This video time compresses no-voice sections to reduce the total run time from 70+ minutes to less than 45. The Time Marked outline is below. We get a few hints of possible coming features.
Prior to the meeting officially starting, Medhue Simoni and Jessicaann Wrigglesworthwere were discussing how Appearance Sliders in Blender-AvaStar work with animation.
The meeting starts at TM: 03:00.
04:30 – Vir Linden working on the HTTPAssets viewer to reduce crash rates. Fixing a couple of bugs. New version out to RC this afternoon (Thursday – v18.104.22.1685825).
05:47 – Medhue is thinking he may figure out a way to make rigged attachment-like-things (num-chuks) that follow the hand. You may remember since the ‘spare’ bones (think unused tail) are not parented to the hand they do not follow the hand.
09:20 – Using mesh-layers discussion (local text chat mostly). The current behavior of the render pipeline is to render layers with transparency as the top layer. This assumes all the layer’s vertices are in the same X,Y.Z positions.
However, while some designers are using this known behavior to support their products the Lab does not consider it a supported behavior. So, at some point it may change.
12:20 – There are overlapping subjects in this and the preceding section. At about 11:10 one of Medhue’s live stream viewers asked, “What’s the rule of thumb when you have a one prim mesh and several faces?”
16:32 – Discussion on changing ALM so that it is supported by lower end machines. A couple of people at this meeting were convinced that ‘most’ people are not using ALM because of the performance hit it causes. They are also convinced that ALM can be improved so that lower end machines can use ALM.
Along about 34:00 I turn my ALM, Ambient Occlusion, and Shadows on and off. You can see the performance hit my GTX 1060 takes.
- ALM+AO+S/M+P enabled = 26FPS
- ALM+AO+S/M+P disabled = 46FPS
- ALM+S/M+P enabled & AO disabled = 28FPS
- ALM+AO enabled & S/M+P disabled = 47FPS
It isn’t the ALM that hits performance, it is the shadows.
I will point out that for a couple of years we have asked Oz Linden for stats on the number of people using ALM. As best I can tell, the Lindens have never got around to measuring that metric. This is one of those places where users and Linden priorities differ. The Linden can’t see the value, or at best evaluate it differently. They aren’t making products to sell in the marketplace.
19:50 – A JIRA accepts a feature request… Vir explains what ‘Accepted’ means in the JIRA. Basically, when a feature request is marked ‘Accepted’ goes on the good ideas list. That does not mean it will ever get implemented. It is in competition with other ‘good’ feature requests for Linden resources and believed value to the majority of SL users.
23:45 – Still talking about how many use ALM.
28:45 – Supplemental animations are mentioned. It is a feature in the proposed upcoming features. These would be like the default animations we have now, walk, run, etc. I am not aware of a list of what animations will be added. I assume; hands, tail, wings, ears, face…
29:20 – Adding Light Maps or layer-able textures on objects.
30:15 – Animated Mesh and Flexi Mesh. Vir says it is in the category of features seriously being looked at…
One of the serous negative concerns is what would happen if someone fills the region up with animated mesh objects. They think adding these would be a serious performance load. I suppose it is a matter of deciding whether they can make them efficient enough to be considered a good benefit to lost performance ratio.
34:18 – Discussion begins on; Animated Materials, Specular and Normal maps.
41:40 – No meeting next week