Viewers, Viewers, Viewers: A Way Out of This Mess

Sunday, September 5, 2010 Sunday, September 05, 2010

Readers may have noticed controversies and discussions over Second Life viewer software in recent weeks and months. In some cases scandal, in others too much change, still others are simply inoperable in many situations. A constant question is, "which viewer should I use". There have been many answers - as many as there are viewers. Clearly there's lots of opinions and few are satisfied.

Meanwhile, new users continue to have challenges learning the environment. Let's face it: SL and 3D worlds are quite a different experience and it takes time for people to get accustomed to using them. Linden Lab knows this, since we've observed them trying to address that problem by deploying the controversial Viewer 2.0.

Viewer 2.0 seems to work for some, and yet not for others. Why is this so?

I've thought this for a long time now, and I'll say it again: there are two different audiences making use of viewers. First there's the new residents, who are frequently confused by the complexity of the interface, give up and never return, and the experienced existing residents, who yearn for more function to customize or simplify their sophisticated virtual existence.

Here's the problem we've been having all along: we're trying to service two entirely different audiences with a single viewer. Hence, Linden Lab creates Viewer 2.0, which includes features designed to simplify life for new residents, while still trying to address the needs of experienced residents. In my opinion, it didn't go nearly far enough in the simplicity direction for new residents. In the opinion of many long-time residents, Viewer 2.0 confused their operations and made their complex virtual lives more difficult. Viewer 2.0 missed the mark for both audiences.

While Viewer 2.0 marginally improved the experience for new residents, it was rejected by a great many experienced residents, a vast number of whom fled to use a variety of third party viewers. Unfortunately, one of the major third party viewers ran into issues and many residents now scramble to find an alternate viewer - and many of them wouldn't consider Linden Lab's Viewer 2.0.

I've seen situations like this before many times in real life, and a good solution approach is to realize the root cause and directly address it. In this case the root cause of viewer difficulties is that there are two audiences that have very different requirements. Thus, the answer is quite simple: we need Two Viewers, one greatly simplified suitable for stark newbies devoid of advanced and confusing features, and the other a more complex viewer filled with all the features one could imagine.

Of course, the challenge will be that Linden Lab is able to muster resources only sufficient to support one viewer. Right now, their efforts try to do it all for everyone, addressing both audiences but in fact doing so in a less than optimal manner.

Here's my proposal to make life a lot easier for everyone: Linden Lab should abandon development of the advanced viewer entirely and leave it to evolve independently by third parties through open source approaches. Publish and maintain a secure specification for accessing the grid and the underlying open source code, and let others do the job of building advanced viewers. This has worked in the world of web browsers and many other environments; why not here too? To some extent, we've already seen this happening, albeit somewhat placed in shadow by Linden Lab's highly visible Viewer 2.0, often selected by residents simply because Linden Lab made it. But imagine how well these third party viewers could advance if they weren't competing against Linden Lab's viewers?

Meanwhile, Linden Lab could then focus efforts on a truly simplified viewer specifically designed for new residents, one that could actually achieve the vision of "five minutes and you're in" that really has no chance of happening in Viewer 2.0. Consider how well they might do if they didn't have to worry about including all the features required by existing residents. New users could "graduate" to one of the more advanced viewers when they feel confident enough to do so.

Let's make things simpler for everyone. Linden Lab desperately needs a simplified viewer; Residents want (and have proven they can make) an advanced viewer. For me, two viewers is the obvious solution. What do you think?


Feline said...

I think you're right, Armi. There are multiple audiences (you can split them into more than two, but two is a good simplification to start with) who need viewers with vastly different feature-sets. I would venture to say that the best new user viewer would be a browser-based solution, if there was a way to overcome the heavy client we download today.

Going the route of separate viewers does lead to a lot of interesting questions. Where do you draw the line at what features to include in the simplified viewer? At first thought, you wouldn't want to include build tools, since they can be intimidating. But on the other hand, build tools are absolutely necessary in order to adjust prim hair and skirts so that body parts don't stick through them. We have adapted to using advanced features like building for simple tasks like wearing hair or clothing, and that muddies the water for sorting out how simple a lightweight viewer can be.

sororNishi said...

Yep, that works for me.

Maria said...

A lot of the functions currently performed by a viewer can be handled by in-world objects.

So, with a Web browser, email isn't built-in -- instead, you would go to or IM isn't built in. Payments aren't built in.
(Though they all these functions used to be part of the AOL browser -- and still might be, for all I know.)

For a 3D equivalent -- your instant messages could be handled by a device you carry. You don't like its usability? Get a new device.

Your clothing and inventory can be handled by a virtual closet -- with multiple companies competing to provide the best possible closet for you.

Payment systems can be handled by different providers. So can search and directories.

We're a tool-using species. Maybe if more of the menu functionality was moved into tools, newcomers to virtual worlds would have an easier time using them -- and, by decoupling them from the viewer itself, evolution could come faster.

After all, how good would search, payments and email be if we were all still stuck using AOL's solutions? We wouldn't have Google. We wouldn't have PayPal. We wouldn't have Facebook or Twitter.

It's hard for one company to come up with innovations and user-friendly solutions in 100 different areas, which is what SL currently tries to do.

-- Maria Korolov
Editor, Hypergrid Business

Nightflower said...


Unfortunately, too brilliant for LL to listen to. But you can bask in the satisfaction of being right :)

Lalo Telling said...

I've seen the two-viewer idea advocated before, and it makes sense -- which, with my cynical hat on, tells me that's reason enough for the Lab to reject the idea.

Otherwise: what Feline said, but add other attachments to the list. I spent my first two years in SL almost without rezzing a prim to make anything -- but I also spent that time as a furry. The confidence I gained, over time, from the necessity of editing attachments enabled me to consider becoming a builder.

@Maria: The idea of add-ons/extensions/gadgets that connect to SL through some kind of API also makes a lot of sense. It's the way of the digital world, these days. But... Third-party asset servers? Ideally, that's what competitive 'virtual closet' offerings would be, not merely another widget to plug into the viewer. Not saying it's a bad idea, only that it's a much higher level of investment for potential developers, beyond writing the code to make it work.

Daniel Smith said...

In the last week, as I come up to speed on the Unity3D environment, I have started to think of the viewer 1.x and 2.x as an evolutionary dead end...

For some time, the most progress in viewers will come from 3rd parties adding on features.

LL is hampering progress by the Wizard of Oz Linden's refusal to consider the 1.x UI on the 2.x codebase. This will dramatically cut into the pool of potential developers. They will simply put forth their efforts elsewhere, so that a greater portion of the community can benefit.

I know of two efforts which use Unity3D as a front end for OpenSim (Rezzable and I expect to see many more. Three factors that will drive development of a full fledged Unity3D front end for OpenSim/SL are:

a) Sheer Quality -- go look at the Tropical Island Demo. SL is nowhere close in terms of graphical ability.
b) widespread deployment -- with Unity3D, you can deploy to Windows, Mac, Web, the Wii, etc all at once.
c) potential developer base -- A very capable version of Unity3D is Free, and it has a large community of developers.

There are, of course, some major hurdles -- coming up with a flexible UI, converting prims to meshes, and how to handle tons of legacy LSL code. Going forward, developers would be able to script in JavaScript, C#, or Boo ... all of them a hell of a lot more capable than LSL.

So, in short.. the existing 1.x and 2.x have too many wheels to reinvent if they are going to get up to the quality of environments such as Unity3D. I sense we will have a day of reckoning, where the community decides that OpenSim + some great game engine is the way forward.

ArminasX said...

Several people have indicated there's more than one audience type, and I agree. Another idea mentioned was the notion of "plug in" modules that could switch modes of the browser, erm, I mean viewer. The viewer could suddenly take on features specialized for say "builder" or "musician", "podcaster", or even "fashionista". These modes would include whatever features are necessary to make that role easier.

While that's a really terrific idea, I don't want to get it confused with the newbie experience. I still think newbies should be exposed to an experience so simple they won't give up. Simply turning off modules would not be the right approach; it must be drastically simpler and friendlier than that. Perhaps it might involve dynamically adding features as the resident becomes experienced, but starting from near zero? Much to think about here.

Open Source projects tend to be really great at adding features, but woefully reluctant to surgically remove the vast amounts of features necessary to sufficiently simplify the viewer. That's the reason I believe responsibility for simplicity should continue to lie with Linden Lab.

Adric said...

"how well these third party viewers could advance if they weren't competing against Linden Lab's viewers?"

Competition is a good thing, but I'm not sure that is the issue here.

The issue is two audiences doesn't even scratch the surface. It's like all the other issues. The fashion folks want their issues front and center, same for the artists, the role players, the whatever.

Second Life may seem like a lot of groups screaming for different things, but that is sort of the cool thing about Second Life. Diverse doesn't begin to describe the grid.

There is no one fit viewer. There never will be. If a third party viewer is so good that no one uses the official one, I suspect Linden will get the picture and self regulate. They don't want to pay those monkeys to code Viewer 2 if no one is using it.

Miso Susanowa said...

The whole 2.0 viewer thing leaves me kind of cold. Just who is the target audience flooding into SL, and is the infrastructure able to cope with this projected horde? Experience says no...

When someone first learns to use a computer, it is very complex and intimidating to them... because it is a complex tool. Many have the same problems with new cell phones, with all their options...

When you want to learn to use a new device/platform/program, there is a learning curve comparative to the complexity and ability of that program/device. That is the "cost of ownership" or "cost of extending your abilities."

Facebook is no cakewalk either for people who have only surfed web pages and used Google before; I know, I have had to train people to use it. The same goes for eBay or Amazon.

Yes, there are some elements that can be simplified in the interface, or unified, but expecting to make it "web-like" is planning for the past, not the future.

Second Life, and the other virtual worlds, are not Google, nor Facebook; they are not web pages or cell phones, and some effort will be needed to learn and get comfortable with the abilities and new knowledge required to access such a platform.

dilspi said...

As viewer 2.0 came out i simply thought it were a bad joke or so.
But, even if i will get shouted, named, hated or simply ignored for this.
Viewer 2.2 is getting something worth to be considered and that for the majority of us.

Also its clear for me, that LL wouldn’t develop more than one viewer line and the TPVs are just (but in some cases real cool) changes and extensions to a LL-viewer line. So the TPV-developers will sooner or later move over to the viewer 2.2+-line.

Yes - for sure - to be ready, also for non-newbies, viewer 2.2 need to change in some essential points and has to have a more feature-completeness as it do has right now.

If viewer 2.2+ goes through some development steps, which need to address this, the TPV-developers will jump on the train and all the back-porting activities could make room for new feature-developments.

Any viewer2-crying wouldn’t help and do not bring us further to a useable viewer. So lets start to name the annoying points in viewer 2.2+!

So, what would it need to make viewer 2.2+? a considerable one?
Some point i named for myself so far are:
give me my corners back! How could a viewer have attachment points in each corner, but the interface start to cover them with sidebars or minimized floaters?
Let me see something! Non-active floaters need to have transparency (this is nearly easy to do with a custom skinning)
the window-cosmetic shouldn’t be so room-eating on my tiny screen!
Click-efficiency! multiple profiles-, group-floaters and so on are essential.

Dil Spitz

Related Posts with Thumbnails

Subscribe Now!

Search This Blog