as always, please refer to the server deployment thread for the latest update.
- The Main (SLS) channel was updated on Tuesday, August 14th with server maintenance package 18.07.20.518086, previously deployed to the RC channels and containing internal fixes.
- On Wednesday, august 15th, all three RCs should be updated with server maintenance packager 18#18.08.10.518612. This comprises an update to the HTTP delivery of off-line IMs and notifications that should ensure group notices and Friend requests are correctly delivered.
Animesh RC Viewer
The Animesh release candidate viewer arrived on August 13th, with the release of version 126.96.36.1998579. For those not familiar with Animesh (anyone?) please see the viewer release notes and the links within them, or check my updates on Animesh, provided as a part of my Content Creation User Group meeting summaries.
Outside of the Animesh update, at the time of writing, the remaining SL viewer pipelines remain unchanged from the end of week #32:
- Current Release version 188.8.131.527973, dated July 30th, promoted August 3rd. Formerly the Quinquina Maintenance RC viewer.
- Release channel cohorts (please see my notes on manually installing RC viewer versions if you wish to install any release candidate(s) yourself):
- Project viewers:
- Linux Spur viewer, version 184.108.40.2069906, dated November 17th, 2017 and offered pending a Linux version of the Alex Ivy viewer code.
- Obsolete platform viewer, version 220.127.116.110847, May 8th, 2015 – provided for users on Windows XP and OS X versions below 10.7.
Region crossing have been somewhat rougher recently than had been the case of late. Simon Linden had been poking at the code in the hope of making some updates, but has been diverted but other working, and is hoping to get back to things soon.
I know the viewer has very similar messaging code as the simulator and there are tweaks I want to make to that which should help crossing issues caused by packet loss … it doesn’t always recover when a crossing goes bad, as we all know. We’re looking at that from two ways … ideally fix things so it doesn’t go bad. But if it fails, be more graceful about it … and there’s a certain amount of hope that the way we force the error is actually what happens in the random normal case.
Part of it now is even having the 2 regions and the viewer agreeing on what happens and how to deal with it … where an AV and vehicle end up and confirming all parties know the same state… I’m hoping this can at least get better if the target region can at least report to the viewer and 1st region a status for how it works out.
Simon Linden discussing region crossings at the Simulator User Group meeting, August 14th, 2018
The question was asked if part of a multiple region crossing (e.g. from the corner of one region to another, passing through the corner of a region adjacent to both) could result in the viewer talking to the “wrong” simulator, to which Simon replied, “I haven’t seen it have issues with talking to the wrong simulator, but then I’m not looking yet at the issue of a corner crossing that might quickly move to a 3rd region.”
So, no changes for the present, but the problem is still getting attention at the Lab, and there is an appetite for trying to improve things from the server-side of the equation, with Simon also noting, “It’s also a classic case of working on SL … we always want to fix things with a minimum disruption. Being able to make something better only by touching the server is a lot better than requiring a viewer update.”