The other day Tateru was blogging about the life expectancy of SLV 1.23 and series 1 viewers. (Reference) She was quoting OZ Linden. Today the minutes for the Open Development User Group posted. Oz Linden is talking about where the Lab is with older viewers and has the quotes. So, how long do the older viewers have?
‘Old’ is always a subjective or relative term. Apparently some residents are using some really old viewers… you know, the ones that come on the stone disk. I guess the Lab just doesn’t bother to block older third party viewers (TPV).
The problem with older viewers is in the way they talk to the Second Life servers. Over time how the servers talk changes and viewers have to keep up. After all when was the last time you saw a rotary dial phone?
The Lab does a transition to the new ways of communicating. For some time they support the older styles. Currently one of the recent changes is in how we download textures. UDP was the protocol by which textures were downloaded. That has changed to the HTTP protocol. As the Internet has changed the HTTP protocol works better and is more reliable and faster. For now one can use either protocol. Older viewers, like SLV 1.23, do not have the code included to use the HTTP protocol. So, when the Lab stops supporting UDP texture download that viewer version will fail.
So, when will older viewers stop working and when will SLV 1.23 become unusable? We don’t know and Oz tells us the Lab has not decided when to stop supporting 1.23. OZ said the SLV 1.23 is showing higher failure rates. The grid is changing underneath it and SLV 1.23 is having more problems and crashing more often.
Over the next few months older viewers will start to have more problems. Clothes are already a problem for some viewers. Using old and new viewers on the same machine certainly creates more clothing problems. How the system remembers what your avatar is wearing changed months ago. So, changing clothes in a new viewer and then opening an old viewer will show your avatar wearing the old/previous outfit.
What is Going to Break?
Mesh is not going to work. Old viewers render mesh as some flat triangles glommed together. When mesh clothes are worn I’m not sure what we are going to see, but it will be odd. So, these problems are coming.
Avatars wearing physics layers fail to render correctly now in viewers not supporting Enhanced Avatar Physics.
New profile features are coming and they won’t be in older viewers. The older in-viewer profiles fail to update and work well with the new web profiles now. The API’s for the older profiles will stop being supported at some point and all the old style profiles in all the viewers will stop working.
The new chat is likely to cause problems for older viewers when it arrives. You probably know the recent plan to upgrade to XMPP chat failed and is chat being revised in some new direction. The team is still searching for the bottle neck that is causing chat lag. Coming server upgrades and viewer upgrades will have more reporting features to provide data on what is causing chat lag.
Most of the application interfaces (API) used by the server are getting upgrades. Viewers that do not get updated to the new API’s are going to have problems and likely fail. Supporting the older API’s adds complication to the system. The Lab does not plan to support the old API’s for long.
Are Series 1 Users Being Abandoned?
The Lindens are looking at what keeps people from upgrading to SLV2. Figuring out what motivates people to stay with series 1 viewer is a significant effort in the Lab. They hope to address the issues and motivate people to update. I wish them luck with that. But, the series 1 users are not being abandoned. At some point the Lab will have to leave them behind or hold everyone else back.
I suspect many of us think the big problem for series 1 users and their not adopting the SLV2 is the user interface. The Snowstorm team started changing the interface to make it more usable. We are hearing hints that a larger team inside the Lab is working on doing even more to change the user interface. However, much of that process has moved to the Viewer Evolution Group.
I don’t expect any big user interfaces changes. I do expect how the viewer parts work to get better and easier to use. We may see new ways to open packages and wear items as Phillip discussed. The Direct Delivery ideas feed into this line of thought.
The handwriting is on the wall. Viewers based on series 1 code are on the way out. They simply will not support the new features in Second Life. Some are still trying to back port features to series 1 based viewers but they seem to be falling farther behind. Plus how the code 1 viewers do things and talk to the servers has to change too.
I take from Oz’s statements that the Lab doesn’t want to block the older viewers. But at some point they can’t continue to support the API’s that drag the system down. When they stop supporting the old API’s the older series 1 viewers will fail.
We don’t know how much time is left. We may find out when the new mesh update announcement comes out at the end of the month. At least we will have new information with which to speculate. For now I am guessing from a couple of months to 3 or 4.