One year ago today Linden Lab finally launched Viewer 2. I say “finally” because the project was originally called “Viewer 2009”, and was running several months late to warrant that name; however, the more accurate definition might have been “prematurely”.
Reactions to the launch announcement were fairly mixed. Leaving aside the fact that generally, people don’t react well to change and all bar the freshest of new users were very, very grounded in Viewer 1.x – so Viewer 2 was going to have an uphill battle to start with, no matter what functionality it offered – it was fairly clear there were issues with the Viewer. Issues that went beyond people’s personal likes / dislikes in terms of UI; things simply didn’t work – and basic things at that. Worse, some decisions had been made concerning the interface that made common tasks a lot more obscure – such as the lack of any CREATE option in the initial Context menus, as well a terms – such as APPEARANCE – meaning different things depending on where you encountered them within the Viewer. There were other issues as well – basic usability issue that should have been more comprehensively considered prior to simply springing the Viewer on an unsuspecting audience, as I commented on at the time.
The biggest criticism concerning Viewer 2 – other than it apparently having been developed by people who spent very little time actively in-world – was no so much that it wasn’t ready for prime time (it was in fairness launched as a “beta”, and therefore was expected to have rough edges), but rather that for a time it appeared as if Linden Lab were not prepared to accept any form of criticism for the major faults within the Viewer, be they technical or with usability. For example, repeated complaints concerning the way the Sidebar both violently interfered with the world view and took up far too much screen real estate were ignored until well after Kristenlee Cinquetti’s S20 hybrid demonstrated just how easy it would be to make this feature both less intrusive and less jarring on the eye.
Beyond this there were security issues with the Viewer raised from day one relating to media usage which are – even now – entirely unresolved. And while it is true to say these vulnerabilities existed in Viewer 1.x – the situation with Viewer 2 is a lot worse, given the way the Viewer has been set-up to enable shared media and the like. Indeed, these issues are currently the focus of much concern among the user community because of the way they are being abused – and rightly so – and that LL apparently considered and dismissed them from the outset in their desire to push ahead with media sharing within the Viewer, has always been something of a bone of contention.
Indeed, such is the situation at the moment that these issues look set to overshadow any celebrations around the launch of Viewer 2. In some respects a shame because – to be fair – it cannot be denied that in many areas, the five iterations we’ve had with the Viewer since its launch have seen significant improvements made on many fronts, such that many early negative issues with the Viewer have (however grudgingly on the part of LL) now been eradicated.
Speaking as one who has taken the time to get to know Viewer 2 through its various iterations, I do appreciate the effort that has been put into it since the premature launch a year ago. However: and here’s the rub – it is still not a Viewer I will use full-time.
This is not because I refuse to learn the new interface – far from it, or that I simply “hate” the Sidebar. I’ve learned the essentials of using Viewer 2 by spending time with with it, with Kirstenlee’s S20 and S21 and with Firestorm; so actually finding my way around the Viewer isn’t an issue: I actually have no preference for pie menus over context menus (both work fine for me) or anything else of this ilk.
No, the reason I won’t use Viewer 2 is that compared to the likes of Kirstenlee’s hybrids and even the “pre-alpha” Firestorm, and despite the improvements made, it is still far too incongruous and intrusive; rather than form a part of my world-view, it remains something that sits between me and whatever I’m, and making its presence felt, rather than remaining largely invisible to me.
Oddly, this is something I don’t find so much with Viewer 2 TPVs like Kirstenlee’s and Firestorm, both of which – while resembling Viewer 2 – offer a flexibility in use that far exceeds anything seen within either Viewer 2 or Viewer 1.x. In this, the developers have shown themselves to be far more responsive to user feedback than Linden Lab has demonstrated in a very long time.
Beyond this, there is the question of information security. The Client Detection System farrago is not going to go away; and while it is true that the answer does not only reside in a technical solution – we need, as has been stated previously a policy to resolve these issues once and for all – the fact remains that it is through technical approach that people are going to feel more comfortable. As reported, one solution is already on its way. As such, the first Viewer 2 TPV that can offer the same options and capabilities is going to gain my support and adoption at least until Linden Lab demonstrate that user privacy is a genuine matter of concern for them, and provide similar technical capabilities as well as a decent policy on these matters.
Doubtless, there will be other issues along the way that will affect things like security and privacy. But as long as Linden Lab respond to them be either going entirely dark (rather than even give a simple, “We hear your concerns, and we are looking into things and will get back to you”), or by becoming somewhat hostile towards concerns (such as by threatening to revoke JIRA privileges), I will always sway towards a Viewer that, when combined with my own common sense, will help me minimise my exposure to situations where people opt to exploit the Viewer code for their own questionable ends.
And, I rather suspect, such matters will cut to the overall take-up of Viewer 2 compared to TPVs as Viewer 1.x code is wound down and deprecated.