There is an article by Hamlet, Future Phoenix Viewer Development in Peril?, that points to a post by one of the Phoenix-Firestorm developers, Tonya Souther. See: Viewer 1 is officially on borrowed time. The title says it pretty clearly. The writing is on the wall, Phoenix is a dying viewer with an expected near term death. While Hamlet focused on Phoenix, there are problems ahead for any V1 viewer. Well… actually any number of viewers brands.
Near the end of June an announcement was made about the projects Linden Lab™ will be working on for the next few months. See: Project Shining to Improve Avatar and Object Streaming Speeds and my article Second Life Changes Coming.
HTTP stands for Hyper Text Transfer Protocol. If you’re curious about the more technical but simple definition read The Webopedia definition of HTTP. For an even more technical explanation see: What’s HTTP? Explain HTTP Request and HTTP Response.
The important aspect of HTTP for Second Life™ users is that it uses TCP. Lots of acronyms, I know. But, spelling them out is tedious. TCP=Transmission Control Protocol. The key word here is: Control.
Many of the current Second Life services and even more of past services use or have used UDP (User Datagram Protocol). UDP’s redeeming quality is simplicity and lack of control, a just do it mentality. The Wikipedia says, “Time-sensitive applications often use UDP because dropping packets is preferable to waiting for delayed packets.” At one time it made sense for the Lab to use UDP. The advances in hardware and technology have improved and connections are way faster, which reduces the need for the Lab to use UDP.
HTTP is more complex. It pushes much of the complexity into the hardware. Routers and network cards take over the duties of error correction, to some extent data encryption/decryption, and data compression/decompression. An application has way less work to do and requires less coding. So, to the application HTTP appears to be much simpler.
The Lab’s HTTP
The Lab is always pressed for time. Their initial HTTP code was not the total master piece of all inclusive code they or any of us would have liked now. But, for the time, it did its job. Now that is changing. The Shining project will build a new HTTP ‘foundation’ for Second Life. I suspect the research into the Avatar Bake Problems (See: #SL Clouds, Grey, and Blurry Avatars) revealed these problems and resulted in, or at least contributed to, the decision to redo the HTTP services of Second Life.
The Lab is rewriting their HTTP library and changing a number of services… well ALL the UDP services and the HTTP services to use the new HTTP library. The Lab will be standing up new servers to provide existing services now provided over UDP to provide them via HTTP. At some point the Lab will be taking down the UDP servers.
A byproduct of the change is the stress placed on older routers will be less. Some of the LinkSys WRT and Belkin routers may, keyword may, start working again. Locations where more than one SL user is on will work better. Wireless will likely work better.
When those UDP servers go down, old versions of viewers that do not use the new HTTP will have more and more services/features that do not work. We know the Lab will not be upgrading SL Viewer 1.23.x. So, it is doomed as are all old versions that are not upgraded.
The Phoenix-Firestorm team was to meet 7/15 to decide what they would do about the Phoenix Viewer. As yet, I have not seen an announcement. The PH-FS Team seems more terrified of the users than do the Lindens. So, I am sure they are waiting to get the preliminary code and information on the new HTTP Library before deciding. If integrating the new code is going to be difficult, then the Phoenix Viewer is likely doomed. If not…