on the tree.

I definitely agree that creating content for Second Life is a more complex and involved process than it was before the introduction of mesh (and prior to that sculpted prims) but that's a slightly different issue to the one you alluded to in the post I quoted. I know that may seem pedantic but in this particular instance the distinction really does matter because the first and probably most important step in brainstorming a solution to any problem is to clearly and accurately define exactly what problem it is you're trying to solve. So, in this instance, the problem isn't so much that creating content in SL has become infinitely more complicated but rather that the current in-world creation tools aren't capable of producing content of the same caliber as content created using dedicated 3D applications. Now obviously the ideal solution to that problem is to update the in-world building tools and introduce new more versatile and powerful features that will enable users to create better content comparable to that created using dedicated 3D applications. Of course not all of that solution is realistic/attainable (and the part that isn't has been struck through just to make it clear) because no matter how much time and resources you invest in developing new tools and features for the viewer you're never going to be able to compete with dedicated 3D applications when it comes to the power and versatility they offer creators. Which leads us to another equally important part of brainstorming solutions to problems, keeping expectations realistic and not spending time trying to fix issues that are insoluble. That's just the price we pay for the level of versatility and creative freedom that SL provides us with but I think it's a fair trade. Well you can add my 2 cents to that because I completely agree with you that in-world building needs an overhaul, the question is what can we buy for 4 cents? I don't make content to sell in SL so as far as I'm concerned if LL could wave a magic wand and introduce features that allowed every resident to create stunning content in-world I'd be absolutely overjoyed but I'm certainly curbing my expectations as to what can potentially be achieved. As for specific ideas on what those improvements and new features may be there have already been some great suggestions including a recent and quite detailed post by @ChinRey on ways in which the existing primitives system could be expanded and improved upon. I'd personally like to see something much more ambitious in the form of a new type of voxel-based in-viewer building system, something akin to Firestorms new Local Mesh feature but with the addition of a rudimentary but extensible voxel-based building tool which would run locally rather than in-world and integrate with the mesh uploader in order to convert the temporary voxel objects to mesh and upload them. Another alternative would be for LL to overhaul the current SL linkset to include proper hierarchies that allow for parent-child relations and sub-trees which can be animated simply by adjusting the transformations of the parent (but since I'm not that familiar with the way in which linksets are currently handled I have no idea how possible that would be). As for creating clothing in-world I've seen a few suggestions but every experience I have with cloth simulations in 3D apps suggests that attempting to incorporate something like that into the SL viewer is a really bad idea unless you're actively trying to drive away every resident that doesn't own a modern graphics workstation/high-end gaming pc. Anyway, regardless of what improvements LL do (or don't) make to in-world creation tools, I think we will have to accept that using dedicated 2D/3D content creation tools will always yield superior results to what can be achieved within the SL viewer (at least in the hands of someone that knows how to use those tools).

My own personal preference is 3D Coat which offers a very comprehensive set of voxel based modelling tools alongside traditional low-poly modeling and high-poly sculpting tools. Modelling with voxels is a very different experience to working with geometry and allows you to focus more on the creative aspect and less on the technical since essentially voxel primitives in 3D Coat are "constructive solids" as you put it and can be merged, cut and intersected to create anything without ever dealing with a vertex or polygon, which makes 3D Coat a lot easier to start creating with for those completely unfamiliar with traditional 3D apps.

It's been a while since I built anything substantial out of prims, in what way have they become more complicated to use? As for Blender, it isn't the only option when it comes to creating content for SL and some of the alternatives are a lot easier to get to grips with.

It's a shame more creators don't make proper use of materials since they can make a staggering amount of difference when creating foliage using flat polygons with textures. For example, here's a set of very basic hand-painted leaves... add a normal map and tweak the specular/roughness a little and this is the result... By adding curvature and a little gloss you can turn a flat plane into a seemingly high-poly surface and completely transform even low poly/LI content into something you'd consider using as a centerpiece.

According to the SL stats page on Daniel Voyagers blog SLs highest concurrency was close to 90k

Don't forget Pressure Washer Simulator! LL could start up a whole new in-world industry... premium members could hire non-premium members to come to Bellisseria and mow their lawns, wash their cars, etc. which would boost the economy and also help with immersion and realism with the addition of a fully functioning class system in SL!

I honestly have no clue, I think there's a "total users online" counter on the viewer log-in screen but it's not something I've ever really paid any attention to.

And before that the infamous Massive Grid Stress Test of 2005 during which they tried to squeeze a whopping 4000 users onto the grid at the same time! 😮 (I can't remember for sure if they managed it or not, but have a vague recollection of the total number of users getting pretty close before the grid finally collapsed and LL declaring it a success anyway)

@Sandra8675 Since you're trying to record the entire conversation rather than just the output from your own mic you're probably going to need to capture the audio from your desktop/application. If you want to capture sounds from the SL viewer along with the conversation then you should set up a group call in SL, if you only want to capture the conversation then you'll need to set up a group call in Discord or Teamspeak or whatever other voice capable software you prefer (you may also want to disable all sound-based notifications in whatever voice app you're using otherwise they'll also show up on the audio track). Once you have your voice call set up click the Add Source button in OBS and select Application Audio Capture (BETA) then, depending on what sounds you want captured, select either the Second Life viewer or the app you're using for voice comms as a source. ETA: You still need to include Audio Input Capture as an additional source in OBS to capture your own voice because the audio picked up by your own mic isn't going to be part of the audio output of the application you're using (otherwise you'd be hearing your own voice through the app every time you talk).

Assuming the objective is to create something which communicates like a human it wouldn't be very realistic if it didn't misrepresent opinion as fact and occasionally make things up entirely.

I'm actually okay with that! On the slim off-chance that they inadvertently create a RL Skynet I have a preference for my terminators being dumb (and also preferably slow and with poor visual acuity)! Also if they could incorporate "questionable fashion choices" into the mix that would be a great callback...

Yes, SL has some very wonderful content and some regions are exceptionally well done! Unfortunately most new users never get to see it because apparently they aren't sticking around long enough to hunt around and find it. They simply look at the Classic avatar and the really old stuff littered around them and think "Ewww this is old and ugly!" and log out never to return. If LL want to capitalize on all the work that they've been doing to bring PBR to the grid they should be planning a complete overhaul of all the new user areas and similar regions in order to really show off exactly what Second Life is capable of in 2023!