Pantera Północy has his 30+ minute video of the 4/7 meeting posted. I’m just getting around to watching it.
Viewer Pipeline
Maintenance Viewer – No new news.
Voice Viewer – Vivox is not providing changes for Linux so Lab is not changing their viewer, which is mostly unsupported by the Lab in any event.
On a better note, the Lab has given up the idea of disabling viewers with older voice packages.
HTTP Viewer – This is the one that begins requesting assets via the Content Delivery Network (CDN). It is in QA, so expect it real soon now. Word is that in all testing it is working perfectly.
Alex Ivy 64-bit Viewer – The question ‘When will 64-bit build instructions be available?’ came up. Answer: as soon as 64-bit is ready. Until then the Lab is working with individual people via the mailing list.
The Lab is testing 64-bit on 7, 8, & 10. Windows 7 has significantly higher crash rate than 8 or 10. 64-bit viewers running on Windows 10 have a very low crash rate, but the sample size is small. So, this could be misleading. Once the viewer goes to RC we’ll have a better take.
The Lindens are also improving the viewer’s OS detection. Currently they are getting lots of unknown OS’s reported by viewers that can’t recognize the OS.
360 Viewer – No change. On hold until 64-bit is working well.
Region & Parcel Access – sets of changes
1 – Avatar Capacity Limits are going up for Premium users. We’ll see that roll out in week #15. The Lindens have not seen any problems so far. I suspect they are considering technical problems. If they are anticipating social drama problems, they aren’t talking about them. I tend to think the Lindens are in the experience circles that would allow them to anticipate such problems.
As other changes are made to the SL system the Lindens will watch for opportunities to increase region avatar capacity.
2 – New OS version of Linux should be rolled out to the servers in week #15. This isn’t something we will notice. There may be performance improvements. But, if the Lindens are expecting that, they are not talking about it.
There is lots of back end work on-going.
New EO/EM Parcel Access Changes – lots of covenants say no ban lines but no enforcement. Adding enforcement. In server. A viewer with the additional controls #15 See the article: New Estate Control, for more information.
In the video, there is a lot of discussion on how this will work. But, little clarity. The Beta Server meeting provided more information. Remember. These are the viewer people we are hearing in the TPDev Meeting. Grumpity Linden, the mover shaker for the change, was there, but she wasn’t talking much.
NVIDIA Problem?
The latest NVIDIA driver seems to be causing some crashes. Version: 381.65 released 4/6. I have had a couple of crashes since I updated. But, I haven’t tracked them to NVIDIA. Not having consistent crashes, I have to say “may be”.
ACI @ Time Mark 20:00 – There was discussion of the Lindens changing the ACI calculations. These changes will very likely change the weight of Belleza and Maitreya mesh bodies. That isn’t the goal. It is just these bodies have high polygon counts and therefore may suffer.
The Lab is accumulating data on the render costs of various assets. They are striving for a change based on measured data they can formulate into a modification for the equation calculating ACI. The idea is to improve a feature that is improving SL.
I saw Penny Patton posting about her unhappiness with the current ACI calcs. I think most of her concerns will be addressed in the coming revision of ACI. It is sounding like polygon counts, texture size, and materials weights will be more costly.
So… ACI get worse then it already is ? i need to move ACK already to 350000 to avoid Jellydolls. still sometimes see it. but there’s no setting above 350000 , except max. and that you not want to use.
I think the lab can better admit that secondlife cannot handle quality mesh. Lucky i see very rare a jellydoll. But the keep a ugly invention.
You have missed the point of ACI. It isn’t that the SL system can’t handle quality mesh objects, the problem is people are not building quality mesh. ACI is the incentive to build quality mesh just as Land Impact Cost is incentive to make efficient use of prims.
While your computer may render poor quality, high polygon count, large texture mesh fast enough for you, there are those with less powerful computers that have a problem with it. There is nothing to require you to be inclusive and caring about others. You can look at the problem from just your self centered perspective and project blame to the undeserving that are striving to be inclusive… I’m being a bit hard on you in the hope you’ll wake up.