OpenSim gains record-breaking 2,177 regions

The top 40 largest OpenSim grids gained a record-breaking 2,177 regions since mid-August, the single biggest monthly gain in history. The new regions pushed the grids to 23,952 total regions, also a new high.

OSGrid was responsible for almost a quarter of the growth, with 498 new regions, for a new total of 10,323 regions. OSGrid is a non-profit grid that allows people to connect regions hosted at home for free, as well as regions hosted by low-cost third-party hosting providers. OSGrid is the largest grid running on the OpenSim software.

Total regions on the 40 largest OpenSim grids.

In fact, all the top-40 grids either maintained their land area or gained regions, with the exception of the education-oriented JokaydiaGrid, which lost just three regions.

Kitely was in close second place with 365 new regions, for a new total of 2,493 regions. Kitely is the largest commercial grid running on the OpenSim software, and is unique for its pricing model — it charges by the hour — and in the way the regions run in the cloud and are only booted up when they’re needed.

Kitely’s megaregions

Much of Kitely’s growth could be attributed to its latest new feature — megaregions.  Customers can now get their regions grouped together into larger landmasses — squares of four, nine, or sixteen regions. The “megaregion” option removes all border crossings, making the new areas look and act like single, large regions, with no border crossing hiccups and easier vehicle scripting.

Kitely CEO Ilan Tochner

“Kitely now offers 16-region worlds that can hold up to 100,000 prims and 100 avatars for a price that is lower than what most other grids charge for a single region that can support a lot less prims and avatars,” Kitely CEO Ilan Tochner told Hypergrid Business.

Kitely users also launched an independent discussion forum, Kitetalk.

“As a result of these developments we saw a jump in the number of interesting worlds made public, people have been spending more time inworld and the number of people who’ve upgraded to our Gold Plan has increased significantly,” said Tochner.

The company’s Gold Plan allows for unlimited use of the Kitely grid and 20 regions for US $35 a month. Each additional region is $1 per month.

“People have been taking advantage of our offering to open many such big worlds and the number of regions we now host has grown significantly as a result,” he added.

In other news, Kitely began working on transfer stations, the company announced today., a move which will help the grid turn on hypergrid connectivity when the next, more secure, update of the hypergrid protocol is released. Users will be sent to transfer stations if they attempting to log in or hypergrid teleport to a Kitely region if its offline and doesn’t load up fast enough.

Popularity contest

For company and school grids, popularity is not an issue — the grids are set up for a specific purpose, and if they meet that purpose, then they are successful. The same is true for grids run by niche communities.

But when it comes to social grids, the rule is: the bigger and busier, the better. People looking to make new friends will go to the grids that already have the most users. Merchants looking to sell content will go to the grids with the most potential customers. Event organizers looking for the biggest audience… you get the idea.

With that in mind, here is our listing for the 10 most popular grids this month.

Top ten most popular grids:

  1. InWorldz: 6,621 active users
  2. OSGrid: 3,515 active users
  3. Avination: 2,365 active users
  4. Island Oasis: 902 active users
  5. FrancoGrid: 880 active users
  6. Metropolis: 401 active users
  7. Craft World: 394 active users
  8. 3rd Rock Grid: 368 active users
  9. Kitely: 242 active users
  10. German Grid: 239 active users

With the exception of Island Oasis and Kitely, commercial grids lost active users this month. Avination lost 32 active users, 3rd Rock Grid lost 31, German Grid lost 30.

The big gainer was Island Oasis, a small commercial grid of just 90 regions, which gained 341 active users. In fact, it gained more active users than all other top-40 grids combined.

Kitely was in second place, with 47 new active users.

The total number of active users on the top 40 grids dropped to 17,303 this month, 61 actives less than August’s 17,364 users, mostly due to the fact that ReactionGrid has shut down its OpenSim grid and no longer appears in our stats.

InWorldz — despite gaining 1,794 new registered users — lost 297 active users compared to last month. That still puts it far ahead of other grids, including Avination.  InWorldz and Avination are the top two closed, commercial grids and directly compete for residents. Both make their money primarily from land rentals.

And InWorldz continued to pull away from Avination when it comes to land area.

InWorldz continued to gain more regions than Avination this month.

OpenSim migration accelerating

Second Life lost 312 regions since mid-August, according to data from Grid Survey. Second Life now has a total of 28,902 regions, a decline of 2,238 regions from this time last year, and a drop of 2,983 regions from its peak in June 2010. Second Life no longer reports active user numbers.

Groups that have made the migration are reporting that they can get more land and better service for much less money in OpenSim.

In addition, many OpenSim grids — over 100, at least count — are hypergrid-enabled, allowing users to easily teleport from one grid to another to visit friends, shop, or attend events. The biggest problems — security, lack of content, and lack of users — are also improving steadily.

“OpenSim is a very likable place, and so are the people involved in it,” said Per Eriksson, founder of Australian virtual design firm Troppo Design. The company runs the Troppo Club OpenSim grid. Its latest project is The Lost Castle, a 29-region recreation of the historic 1613 Mori Tadamasa’s Castle from Tsuyama, Japan, which is a setting for machinima-style animated film production.

At first, Eriksson said, he thought that Second Life and OpenSim would grow closer together. However, despite a successful Second Life-to-OpenSim teleport test by IBM and Linden Lab in 2008, the inter-grid teleport project was abandoned.

Per Eriksson

“I can’t see that happening now,” Eriksson said. He said that he will shut down his last Second Life region before the end of this year. “The hypergrid is really what it’s all about.”

Both of Eriksson’s OpenSim grids are hypergrid-enabled, which means that users can teleport to and from other grids. Today, the hypergrid allows users to make friends, send instant messages, and create landmarks. The next version of the hypergrid protocol is expected to have new security standards, allowing creators to decide whether individual objects can leave their home grids or not.

Jokaydia’s Jo Kay said that she is cutting back on her Second Life holdings, as well. The education-focused Islands of Jokadia community has been reduced from four regions to one region.

“The original Jokaydia sim remains and will continue to be our headquarters inside the walled Linden garden,” she said. Meanwhile, the hypergrid-enabled JokaydiaGrid now has 83 regions.

And Chris Collins, founder and manager of the Center for Simulations & Virtual Environments Research at the University of Cincinnati, recently wrote a much-discussed post calling on Second Life users to expand to other platforms. Collins has also long been active in Second Life, including running last year’s community convention and serving and the director of the education-oriented Chilbo Community, which has also been reducing its Second Life land area.

“Five or six years ago, you could not have found a more enthusiastic and engaged supporter of the Second Life platform than me,” Collins wrote.

Since then, she wrote, it has become clear that Second Life has abandoned its early vision of a metaverse future.

“They are simply not interested in having their virtual world participate in any of this metaverse stuff at all,” she wrote. “Does this mean you have to leave Second Life? No of course not, even I haven’t done that. I still have projects for work in Second Life … and I still have a strong connection to my friends, colleagues, and communities in Second Life. But I have tiered down, way down, and I have begun to invest my time and money largely elsewhere… I think if you care at all about making the metaverse a reality, that’s what you should do, too.”

Collins, also known as “Fleep Tuque” in-world, is the owner and founder of the 16-region hypergrid-enabled FleepGrid.


The following 33 grids seemed to be suspended this month: Carohome5, Carpathes-Grid, Cholul, Dawn Grid, Destiny Zero, Fantasy World, Fargis, Free Open Grid, GstarCAD, HewittSim, IceGrid, Kelly & James Fantasy Playground, KorolovGrid, Meridian Grid, My3dLife, NewtonGrid, OpenMoodScience, Pandora, Planet Envee,  Sirius, Southern’s Place, Swondo World, Tanjas World, The Hive, ThunderLife, Verkosis , Winter Heat, World of Eternity, Xscape Grid, Yeah Right, and You3D.

In addition, ReactionGrid’s namesake grid shut down, as ReactionGrid the company refocuses away from OpenSim to development of its propritary Unity-based Jibe virtual world platform.

New grids

This month, we’ve added several new grids to our database, including The Lost CastleVaginchen TwigCSC9N5 Virtual World, MoehlenhoffPonyLandSirinHGpole, Virtual Marin, and the Greek-language AVIlife.

If there’s a public grid we’re not tracking, please email us at [email protected]. There’s no centralized way to find OpenSim grids, so if you don’t tell us about it, and Google doesn’t alert us, we won’t know about it.

Well, there is one centralized way to find some grids — the New World Studio grid directory. It currently lists 1,032 different worlds, totaling 9,398 regions. There are many duplicates on this list, however, and few worlds were up when we checked.

However, there are also statistics for one popular version of OpenSim, the Diva Distro, a four-region, hypergrid-enabled, pre-configured minigrid.

The Diva Distro has been downloaded 93 times over the past month, possibly because people were waiting for the new update, which was released earlier this month — and updated yesterday. The total number of Diva Distro downloads now stands at 16,379. This does not mean that there are sixteen thousand mini-grids out there, however — someone might download the software but never use it, or download it once and use it to set up many grids. And it doesn’t include the Diva Distros used as part of the New World Studio distribution of OpenSim.

Diva Distro is also part of Sim-on-a-Stick, a version of OpenSim packaged to run on a USB stick. According to Sim-on-a-Stick creator Ener Hax, the USB-friendly OpenSim package has been downloaded 1,020 times over the past month, bringing the total of these downloads to 14,795.

Meanwhile, according to data from The Hypergates, the number of hypergates increased by 53, to a new high of 613. Hypergrid jumps decreased from 2,508 to 2,484 this month. The Hypergates did not report the total number of travelers this month.

This data is very limited, however. For example, not all hypergates are part of The Hypergates network — anyone can create their own hypergate by dropping a script on any object, such as our touch or walk-through single-destination hypergate script. In addition, many people do hypergrid jumps without using any gate at all, simply by typing a hypergrid address into Map-Search, or by using a hypergrid landmark created during a previous jump. There is currently no way of tracking that traffic.

Our own Hyperica directory now tracks 102 grids that are accessible via hypergrid, up from 101 last month. (We’re running behind on adding grids to the directory.) This past month also saw 86 unique visitors to the Hyperica in-world hypergate terminals, up from 82 the previous month.

September Region Counts on the Top 40 Grids

We are now tracking a total of 360 different publicly-accessible grids, 194 of which were active this month, and 156 of which published their statistics. Many school, company or personal grids do not publish their numbers.

The raw data for this month’s report is hereA list of all active grids is here.

Related Posts'

Maria Korolov

Maria Korolov is editor and publisher of Hypergrid Business. She has been a journalist for more than twenty years and has worked for the Chicago Tribune, Reuters, and Computerworld and has reported from over a dozen countries, including Russia and China. Follow me on Twitter @MariaKorolov.

21 Responses

  1. Ener Hax says:

    thanks for the great post and all of the insight into the various grids =)

    • I want to thank SpotON3D for sharing their region counts, and InWorldz for sharing their active user numbers.

      For commercial grids in particular, growing land area is a sign of the health of the business.

      And traffic numbers are very useful for folks and businesses looking to grids to check out. And while many would gravitate towards the busiest grid, some people are also looking for new, smaller markets to explore, or a more personal, cozier grid for socializing. The traffic numbers can show which grids are gaining active users, etc…

      Meanwhile, here are some extra stats:

      Total number of regions on the 156 grids that release these numbers: 25,545
      Total number of registered users: 265,015 (down a bit since last month because ReactionGrid’s gone)
      Total number of active users: 18,316

      Total number of active users on hypergrid-enabled grids: 6,859.

      (This includes only the grids I definitely know are hypergrid enabled — I haven’t checked all of them, and doesn’t include many hypergrid-enabled personal and school grids that aren’t in our directly, such as the Georgia school district’s grid. But then again, some of the smaller grids count hypergrid visitors as part of their active users total.)

      So there are more users on the hypergrid right now than on any individual commercial grid. And with hypergrid-enabled friends, instant messages, groups, and landmarks its getting easier and easier to reach these folks. Of course, commercial grids have a marketing advantage in that they can send out messages to all of their users, or put up announcements on their login screens. There is currently no way to simultaneously reach all users of the hypergrid, and probably will never be — it’s like trying to reach all the users of the Internet.

      So, for merchants, being on a commercial grid gives them the opportunity to buy ads that reach all the members of that grid. (Kind of like buying ads on Facebook.)

      •' Dan says:

        The love of merchants goes right against the socialistic nature of Opensim and Open Source, it kills friendships, the sharing of creative knowledge and just attracts the exploiters and sharks. Second Life is full of shopping malls, hawkers, griefers, paedophiles and other perverts, do we really need a repeat of all the above with Opensim grids, and how are the Opensim grids an advancement over what Second Life has already done?

  2. These numbers are very good to have. Although I would however like to see clarified that active members and total members are two different things. For example when I log into Inworldz I only see 200 to 300 people logged in at a given time. The number they are giving is actually the total number of registered accounts but not the active number.

    • Total registered users: the number of people who have ever set up an account on a grid. InWorldz now has 63,929.

      Active monthly users: the number of people who have logged into the grid at least once over the past 30 days: InWorldz now has 6,621.

      Currently logged in: users who are, right now, logged into the grid. That’s the number you see in your viewer, usually around a couple of hundred.

      • Okay that sounds a bit more accurate. I usually don’t go by the exact number they give as that number also takes into account alternate avatars, avatar accounts used as bots, and multiple logins from the same account (this one is possible). This probably is something that should be figured out as to how to break those numbers down to get a more accurate reading.

        • The 30-day active user number DO include alts but do not count additional logins from the same account. So if you logged in 100 times or 10000 times, you would still be counted as one active user.

          There’s nothing much you can do about alts. Even if two users log in from the same IP address, on the same computer, it could be one person with two accounts, or two people who share a computer.

          Maybe someone can do a survey, asking people how many alts they have, how many grids they’re on, and how big those grids are. See if there’s any correlation. I’d guess — with no evidence either way — that the larger the grid, the more likely people are to use alts because there’s more stuff to do with them.

          Of course, if a grid ranks traffic to individual regions, then region owners might use alts to beef up their traffic numbers. The InWorldz folks tell me that they don’t do that, so there’s no incentive to inflate the stats that way.

  3.' Minethere says:

    These numbers are good for getting –an idea– of grid health…but as fred points out they are not really all that accurate…in addition, for example, as anyone can see on the inwz splash page –they have a large number of –sponsored sims–if one watches this number, it fluctuates quite a bit…and it is used for the total number of sims count

    so, if, for example, a grid owner knows when the numbers are crunched, they can play with their numbers quite a bit [I’m not saying any grid owner would be dishonest, of course]

    But I did know that IO numbers were doing very well, as I am one who helped it to happen…lol

    • Yes, grid owners could rig the stats in any numbers of ways. And some have. (Whether deliberately, or accidentally, I don’t know and I’m not going to comment.)

      I’ve written about this before.

      The thing is, it’s hard to hide that kind of thing. And it’s bad for your grid’s reputation if it comes out.

      Plus, in my experience, grid owners have a lot of other stuff to worry about, and a lot of things they can be doing to promote their grids which would be way more effective, but they probably don’t have time to do because they’re busy dealing with all their new customers.

      So say a grid owner decides to roll out a bunch of new regions to cover up the fact that they’re losing customers. Next month they’ll have to add even more regions. And add more regions the month after that. Meanwhile, people are going to start noticing that a bigger and bigger proportion of the land area is empty. A few empty regions are to be expected — the new owners haven’t arrived yet, or a public construction project has been planned but hasn’t started yet. (Everything takes longer than you think.)

      But too much empty land is a worse sign of health for the grid than bad monthly stats would be.

      •' Minethere says:

        This is very true, of course.

        In addition I would like to point out that inwz private sim numbers have been dropping over the last couple of weeks…this, along with the reduction in active users…along with the numbers you will see next month, will be quite interesting, and telling.

        IO has some really great potential…and if one looks at their land prices and options…as people are doing…just watch the numbers next month.

        It is on a roll right now, with much in the works….is this an ad for IO? well, ya, but I am not on the staff…maybe somebody can pay me tho??? lol

        •' Dan says:

          In my honest opinion IO are just like IBM wannabe Microsoft’s or in this instance Second Life. They are not going to go anywhere and are not the future it’s just a case of they want to be the Lindens of their World (Walled Garden) and don’t think for a minute that if a lot of people by miracle decided to join IO do you think they would not increase land prices like LL has, just because LL has setup fees many Opensim grids have them albeit much lower at present anyways. What has IO given back to Opensim? Nada.

  4. RULosingHair says:

    “For commercial grids in particular, growing land area is a sign of the health of the business.”

    How do you come to this conclusion? Concurrent users have the purses.

    How about counting skyplatforms as well?

    • Active users are important for in-world merchants, and grids to make some money from taking a cut from in-world transactions and selling in-world currency.

      No commercial grid is releasing their internal economic numbers, however.

      Meanwhile, commercial grids rent regions directly to users, typically around $60 a region. Third party hosting providers sell similar regions for around $30, so I’m figuring that commercial grids have $30 to spend on things like customer support, marketing, in-world activities, and providing common areas such as welcome areas, sandboxes, and other grid-owned locations.

      I’d guess that anything left over, grid owners are plowing right back into the grids — improving the technology, the Web interfaces, increasing support, etc…

      It becomes a virtuous cycle — the more money is coming in from land rentals, the more money there is to improve the grid, which makes it even more attractive to users.

      Meanwhile, a grid that has a steady decline in land faces the opposite problem. They’re geared up to provide a certain level of service and support based on their revenues, and when the revenues fall, they have to either invest money to prop it up, or cut back on services or technology.

      • RULosingHair says:

        Statistically spoken, your definitions of simulators, regions, lands and other whatnot etc. is not consistent with what people call “statistical discreteness”, your input compares apples and pears.

        So your output counts and derived conclusion-making unfolds to be a haywire of unvalidated inductive statements and hypotheses.

        How does all that conclusion-making match with this piece of deductive datamining and hard raw data

        • Hi, James OReilly.

          The data comes from the grid operators themselves. There are common definitions for “region”, “registered users” and “active users”.
          You can see the full report here, and make your own analysis from it:
          If you know of anyone else collecting OpenSim usage data and putting our reports, I’d love to know about it.

          • RULosingHair says:

            Business readers know about statistics and statistical discreteness, even if you don’t.

            1. Well as to the gibberish “lack of statistical discreteness” you are creating…

            You know Kitely uses one simulator with a capacity of 100k prims 100 avatars, meaning one Kitely world, or one region from the very beginning.

            – Now one Kitely simulator can be subdivided into more regions or land as you now call it, the capacity of 100k prims 100 avatars remains the same in the course.

            – Well land can also be extended by adding skyplatforms, so you could add like maybe 100,000 platforms, but the you would be magically transforming prims into a new definition of land for your bodycount.

            – One Aurora simulator offers “unlimited land” only limited by your hardware RAM.

            You are inflating the number spacial words only to inflate your bodycount into wasteland. That wasteland is also tangibly visible in the ups and downs of your monthly bodycounts.

            So how can you jump the gun into economical conclusion-making for knowledeable business readers with all that statistical haywire and in-discreteness towards wasteland?

            2. You missed answering how all your conclusion-making fits to the discrete and declining search results of Google Trends…

            This information does not come from those developing utility providers and sellers, moreover, from procurement driven clients seeking information to the term Opensim.

            As we have a Buyer’s Market and not a Seller’s Market, it would seem much more viable and bona-fide to see the procurement and marketing side of things, rather than the seller’s side of things.

          • I’d love to see some buy-side data. Unfortunately, Google Trends isn’t really a reliable source of information for buying interest given the small size of today’s market.

            For example, look at the Google Trends for Apache Server:

            And here’s the graph of the number of websites running on Apache:


            As you can see, the two run in opposite directions.

            The problem with getting buy-side data is that the OpenSim penetration is too low to do a general survey, individual buyers too hard to find — nobody is tracking downloads — and vendors are reluctant to disclose customer counts.

            The region counts are not a perfect measure, but it’s currently the closest thing we have to a proxy for buyer interest.

          • RULosingHair says:

            From a Client Procurement Viewpoint or Buyer’s Market Approach, the following applies:

            1. Google Trends is indeed a validated tool for marketing… just check this article… Google’s New Tool is meant for Marketers

            2. Next Google Trends, you have Google Insights, and the search for Opensim delivers the “same declining pattern” as in Google Trends


            “The collection of search queries that people type into Google has been called a “database of intentions” since it is a window into what people are interested in and, sometimes, what they are interested in buying. Insights allows anyone to analyze the results in much greater detail than Trends does. ”

            3. I mean really, does Opensim actually need a repeat “Second Honey Trap” and “Wastelands” like the 2006-2011 ways of Second Life? It’s a dis-service to repeat that known mistake.

            Feeble conclusion-making shot from the hip is not working sustainably… and also prevents a Turnaround Strategy.

            A Turnaround Strategy means:

            – Getting the Buyer-Side Decline to reverse and go up,

            Who cares if that happens on one simulator.

            Zillions of users on One simulator is better than One user on Zillions of simulators (respectively regions, worlds, platforms, lands, parcels, megas, and other whatnot etc.).

          • James —

            Again, if you use Google Insights on “apache server” you’ll also see a steady declining trend. It may be a useful tool for marketing, but it’s not an accurate indicator of buying interest. At best, it can help you choose search key words to use to optimize your Ad Words advertising campaign.

            The best way to track buyer interest is to follow the money. Analyst firms do this — they track shipments, purchases, and survey buyers about their spending.

            OpenSim is too small a market right now for analyst first to cover — plus, there are no publicly listed companies in this space, except for IBM and Intel, and OpenSim doesn’t really affect their bottom lines. So we’ll have to wait for the analyst research.

            All the companies involved in OpenSim hosting, including Kitely, are privately held, and don’t want to see their revenues, expense ratios, or profit numbers in print.

            The region data we have is the only currently available indicator of spending, and it’s not a direct indicator, but a proxy, more useful in terms of trends rather than single-month data.

          • RULosingHair says:

            Well, your monthly statistic is certainly not a scientific indicator of Opensim Grid Health either…

            Too many questions about its system design and discreteness are left open and unanswered, you admit statistical gaps…

            Things are going haywire if you make sweeping economical conclusions based on this system full of gaps.

            Unfortunately, some people get the same honey-trap twice, like in Second Life.

            Consumer Detriment is the running term for this.

        • And if you have an actual question, can you please ask it in plain English? I don’t mind you asking questions about the data.
          But comments filled with gibberish aren’t useful to anyone, and if I can’t figure out what you’re saying, I’ll moderate them.

          If you’re saying that there’s no standard definition of the word “region”, that’s not the case. Prim counts and avatar capacities do vary greatly, as does uptime. OSGrid has many home-based regions that aren’t up all the time, so do many personal grids. Kitely regions are only booted up when folks teleport in.

          So there are variations in how regions are implemented, but the standard region is a land area 256 meters by 256 meters. That’s pretty specific and it’s something you can compare across grids.