Rodvik shares – and so do I

So here we are in October, and we have the news that the simulator software is undergoing some kind of “critical software upgrade” that is affecting all channels and could see a higher than usual number of restarts over the course of the next week. At least updates on this are being pushed out via Twitter and the Grid Status pages (rather than it being one or the other in a lot of instances of late).

As I’ve been away for the last few days sunning myself and enjoying Swiss hospitality, I’ve no idea as to whether LL have finally recommenced in-world announcements where they know that restarts, etc., are going to impact object rezzing, transactions and so on. If they haven’t, it’s really about time they did.

LL CEO Rod Humble

We also have Rodvik sharing a post as a follow-on from comments relating to the platform made earlier this year, and which is intended to “Update you on each area and share with you our plans for the remainder of the year”.

Certainly, there were some tidbits – but I couldn’t help reading some elements of the post with a sense of disappointment at what was left unsaid.

Let’s take the items Rodvik discusses in order:

Viewer Usability

The two modes (Basic and Advanced) to be merged “soon” allowing changes made to the Basic mode to be integrated into the Advanced mode. Precise details are scant, but Rodvik specifically mentions avatar and camera movement. Ergo, expect to see the Basic mode “click to move” and perhaps the “goto” movement options appearing in the Advanced mode.

One would hope, as well, that the HOW TO button finds its way into the Advanced mode – and has potentially been extended to cover other important functionality, a point I’ve touched on in the past.

How To: cue-card style help in Basic mode: soon to be seen in advanced? One hopes so

At SLCC 2011, Rodvik indicated that the Sidebar’s days were numbered – and it would probably be a better than even guess that this is what he is referring to when he states, “After these modes are merged and deployed, you can expect us to release an improved UI into the Viewer”. I’ve little doubt the improved UI will see other nips and tucks made – but  anyone interpreting this as meaning the V2.x/3.x UI is “going away” or going to receive anything approaching a major revamp … is most likely going to be disappointed.


This was the real eye-opener statement in the most negative sense. “Thankfully this massive feature got deployed on time”. Hardly the best opening line to describe what is supposedly one of the biggest changes to SL’s capabilities. Rather than championing a new and exciting medium, this reads almost as if Rodvik is saying, “Thank God that’s over!”

It is certainly an odd statement to make given LL went to some lengths to avoid being tied-down to any precise dates. Even the May “mesh roadmap” was so full of caveats and warnings one fully expected mesh to slip beyond the broad deadline of “the end of August” for full deployment.

Yet now it would appear that – as I’ve commented upon in the past where LL is concerned – dates were the driving factor in the deployment, not matters of usability. How else can one explain deploying a mesh upload dialogue which was, by LL’s ownadmission at the time, as much a cause for confusion as anything else when it comes to trying to optimise mesh for SL use?

Beyond this, there is the fact that we know from Charlar that mesh in SL is due at least one more “non trivial” update – possibly before the end of the year. Whither then, is the pointer towards this in Rodvik’s comments?

True, in the context of this post, he’s not directly talked about the future of mesh in the past, but that doesn’t mean he cannot give at least a hint that LL are striving to overcome shortfalls and issues. Even something like, “We are working on a series of updates to improve how mesh objects are uploaded to SL, and how you can better understand their resource impact, and we’ll be blogging about that soon,” would be better received than what amounts to something of a brush-off – especially for those suffering under LL’s refusal to respond to the likes of JIRA SH-2374.

Improved Service

This was more honest: improvements have been made, but more investment is needed and will be given. This is good news all around. Of greatest interest to me was the comment relating to revised policies. At SLCC 2011, Rodvik hinted that things could soon get particularly unpleasant for griefers in-world. At the time, I got the impression he wasn’t just talking about deploying tools estate owners could use to handle griefers directly. Are we about to see a revised ToS that also makes the consequences of anti-social behaviour in-world a lot clearer to all?

Better Customer Value

This came across as a very self-congratulatory pat on the back. Frankly, I’m not sure it is due. It is good to see LL moving to make Premium accounts more relevant and attractive – but much more needs to be done in this area before LL can really start to feel good about things. This is particularly true with regards to re-engaging with people like myself; those who were Premium members but who “downgraded” to free accounts.

Rodvik, stipends here don’t cut it – I don’t particularly care if I get “most” of my subscription back in stipends (assuming I pay annually, that is) – I’d still be spending $24 (incl VAT, thank you) a year to enjoy what I already have. Nor does anything else currently on offer present anything approaching a “must have” factor for me – and I’m not alone in this. You need to give more thought to making Premium attractive to your established user base.

Shiny New Things

This section offered perhaps the most interest. At SLCC 2011 (again), Rodvik made mention of the introduction of NPCs – Non-player characters with whom it would be possible to interact with to a degree, and which can be set to perform specific tasks.

While bots, etc., have been possible within SL, they’ve tended to rely on a mis-match of enabling technologies: external data servers in some cases, or heavy reliance on server-side scripting resources (such as with many breedables) that draw down the anger of those impacted by such entities.

NPCs: New opportunities

Second Life itself actually had the capabilities to manage artificial life at one time – specifically plants and animals. Whether or not this capability is being revamped and re-introduced into the platform or not is really beside the point (although if it is there and is robust enough, making use of it would seem to make sense).

What is important is the fact that a massive capability is going to be added to the platform that could have a huge range of potential applications in-world. NPCs alone – human, alien, animal – have a wide range of applications in the likes of role-play and so on. As Rodvik states in his post: imagine a town filled with NPCs going about their “business”, allowing for a certain degree of interaction and so on, generating enormous depth to any gameplay or role-play environment.

Nor does it end there – assuming it can be done robustly and intelligently – the opportunities for using such “artificial life” capabilities has implications well beyond those of gameplay, many of which fall into the realms of education and practical research. Self-navigating agents? The avenues for schools and educational institutions to engage in things like robotics are simply mind-boggling.

Native AI in SL: potential educational / research / modelling benefits beyond RPG

According to Rodvik, we can expect some of these new capabilities to commence testing in December of this year, although (wisely) no roll-out data has yet to be pinned to the wall. Given the impact and benefit such capabilities could bring to Second Life, I really hope that LL strive for two things:

  1. Make sure the capability is properly developed and implemented – sorry, but at the end of the day, mesh came across as a half-arsed implementation to many, and I’d suggest further that it is a practical demonstration why (again) the push, test, polish, test, polish approach can actually do SL as much harm as it can potentially do good.
  2. They openly engage with a wide range of potential “beta testers” from within and without the SL community to ensure as wide as possible spread of potential use cases are identified and catered for and are the determining factors as to when the functionality is rolled out. Let’s not have the calendar again dictating what can and cannot get done in the “first release”, because many using SL already equate “first release” with, “That’s all LL are prepared to do”.

9 thoughts on “Rodvik shares – and so do I

  1. the impression I got was that the npc is next years’ summer blockbuster (as mesh was this years’) and that will be the excuse for not actually fixing the technical issues that their current user base would rather have fixed. I suppose there’s more money in games platforms than there is in open ended virtual worlds so it makes sense to go for that market.

    hey ho and all that.


    1. That is the sad element. We’re now looking to the Next Big Thing rather than discussing refinements to the Almost Was Big Thing We Just Had.

      The whole AI / AL thing has a huge potential, but I do worry that LL will simply focus on the “gaming” side of things and let the rest sail right over their heads. Let’s face it, despite their platform currently being used for conferences, etc., the fact is that they are out of the educational market to a get extent purely on the basis of pricing. It’s odd, but I came away from SLCC with a high level of optimism; the last month has wiped a good portion of that optimism away. Even so, I’ll keep fingers crossed. For now.


  2. Hey folks a new toy…Aww..Broken! Hey, never mind look at this new “shiny”…Ooh! They Never Learn.


  3. Hmmmmmm. This is intriguing! For purely personal reasons, I’m very curious about what LL comes up with. My line of academic research has been mostly about “giving tools to people for things that Linden Lab doesn’t want to implement”, and since NPCs never really were on LL’s horizon, part of my PhD work is about integrating AI-driven ‘bots, which is pretty much all we can do at this stage.

    It would be rather ironic if LL came up with something in 1-2 years that pretty much renders all my work useless, so I’m actually sorry that they’re so secretive about it… I’m certainly quite interested in see what Linden Lab comes up with!


    1. The AI – or “A-Life” capability was there in the early days, as I’m sure you’re aware. Some have speculated this will be the basis of the new capabilities – but as you say, with LL being so secretive, it is hard to tell. Certainly, I do feel for you and others who may see a lot of work in this area undone should LL pursue a course of their own choosing.

      In this, I feel they are still getting the balance wrong: on the one hand they are banging out NDAs relating to the enhancement of existing capabilities when they should perhaps be thinking about being more up-front and open in their engagement with users. Then we have this situation where one can understand the need for security – AI / A-Life could have huge repercussions across the metaverse as a whole for those that develop the capability in a really robust manner first – and as such, it would perhaps be more fitting for LL to offer to engage with those already researching / developing the capability, and the use of NDAs would seem entirely appropriate in such an instance.

      I do continue to wonder as to which void in particular LL will be aiming at; going for a relatively basic capability suitable for the creation and deployment of bots, etc., intended for the “gaming” aspects of SL – or something broader & deeper that could, as mentioned in my article, have far wider uses….? My suspicions at this point is it’ll be more the former, and likely to be rolled out (as others have commented here) as next summer’s “blockbuster” feature. Something which, in and of itself given recent developments vis-a-vis mesh, doesn’t actually bode well in the usability stakes.


  4. What’s somewhat ironic is that doing more to, say, improve stability would do FAR more to make SL more game-friendly than adding in AI NPCs would. NPCs are just a new toy, and while nifty don’t change SL’s underlying futziness. Unfortunately, I think we all know which option is harder, and which is more sexy.


    1. In fairness, LL are working on stability, and have been for some time. Progress may be slow, but they are attempting to sort a lot of long-standing issues out at both ends of the spectrum – Viewer and server.

      As to new stuff – this is where I do feel sympathy for LL. Things like NPCs and mesh are responses to user requests made over and over and over again through the years – years in which LL get hammered for not implementing them. Then, when they finally move on such things, they still get hammered for doing so.

      We’re also quick to boot them very hard when errors / mistakes occur that lead to code breakages – but at the same time, when improvements are made to the system, the majority of us take it for granted.

      I would suggest that it is a mistake to think that the Lab are only focused on the new and “sexy”; it’s what is newsworthy, sure, so that’s why they headline it. But it doesn’t mean they are carefully tinkering away elsewhere to fix things.

      Of course, it has to be said that they don’t exactly help themselves when they make comments that imply that some matters are a case of, “Job done, move on”. This pretty much seems to be the message where mesh is concerned; both with Rodvik’s blog comments, and with Charlar’s JIRA admission that there are no major resources targeted at mesh development any more, and things are going into the “backlog”.

      When it comes to fixing things, one of the biggest problems is the sheer lack of communication from LL on the matter. This is particularly frustrating when ushc issues – be it the return of the sim lock-ups experienced when avatars teleport, or the issue relating to upmarket nVidia graphics systems – are widely known.

      While it would be unfair to expect weekly updates on issues like these, LL could go a long way to helping themselves if they added a section to the end of their monthly updates that provided a brief update on them, given they are so widely known and are causing upset and frustration among users. Even a couple of lines showing someone is trying to address the issue is a damned sight better than continued silence over periods that, in some cases, are approaching months.


Comments are closed.