Drama swirls around the coming end of the Second Life Viewer 1.23 (SLV1) and SL Snowglobe 1.4 & 1.5 viewers and possibly some Third Party Viewers (TPV). If you are interested in getting perspective and an idea of what is coming and why, read on.
It appears to have many upset and predicting the end of SL over the viewer change from the SLV1 to the SLV2 series viewers. I personally think these are people not paying attention to what is happening at Linden Lab.
Consider. A viewer must work with the SL grid to be used. LL is changing server side code and adding features older viewers don’t support and they are doing it at a faster pace now. Take Collada meshes for instance. Old style viewers can’t render them; they just won’t rez for a person using an old viewer. Lindens like Oz are telling us even more features are coming. These are features and changes that have yet to come into the new transparent process. So, there is even more coming older viewers won’t render or otherwise be able to use.
TPV Dev’s building SLV2 series viewers will have to integrate the new mesh code and other changes with the old code to get meshes and new features working. That is way more work than plugging the feature into viewer 2 code. Some TPV Dev’s are porting SLV2 code back to SLV1 style viewers.
TPV Developers have long complained that LL code was a mess of spaghetti code and needed to be made modular. Now Linden Lab is doing that. But rather than start with the SLV1 code they started with the new Viewer 2 code, SLV2. It would be like putting an electric motor in your gasoline powered car, lots more changes would be needed than if you were putting a new more efficient electric motor in your already electric car. Those developers using V1 code may not be able to keep up with LL’s planned fast paced changes.
Much of the viewer controversy swirls around Oz’s statement,
Don’t waste everyone’s time suggesting that we throw away Viewer 2, or that we revert the UI to Viewer 1. It is absolutely not going to happen, and any suggestion to that effect will be ignored.
Many jump to the conclusion that this means we are stuck with the Viewer 2 interface. Other use it as an example of LL continuing to refuse to listen to residents. Anyone that is using the Development Viewer (several thousand now) knows both claims are silly. The Lindens have done their Mea Culpa and are changing Viewer 2 as fast as they can. PLUS, they have made it easy for residents to have input as to what they want. See: Snowstorm Project
Oz has also said they are transferring parts of Snowglobe (viewer 1 series) parts to the Viewer 2 code. Oz also has no problem with parts of the SLV1 interface being moved to the new SLV2 series. As LL makes these code transfers they are making the code more modular, just what developers wanted. One really has to wonder about the whining we hear now that LL has decided not to put effort into older less modular code.
Oz has repeatedly stated there is no plan on when to turn off series 1 viewers. It won’t happen for 3 months, for SURE, and may be not for up to a year. The reason; “We know that the UI issues for v1 users are still much too serious” and “We don’t need to yet – the number of SL features that they [v1] don’t support is still acceptably small.”
We know that we’ve got to solve the second problem [v2 user interface] before the missing feature support in v1 viewers becomes too big a problem.
The bottom line is that we can’t afford to develop two families of viewers, so we have to stop investing in v1 completely (and yes, even just keeping it up and building requires significant work) so that we can fix v2 faster.
I hope that we can change the traffic on this [open source mailing] list to be more focused on consideration of the actual work to be done, and how we can collectively do it best, and spend a bit less time on Pronouncements Of Doom and other drama.
Henri Beauchamp, Cool VL Viewer developer who’s code makes it into many TPV’s, disagrees with Linden’s direction and stance on SLV2 and the abandonment of SLV1 code (Snowglobe is his specific meaning) and user interface. His thinking is many TPV’s are using Snowglobe code and back porting new features to it and will continue to do so for the next couple of years. Time will tell us if he is right. However in the mean time, he is making back ports and publishing them for other developers.
Again, TPV developers are not going to bother with v2 unless YOU, Linden Lab, change your stance on the UI and do accept reversals to the v1 way, where needed.
I’m not sure how Oz can be any clearer that they are taking parts of Snowglobe and are open to changes in the SLV2 interface and are, in fact, changing the SLV2 user interface as fast as they can. Also, we already have several of the popular TPV teams saying they are moving to SLV2 code base. Henri says otherwise and that no TPV Dev’s are submitting changes. I don’t develop a viewer, so I don’t have a dog in the race, and I obviously see things differently than Henri and most other naysayers. A TPV Dev stating they are changing to SLV2 code base and reaching the point where they can submit changes takes time. So, it may be some time before we see evidence of a change over even if it started some time ago.
The Lab is working toward dividing the render engine code and server interface code from the user interface code. Parts of the viewer and server will be placed into the same code fork as they once were. This will allow the TPV Dev’s to more easily create different interfaces and add viewer features. So, I can’t see TPV Dev’s staying with Snowglobe code for the couple of years Henri predicts. I also think the changes at Linden Lab are going to speed things up, something I think many are forgetting.
If you are wondering which interface will win out and become predominant… SLV1 or SLV2, my guess is neither. The winning interface is going to be much more SLV2 like than SLV1 like for the simple reason that some metrics indicate 70% or more of new sign ups drop SL because of the SLV1 viewer interface being difficult to learn. (That 70% number is difficult to verify, one just has to make a call as to who they believe). That apparently improved a tiny bit with the SLV2 interface. This reason alone means the interface WILL change as the Lab tries to improve player retention.
I hope this gives you some perspective on what is coming in the way of viewer changes. I think it is obvious that Linden Lab has to do something to improve player retention and increase time in world. Some significant part of retaining new players is the viewer and how easy it is or isn’t to learn. Along with those ‘ease of use’ changes are the new features. One of which will be; how we buy, open packages, and use/wear stuff. Phillip says that has to become easy and fun. I agree with that. All these things mean the viewer has to change.
I have no doubt that Linden Lab knows the SLV2 user interface is unacceptable to the community. Especially as they say as much. The SLV2 user interface will change. How it is going to change is largely open to the residents. Plus now the viewer development is a transparent process. So, we should see more of what residents want appearing in the viewer.
Linden Labs past performance in many ways has been to over promise and under deliver. So, I’m not surprised many are skeptical of LL’s statements that they are changing. Time is going to tell if this is a change or whether the Lab will slide back into old habits. They are still figuring out how to make these changes and getting everyone at the Lab on board. It takes time, but I am hopeful it is real change.