I really don’t like spending my time testing sim crossings, but my curiosity continues to drive me.

Is the Sim-Crossing Problem Fixed?
I’ve been thinking about recent sim crossing improved experience. I’ve learned that when I’m with friends with different viewers, the smooth sim crossings I’ve been experiencing isn’t necessarily shared. I’ve also noticed that the seas are a lot more crowed lately with peeps in boats & planes, and many peeps are crossing sims at high speed. So, I believe the viewer plays a big part (of course, other factors include broadband speed, computing power, strong enough video card.) But I believe the Lab is on the right track and this problem is Fixed for a noticable number of peeps.

So, When Did Problem Get Fixed?
I looked back in my main blog for clues as to when the problem was fixed. I found a post on November 6, 2011 (about 5 months ago) that indicated to me that problem was in full force. And I recall the agony of being tossed out of my boat, relogging, then going to find the boat.  So it wasn’t fixed back then. I was running Phoenix 1.5 (1185) at the time. I’m sorry to say I don’t know when I converted to Phoenix 1.6, but I know I was reinstalled Phoenix 1.6.0.1600 after my failed attempt to use Firestorm 3.3 on February 26th.  Here’s where it gets grey because I’d become accustomed to slowing down for sim crossings during that time and I’m sorry to say I didn’t notice the improvement until March 27th.

Was It the Video Drivers?
I’ve been looking for explanations, who fixed the problem? When I checked, I noticed that I had installed Nvidia driver update 295.73 on February 21st. But as I’m pretty sure that didn’t fix the problem because of post I did on February 26th, “… I retested travel on the seas and although the seas are not so great lately, I was able to travel over a dozen sim crossings without crashing.” So this may rule out the improvements being related to the new drivers.

Is The Fix Viewer Specific?
In my testings I’ve proven beyond a reasonable doubt, for my configuration, the sim crossing problem clearly has been fixed in Phoenix 1.6.1. I’ve also proven it is not fixed in Firestorm 4 and I’ve seen some indications that the problem still exists in Viewer 3. Some of my tests were done with passengers, and about half of my passengers experience the sim crossings without incidents (one of my successful passengers was running Phoenix 1.5, the other running Exodus). So, I’m coming to the conclusion that it wasn’t something Phoenix fixed, but they have compatible code, as does Exodus.

Then Why Doesn’t Viewer 3 Work?
My tests were NOT rigorous or scientific, but I’ve made a case that some viewers (Phoenix & Exodus) in some computer configurations can make virtually seamless sim crossings. And I believe I’ve ruled out updated video drivers as the reason.

So, I believe the Fix occurred between February 26th – March 27th.

I’m left with the conclusion that something that was done at Linden Lab during that period resulted what I believe is a Fix. If my skimpy evidence about LL’s Viewer 3 is correct, then why didn’t the Fix work on the Lab’s principal viewer?

More Questions
There are still a few of questions. I haven’t really determined if the Fix works on mainland sims (class 4 servers), but I’m reasonably certain that estates (class 5 servers) and homesteads are good (certainly if they are mostly water), so I need to give the mainlands some testing. And is I haven’t given other viewers, specifically Viewer 3, a test. So those questions should be answered.

*But the big question is, did Linden Lab make a modification to server code that was intended to fix the sim crossing problem?

UPDATE: Cinders (@lointaine)
I can narrow that down for you: Bluesteel 12.03.20.251556
This is again the “threaded region crossing” project. scheduled 2012-03-21
Features
Prototype threaded region cross for agents. Uses the new threaded RPC/thread state machine mechanism.
Also, last month LL did a huge upgrade moving many of the Class 5 servers to Class 7 servers, and therefore better performance all around.

Yes, there was a single sim crossing blip
on one of the 30 sim crossings
I made on Easter.

I think its fair to say, Bravo Linden Lab!