Second Life to Kill OpenSim?

Hypergrid Business has an article about what Linden Lab’s® move to the cloud may mean for OpenSim grids. By knowing what may happen to OpenSim we can infer what is likely to happen to Second Life.

David Kariuki thinks moving to the cloud will allow Linden Lab to run on demand regions. Meaning if no one is in the region, the region would spin down and drop out of the servers, go offline. When someone is on the grid next door or TP’s there, the region loads into a server and spins up.

Brand New Colony (Sept. 2017)

Brand New Colony (Sept. 2017)

The result is fewer servers would be needed, a huge cost saving in hardware and electricity.

It sounds to me like this could be a performance problem. Would we have to wait while the region spins up? Or can a region load fast enough we wouldn’t notice? We don’t know.

What we do know is that when we cross into a new region all the scripts and animations we are running barely miss a beat. While it takes a bit for a region to render, the simulators aren’t rendering regions. They just tell us about them so our viewer can render them. So, the simulators may be able to spin up really fast using something akin to a laptop’s hibernation. With memory and SSD’s getting cheaper it may be possible to load a running server image, eliminating spin up, in milliseconds. But, we’ll have to wait and see if they can do it.

He also speculates that the cost savings may go to fund Sansar. So, we wouldn’t see a lower tier? We don’t know. We do know Sansar funding is a combination of venture capital and revenue from SL, Blocksworld, and other Linden products that fund Sansar development.

Also, we know the SL Team at the Lab is always considering ways to make SL more popular. So, tier has to be on their minds.

So, if SL tier gets cheap, will OpenSim users flock back to SL? From my experience on OSGrid, I would say no, not everyone. A significant number of OpenSim users rage quit SL and hate SL and the Lab. They bad mouth SL any chance they get.

However, some number of OpenSim users use OpenSim for building and testing. I ran a 9-region sim for a couple of years and did building there. I moved finished things to SL when they were done. Plus, they were my play regions.

So, there is a thought in the OS community that the Lab could wipe out OpenSim in a matter of months if they drop tier to competitive prices. Saying ‘wipeout’ may be an overstatement, but they could put a serious dent in OpenSim user numbers. It all depends on how low they can take the prices. I personally think Sansar is the bigger threat.

There is also the matter of what people are willing to pay for land in SL vs OpenSim. What is it worth to own/lease land in a place that has 30,000 to 55,000 people in-world every hour of the day versus 0 to 300 in-world every hour? What does that do to sales?

Then there are High-Fidelity and Sansar. HiFi is open source so you can run your own host, like what I did on OSGrid. I’ve played with HiFi and I have NOT been drawn to it.

Sansar has possibilities. But, we are a year or more away from it being something I can do much with.

Second Life’s move to the cloud should probably be labeled a project of exploration or hope. The servers are being improved with the secondary idea of moving them to the cloud in mind. But, I don’t see it as an inexorable move. I get that it is a more tentative project… or maybe timid or cautious is a better word.

Whenever the Lindens lack solid information the move cautiously. They take the time to build a part, go a little distance and then check to see if the changes meet their expectations. Then the next step and more testing. They either reach the goal or change the plan.

So, this SL-to-Cloud move may take a year or more… or the idea may be given up. It wouldn’t be the first idea to fail. But, the Lindens are confident they can do it, which is why we have heard about it. We don’t hear about ideas they think unworkable. So, I expect them to pull it off. We just have no idea what it will mean for SL and other users of virtual worlds.

But… we can speculate like crazy… :))

11 thoughts on “Second Life to Kill OpenSim?

  1. To speculate on this subject, there could be two kind of pricing. One for people who need to have their region always online. Another cheaper for customers coming to second life only during their spare time. For example, if I only come during the weekend on second life, I wouldn’t mind to wait for my parcel to come up when I login the first time if I can get a parcel at a lower price. If the delay before an empty region is shut down is 12 hours, for example, I wouldn’t wait when I come during my weekends, since I would certainly connect frequently enough to be in this delay. I hope that I’m clear, since suddenly I’m not sure anymore 😁

  2. Mainland implications could be interesting.
    Even at the present level of Mainland occupation it is rare to have 2 empty regions between avitars so I assume all Mainland simulators would remain up and running.
    Would that make Mainland the new “no lag” property? (with regard to region crossings and TP access).
    Given that, with the current setup, there are 20 or more Regions/Simulator one would wonder just how many simulators could actually be “spun down”

  3. I live near a 4 corner region, the neighboring regions are usually blank spots -vacant holes when I fly in—- then they rez— and if I stay a while– they disappear again.

    Really destroys the illusion.

      • Going to the cloud they might change from one CPU per region to either multiple CPU’s per region or multiple regions per CPU… or stay the same. Possibilities.

  4. “On-demand cloud-based regions would only be up and running if people are currently visiting them[…]” – LOL righty! Either this guy has no idea how things work or he blatantly ignored this won’t work at all. Or how does he expect the scripts running on a region continue to execute when the region gets suspended? You know, these scripts that run vendors like CasperVend or sending out automatic updates or…

    • They’re just making a general point in the sense that there are many regions (obviously not all of them) that don’t use such scripts like you’re referring to. So, those can be safely suspended and called on-demand.

      Consider a region that is purely for photogenic photography. If no one is there, I see no reason to keep it up unless you need something very particular to constantly run for whatever reason.

  5. I run three OpenSim servers using DreamGrid for the fun of it and let folks use them for free. I am sure I am not alone. SL cannot hurt my usage that’s for sure.

Leave a Reply

Your email address will not be published. Required fields are marked *