Viewer 2.0: Another reason to love Imprudence?

… or Emerald, or Meerkat, or CoolViewer, or…

If Tateru is right (and she usually seems to be right about these things), the story goes something like this: a Long Time Ago, when the Grid was young, the Lab dreamed about the Next Big Version of Second Life, under the obvious name Second Life 2.0. As the months went by, though, and SL turned out to be fsking wonderful and wildly popular, and much of the Lab’s time was taken up in keeping the Grid from crashing, rather than aiming at a Big Flag Day when everything would suddenly change, things were introduced more incrementally (the end of telehubs, the introduction of flexies and sculpties, and so on). On the client side, we got Windlight (which is still not done, grr), the abortive attempt to introduce viewer skins (which failed because everyone hated the first one they did), and not much else.

Now apparently the Lab is still thinking in terms of the Next Big Rev of the viewer as a comparatively large and sudden change, under the name Viewer 2.0. What will Viewer 2.0 be like? Hard to say.

Last year when bopping around the official Wiki, I came across the Landmarks and Navigation Project page, which contained among other things a rather painful section called “Food For Thought”, which argued that it might be a good idea to get landmarks out of inventory, and to “deprecate” user picks, apparently based on someone from an outside design company hired to think about Viewer 2.0, observing that:

  • in the ten minutes they’d spent in SL (or, or Metaverse, or maybe it was Club Penguin, who can keep these things straight?), they hadn’t found landmarks and picks useful, and
  • web browsers don’t have them, so they can’t be good.

Also in the Food For Thought section were some thoughts on how to make search in SL more like Web search, since as we all know SL is basically a website, and Web search is perfect for websites.

I winced heavily at this, as would anyone who’s actually used SL and knows how important LMs and Picks are, and (since it’s a Wiki) I went in and made some changes to make it less painfully clueless. Shortly thereafter, someone commented on Second Thoughts that this whole “Food For Thought” thing had been essentially retracted in an SLDEV posting, where the external team says that actual Resis had told them that getting rid of LMs and Picks was an awful idea, so those things were no longer on their list. (Why they were, and are, still in the Wiki is a bit of a mystery.) So I reverted most of my Wiki changes, and just stuck in a note pointing at that SLDEV posting.

The “Landmarks and Navigation Project” page has been pretty much moribund since then, and I haven’t heard the name of that particular outside design company (“VectorForm” I think is was) mentioned again. The Lab hired “Big Spaceship” (apparently another design company, not a children’s cartoon) to improve the user experience, and they applied modern user-centered design principles by adding lots of extra black pixels and decorative design elements to the Second Life website. The decorations were removed or scaled back when Residents complained about not actually being able to see much information because of all the screen space they took up, but I believe the extra black pixels are still there on the homepage, helping save electricity or something.

Last March, the Lab put up a weblog posting about its goals for the next viewer (under the name “Viewer 2009”); they were very vague, on the order of “make it better”, only with UI buzzwords.

Then back last June, Tateru Nino got hold of an early Viewer 2.0 prototype that had somehow found its way onto the public daily-build servers. My main impression of it was that it wasn’t all that interesting, except that it added the ability to organize one’s friends list (a good idea!) and replaced the right-click pie menus with traditional dropdowns (which seemed like a bad idea, in a sigh-whatever sort of way).

Shortly after Tateru’s posting about it, that viewer vanished from the build servers, and as far as I know it hasn’t been heard from again.

Then just the other week, Tateru (again!) posted an interview with M Linden in which (being M Linden) he says various painfully chirpy things about the Viewer presentation at SLCC09 (which I didn’t see) and the new viewer in general, including talking about (sigh) getting rid of the pie menus, and also notably referring to:

the browser-like location bar where you enter a region name (e.g., p squared), hit return and voila you are magically teleported there. It is really slick and every time I use it I smile.

This worries me because:

  • He says “browser-like” as though it’s obviously a good thing,
  • He either doesn’t realize that the slick new thing he’s describing is basically already on the Map dialog, or he’s assuming that the people he cares about don’t know that, and
  • If that’s what makes him smile, I have no idea what else he might think is “slick”.

He also said “I really like the sliding side panel.” Gah. Too much emphasis on form over function imho. He didn’t even say “we’ve done studies, and determined that this sliding-panel design is better than what he have now.” He just “likes”, it. Oh boy.

Over on Second Thoughts, Prokofy Neva opined that the passage I quoted there probably means that the Lab was going to get rid of landmarks as inventory objects, in favor of just cutting-and-pasting in and out of the location bar that makes M smile. I think this is pretty unlikely, myself, because of the apparently very strong negative feedback that VectorForm got when they originally floated that thought. But you never know! I have signed Prokofy’s petition on Landmarks, and I’d urge you to do the same (if only so that M doesn’t see a petition with just a couple dozen names on it, and think “heh, I guess people don’t care about LMs anymore!”).

Apparently some other people don’t like the new viewer nearly as much as M says he does. The Herald (an SL humor magazine) published what is apparently a collection of early tester comments on the current incarnation of Viewer 2.0, and the impression that I get from reading it is that, once again, it’s been designed by some wet-behind-the-ears UI designers, who don’t have any deep knowledge of Second Life or Resi usage patterns, but who do like “slick” stuff like sliding panels and cutesy icons.

Do read it for yourself and form your own opinion on what’s going on. But it made me shake my head.

However! To some extent it doesn’t matter. Viewer 2.0 itself can’t actually get rid of landmarks in inventory, because landmarks being in inventory is a server-side thing. If the Lab is really clueless enough to make LMs hard to manipulate and trade in the mainline viewer, I’ll just keep right on using Imprudence, or Emerald, or any of the other third-party viewers that are owned by people with a bit more practical sense. They will continue to have nice useful landmarks unless the Lab actually breaks them on the server, and there would be so much uproar about that due to broken content that I really can’t imagine them doing that.

This goes back to what I said the other day about third-party viewers: not only do they speed up innovation and unleash creativity, they also allow us to mostly just ignore any silly mistakes the Lab might make with the mainline viewer. It doesn’t entirely free us from the consequences of such mistakes, because some people (and especially newborns) will always be using the mainline viewer, and if it becomes harder to interact with those folks, and harder to help them out, that would be unfortunate. And more subtly, some changes to the mainline viewer could cause cultural changes that would eventually impact all of us (if for instance landmarks were made less useful, landmark-givers would probably become rarer).

But in general I think the third-party viewers are a real help here. While I don’t think they’ll go as far as removing landmarks as inventory items, the Lab will make mistakes with Viewer 2.0. But because no one is locked into that viewer, no one has to suffer from those mistakes if they’d rather not, and since they presumably know this, the Lab may think a bit harder before making random changes based on someone thinking a sliding sidebar looks slick. Even if it’s M Linden who thinks that…

Imprudent Emerald Meerkats

Imprudent Emerald MeerkatSo the recent Second Life weblog posting on third-party viewer policies has caused quite a stir in the comments, due in part to one particularly vocal and vehement and misinformed Resident, but due also to some quite understandable concern about the current and future role of third-party viewers in Second Life.

Ever since the Lab open-sourced the viewer, way back in some year prior to this one, there have been third-party viewers available, based to a greater or lesser extent on the main Linden viewer. A bunch of alternative viewers are listed in the wiki; the ones whose logos I’ve snatch for the adorable picture here are Imprudence, Emerald and Meerkat. I’ve used both Imprudence and Emerald, myself; currently Emerald is my favorite just because it has so many nice small features, and they seem to have thought hard about the overall user experience: lots of the little things that I want to do are just that one or two mouse movements and clicks closer to hand than they are in any other viewer.

And of course there are the enhanced avatar physics. :) But I was using Emerald before that feature was added.

I’m hoping that Imprudence, say, will be goaded by a friendly sense of rivalry to make their viewer even more pleasant to use than Emerald. Competition is good. :)

Third-party viewers are really popular on the Grid these days, especially among people who have been around long enough to know they exist and try them out. One nice thing about Emerald is that it can often tell you what viewer other Residents that you see are using (although they can turn that off if they’d rather you didn’t know). The fraction of the crowd using Emerald seems to vary from around 10% in your average dance club, to 40% or higher in places where the people tend to be more oldbie or more clued in general (I’d say it was close to 50% at the Lamplighter procession the other night).

The new policy that was announced on the SL weblog was rather vague and unformed. It said that people using third-party viewers to violate the Second Life Terms of Service would be warned and/or banned, as appropriate; but since that’s true of anyone who uses any viewer to violate the ToS, it doesn’t really mean anything new. It also mentioned that the Lab might provide a registry of viewers known to be nice. And in a letter to the developers of third-party viewers, Cyn Linden apparently mentioned the ability to encrypt chat and IM as a function that is “at odds with” the ToS, something that has left quite a few people scratching their heads.

Toward the end of the comment thread linked above, Blondin asked for people to summarize their thoughts on the subject. Here’s what I wrote:

Always nice to have a chance to summarize. :)

My basic feeling on the general subject is that third-party viewers are a primary source of innovation in virtual worlds technology, and I would hate to see Second Life cut itself off from that stream. This is largely for selfish personal reasons: I think a Second Life that closes in on itself that way would quickly fade from the scene, and re-establishing on OpenGrid or elsewhere my social networks and my reputation (not to mention my collections of nice tank-tops and dark-red hair) would take time and effort that I’d rather spend on other things.

In itself this new policy doesn’t, I don’t think, put too much of a burden on the viewer innovation stream. If the lab publishes a reasonable set of rules for acceptable third-party viewers, and is reasonable about enforcing those rules, that would be fine. The rules should include a ban on functions whose main effect is to enable copyright violation or griefing. If the Lab has some good reason for it, they could also include a ban on end-to-end encryption, but I would really like to hear the reasons behind it. And the rule should not say that the encryption function may not be in the viewer at all; at most it should say that the function should be turned off when connecting to the SL grid.

So as long as this policy is truly just an attempt to limit the use and distribution of malicious viewers, I think that’s all to the good. To the extent that it’s a step toward closing the grid to independantly-written viewers in general, I think that would be a very bad thing for Second Life as a leader, or even a significant player, in its field.

It’s been loudly suggested that the third-party viewers that we have now are a threat to content creators, and I’d like to take a moment to address that issue, because I don’t think it’s true, and I think an artificial conflict between viewer developers and content creators would be harmful to both communities, and to Residents as a whole.

The current third-party viewers include, as everyone knows, various features that are useful to content creators: things like more flexible and precise build tools, and temporary texture caching that doesn’t use the asset servers. (The argument that if users get used to free temporary ‘uploads’, they will come to demand free real uploads is unconvincing to me: one could as well argue that allowing free uploads on the Beta Grid will lead users to demand it on the main grid as well, and that hasn’t happened. Users understand that if they want to store textures on the main grid asset servers, they need to pay the upload fee.)

But as well as features that are directly useful to content creators, third-party viewers also contain innovations that help protect content. Consider a skin creator, for instance. Unless the main Linden viewer has changed since last time I looked, every time someone using that viewer walks into a club wearing a skin that that person created, a copy of that skin is saved in cache on the hard drive of everyone in the club who can see them. On the other hand, if someone wearing that skin walks into a club using the Emerald viewer, with the appropriate box checked (I forget whether it’s the default), the only thing that ends up in all those caches is the fully baked and composited AV texture; the skin itself isn’t sent anywhere.

This kind of feature protects, not the Emerald user as such (who probably doesn’t really care, as a user, about what’s in other people’s disk caches), but the creator of the content that the user is wearing; the content creator gets this protection regardless of what viewer she herself uses. This feature has been discussed for the main Linden viewer for some time, but because the Lab is always busy, and has things that it considers more urgent to attend to, it hasn’t been (at least last time I looked) implemented there yet.

This is the virtue of the third-party viewer system in general: it allows innovation and experimentation with all sorts of features, to the benefit of the Residents using the viewers, the producers of the content that they use, and the grid as a whole, without the burden of the Lab’s review cycles, internal politics, and resource constraints. Malicious innovation is of course also possible, and I think the viewer policy that we’re discussing here, if carefully and thoughtfully implemented, will be a good way of minimizing the impact of that.

On the whole, it’s my feeling that Second Life will stay in the game only if it stays tied to key sources of innovation. If this policy is a way to do that responsibly, that’s a good thing. If it is part of an effort to break that tie, that would be very ill-advised.

It’s always hard to tell what’s going on in the collective mind of Linden Lab. I hope that this new policy will be made sensibly and thoughtfully, and keep Second Life going strong. If only ’cause I have a lot of tank tops. :)