Second Life shutters orientation islands

Linden Lab has chosen to close its existing task-based Second Life Orientation Islands, which have been problematic in so many ways. Instead a modified version of the so-called Help Islands will be used as default user-entry-points instead. These modified islands were a part of an A/B testing program last month, and Blue Linden, Linden Lab's volunteer coordinator, said that the modified Help Islands 'are proving to do a bit better' according to the internal metrics that Linden Lab keeps. Just how much better, we're not sure, as Linden Lab generally plays those cards fairly close, but one Linden staffer is reported as saying that the hub-structured, task-oriented Orientation Islands are currently only yielding about 1%.

We're not even sure whether to believe that grim statistic, as it seems to be shockingly low. Linden Lab's testing seems to indicate that the task-oriented nature of the Orientation islands is a mark against them, while many others have suggested that it is the hubbed layout that is at fault. Certainly everyone agrees that the outdated and faulty Orientation HUD system (an attachable user-interface item that is intended to guide and instruct) is a major strike against the existing orientation experience.

	Are you a part of the most widely-known collaborative virtual environment or keeping a close eye on it? Massively's Second Life coverage keeps you in the loop.

Indeed, Linden Lab have been recently working on discarding that Orientation HUD with an embedded web-based variant, driven by the embedded Gecko/Mozilla technology in the Second Life viewer. The creation of artwork and assets for that system have already come well along, in a variety of languages -- though we've not seen any sign of the supporting code in the Second Life viewer itself yet. We believe that (unless this project has been abandoned) that internal alpha test versions of the system already exist.

Unfortunately, the web-based retrofit of the orientation HUD is just addressing a symptom -- indeed, that's all that is really happening here; the addressing of symptoms.

The Orientation HUD frequently fails to function because the underlying platform fails to provide stable-enough support for it to do so. Thus, the underlying bugs and glitches are ones that affect much more than just that system. Migrating the HUD to a new, web-based form certainly may deal with that, and with the issue of it being out-of-date (though, we rather think that it won't, judging from Linden Lab's performance in keeping other material current) -- and neatly sidesteps the issue of an unstable code-base underneath. One which, we are repeatedly told is a priority to fix.

It's also true that Linden Lab is somewhat drowning in top priorities.

Aside from simple stability, Linden Lab has now spent a little more than five years trying different methods of orientation, which in practice actually amounts only to a small handful of changes as it typically takes months or years for the Lab to bring each one from inception to production.

The user-interface also is widely considered a highly culpable item in early retention, or lack thereof. Though very few new users ever seemed to have much trouble getting to grips with it in the early stages. Users rarely start complaining about the user interface until they've been in Second Life for some time.

All in all, the metrics which Linden Lab tracks to measure success of orientation have shown very little real change despite the various attempts to improve the new-user experience, according to Blue Linden, 'All of these experiences perform roughly the same. I've heard people suggest that that there are easy wins, but we really haven't seen them yet. So remember, it's a very complex problem, and we might never get above, say 20% retention ... that might just be the nature of the beast.'

And he's right. Mainstream usability is a very tough nut to crack. Of course it's also possible that the metrics used to measure 'success' are off-beam or incorrectly weighted. If they are, then mainstream usability will only ever really be achieved by serendipity.