Last Friday there was a Third-Party Dev UG meeting that ran 48 minutes. The video is here. I’ve included the interesting stuff mixed in.
The main channel will stay with version #19.04.22.526534. No restart. This is the version pushed to reduce disconnects. In my experience, things are much better.
Blue Steel, Le Tigre, and Magnum will update to version #19.04.25.526669 on Wednesday. The version is listed as having Internal Fixes and Fixed EEP Regressions.
Rider Linden met with an SL group, Visionaire, to explain EEP. It is an hour plus video.
From the Server-Scripting UG: [12:34] Oz Linden: And we’ve already put in place a bunch of improvements [think detectors & stats] to our RC monitors to help prevent a similar recurrence (with more coming soon)
As of the 4/18 server change was to roll the server OS back. In doing so some are seeing less script time available. One or two have seen a drop from 100% of scripts to only 30% to 60% run.
Oz says they have lots of performance data for that metric. So, they will be checking as they move to re-upgrade the OS. The simulator OS has to update as part of the pre-move to the cloud.
There are other script performance things planned. So, the Lindens will add in a look at the script slowing issue along with those. Continue reading →
There is news. Everyone has been back to work long enough post-holiday
for things to be getting completed. So, changes are in progress. I am behind,
again. So, I’m just getting this converted from my notes to something readable.
We did get a Deploy post, appearing Tuesday afternoon.
The Rainforest Photo Contest, Entry # 2
The main channel
updates to version #19.01.09.523003. This version is listed as having logging
collecting data to build a baseline performance measure. Also, more internal
My empirical take is things have degraded. But, without some
objective measure that is only an opinion. As I started flying a complex Shergood
helicopter in July 2018 my perception of region reliability has likely changed.
But, my observation of what happens when crossing region boundaries is more
Using the Firestorm Viewer, I have the option to cross
regions in two ways. The default is to cross using the basic viewer prediction
of what should happen for a smoother appearing crossing. The other is sort of
stop-action crossing. (Preferences->Move
& View->Movement->Movement at region crossing: Predict or Stop)