Digital Lisbon raising funds for grid hosting

Virtual Lisbon. (Image courtesy Carlos Loff.)

Virtual Lisbon. (Image courtesy Carlos Loff.)

Virtual city recreation project Digital Lisbon, currently a part of the AviWorlds grid, has launched a crowd-funding campaign on IndieGoGo to set up its own, dedicated world.

Carlos Loff

Carlos Loff

Founder Carlos Loff is looking to raise $2,000, which will cover half a year of hosting and provide money for events and content. Loff plans to host the 144-region project on CloudServe, on the “VWorld 4” package, which costs $140 a month for six CPU cores, and includes a website and member control panel. The new grid will be called the Lisbon Grid.

Supporters will get rewards such as special access, free land plots, and even entire regions, depending on the level of support, and there is no risk to making a pledge.

“Either the whole goal is reached, or all donations are refunded,” said Loff, who is an experienced virtual worlds builder and has been a community manager in Second Life since 2007.

Belem tower in Digital Lisbon. (Image courtesy Carlos Loff.)

Belem tower in Digital Lisbon. (Image courtesy Carlos Loff.)

“Digital Lisbon is by far the biggest city replica in the whole metaverse, with nothing less than nine clustered megaregions of 16 regions each,” he said. New and noteworthy in-world destinations and activities include watersports on the entire length of the Targus river, a hypergrid job center, and a highly accurate replicate of Lisbon’s best known monument, the Belém Tower, he said.

The Lisbon Grid will have an almost 70 percent degree of geographical scale accuracy and structural continuity, and will cover the equivalent to nine square kilometers of virtual land, he added.

Virtual Lisbon. (Image courtesy Carlos Loff.)

Virtual Lisbon. (Image courtesy Carlos Loff.)

“People will be able to furnish their own homes, attend concerts, fairs and workshops, explore virtual museums and monument replicas, rent a store or office and run a bar, theater or school,” he said. “What makes Lisbon special are elements such as unparalleled city size and continuity, building and texturing accuracy, full usage and interactivity of buildings, objects and living things, everything brought alive with many community activities and dynamics, including hiring many artists and lecturers.”

The crowdfunding campaign ends on October 18. As of this writing, the project has raised $150 of its $2,000 goal. If successful, the launch of the grid is scheduled for January 2016.

This is the Loff’s third attempt to get the project funded. Last spring, Loff launched a crowdfunding campaign on GoFundMe to build Digital Lisbon on Kitely. Then, in July, he relaunched that GoFundMe campaign as part of the move to the AviWorlds grid. The GoFundMe campaign raised a total of $210 over the course of six months.

To visit Digital Lisbon while it is still under development on the AviWorlds grid, teleport to You can follow the project on Google Plus, Facebook, and Twitter.

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.

26 Responses

  1.' skylifegrid says:

    FYI Carlos The Server your on now Is much faster than the “VWorld 4” you would be looking at … What is the need for 2 places ? We sure hope you will stay with us at AviWorlds!

    •' Carlos Loff says:

      I will allocate very few regions to it – Lisbon + about 10 more so Im counting on that relation server/regions – I do not plan to erase the Aviworlds Lisbon because I intend to develop different plans for each one – The main reason I must use my own grid is the need of full specs control because a Grid that is built to serve the whole Hypergrid visitors needs very specific tunning

      • Frank Corsi says: manages its own Dedicated Cloud system. The vWorld4 is normally placed onto an Intel based server node. E5-1620v2 or an Intel E5-2650v3. Both options are far superior to anything at can find at

        •' Alex Ferraris says:

          Frank being that virtual worlds require many processing instances and within these instances even more processing a VPS is not a very good choice in my opinion.
          A VPS (virtual server) is a division within a server creating that specific space and yes with its own IP address. That means a server is processing all these VSs using a limited amount of processors. A server subdivided in many compartments processing perhaps more than 10 grids.
          It is not the best SOLUTION in my opinion. I may be wrong but even Dreamland Metaverse , Mr. Snoopy has an entire explanation as to why not use a VPS.
          and I quote “Most of our competitors use virtualization technologies, Virtual Private Servers (VPS) or Cloud Computing, which are unsuitable for massively computing and memory intensive real time applications like virtual worlds. Or they run many OpenSim regions on the same OpenSim process, to simplify their administrative tasks on the cost of lower service quality and less stability.”

          •' Leslie Harris says:

            As CEO of the corporation under which CloudServe operates, I wish to thank you, Alex, for your input; however, I understand you can’t know how this hosting company operates. I am sorry you have placed misinformation in your comment to the news article and attributed incorrect analogies as to CloudServe’s functionality. As with other corporations using the cloud, such as Adobe, Intel Corporation, and Microsoft to name only a few, the information you presented was based on perhaps lesser quality cloud servers which have no correlation to CloudServe’s set up.

          •' Alex Ferraris says:

            Well I am only using comum sense plus experience in the field. If you subdivide anything it gets weaker. A rope for example. You subdivide a rope it will brake easy.
            My experiences with cloud based hosting and VPS is not a good one and never have been. I even showed a quote straight from Dreamland Metaverse above stating the same as I have.
            The more processing requests you place in a single server less power it will DEDICATE to one GRID.
            The more students in a class the less they will learn. It is just as simple as that.
            Carlos has mentioned in one of his posts that he could not believe how powerful and lag free AviWorlds was and he wanted AviWorlds secret because he had been in many other grids including ones HOSTED BY CLOUD SERVE and never experienced anything like in AviWorlds.
            I do not need to know how Cloud Serve operates. I am sure it is a very respectful company and this has nothing to do with them.
            VPS is not better than a hard dedicated server that is paying attention to ONE GRID ONLY. Simple and clear.

          • Frank Corsi says:

            Research Cloud Hosting plans VS Virtual Machines!
            I HOPE we can all get along this year, competition is great if your educated on the real facts.
            The “Mine is better than Yours” is going to become childish!

          •' Alex Ferraris says:

            Yes Frank. I just want to know the difference between VS Virtual Machines and a Dedicated Server hosting ONE grid only one robust. But yes we can all get along as you said. Competing is great ! Love it!

          •' XMIR Grid says:

            A physical server will always perform better than a virtual server running on the same hardware. – Always.

            The requirements of running a simulation like OpenSim are different then running websites or streaming services. Particularly with mono where you run a virtual machine environment and the mono JIT compiler on top of a virtual machine running the operating system, there is a LOT of overhead compared to running mono on real iron.

          • The virtualization layer does add a bit of overhead. Instead of taking one machine, and running software on it, you’re installing a hypervisor that makes it think there are several machines instead of just one — then each of those machines needs an operating system, etc…

            So technically, it is true that virtualized cloud setups are less efficient than “bare metal” physical servers.


            That’s not how clouds are actually used.

            Some examples:

            1. In the old days, companies would set up a machine for each application. If an application didn’t use up all the resources on that machine, those resources were wasted. In data centers based on this model, up to 90 percent of all server capacity winds up being wasted, since not applications are running all the time, and not not all running applications are running at maximum all the time. Virtualization allows you to get this wasted space back, and assign it to whatever application needs it at the time.

            2. In the old days, if a company needed a new machine, they would have to order it from the manufacturer, wait for it to be delivered, then install the operating system and the software and hook it up to the network, and test it to make sure everything was done right — a process that could take days, or even weeks. With virtual machines, a company orders a new machine from a cloud provider and gets it in minutes (or seconds) and then loads it with an image — a pre-made package of operating system, software, and everything else needed, all pre-tested, and ready to go. Oh, and instead of having to pay the full price of the server up front, the company would have no up-front costs, and would just pay for the time it spends using the virtual machine.

            3. If a company suddenly needed 1,000 new servers for testing, or because of a holiday spike in business, or a one-time event of some kind, in the old days it would be out of luck. In a cloud environment, it would take it’s pre-made image, and install it on 1,000 new virtual machines, then take them down when it was done.

            The trick, when renting virtual machines, is not to think about the underlying hardware, but to ask what you’re getting as part of that particular virtual machine — how much processing power? How much storage? — and plan accordingly.

            Kitely, for example, runs 100% on the Amazon cloud and has had excellent performance results.

            In fact, most businesses today — both startups and established corporations — have embraced virtualized cloud infrastructure in a big, big, big way.

          •' XMIR Grid says:

            Yeah, I was not discussing data center performance and economy, but raw performance on bare metal vs a virtual machine, particularly when you are running a JIT and virtualization in the virtual machine (mono).

            Data centers typically have very fast IO channels connected to their servers, in addition to better redundancy, cooling, power supplies, duplicated network access paths, automated backups and what have you.

            A small to medium size grid can happily manage to produce a fast and reliable service on their own kit if they have the right skill set, but for any larger size operation moving into a datacenter fast becomes a necessity.

          •' Alex Ferraris says:

            Thats my point. And placing more than 1 region in an instance of ooensim is also not good. I have been a customer of most these opensim hosting companies and the results have not been favourable. That is why I know exactly what a grid needs. Like the hair club for men guy always said. Im not only the president but im also a custimer. 🙂

          •' Rene says:

            I need to correct your misunderstanding of the virtualization overhead. The virtualization layer is almost always hardware these days. One example is that Hyper-V virtual CPU overhead is less than 0.1% (one tenth of one percent) and for the most part it is not cause by code running but instead by the various mapping tables that the hardware (CPUs, memory controllers, I/O interrupt routers) access to maintain the separation wall. VMware has similar low overhead This miniscule overhead is often not measurable across the entire span of workloads. Likewise, for dedicated/pass-through devices, the overhead is again less than 0.1% because access to the I/O registers by the virtual machine OS drivers is direct. In modern (and by that I mean since the last 7 years) hypervisors, all the critical and performance sensitive virtualization happens in hardware.
            Overhead using shared resources is higher, around 1% (one percent) for block storage devices and networking. Slightly higher overheads happen when multiplexing virtual processors onto cores. Nonetheless, this overhead is so low as to be overshadowed by the benefits of dynamic resizing and massive cost and infrastructure savings brought on by machine virtualization.
            The question of whether to dedicate physical hardware to one workload has long been answered as it’s not worth it. Virtual world simulation is well suited to run in high performance virtual machines.

          • Rene —

            Yes, the hypervisor is getting more and more efficient, but you still have the issue of having a copy of the operating system in each virtual machine. Containers do away with that, but lose some of the isolation benefits of virtual machines.

            What’s interesting to me is what’s going on with running containers on bare metal, with a thin container hypervisor, and there’s some interesting stuff going on with better isolation of the individual containers. But

          •' Rene says:

            “Yes, the hypervisor is getting more and more efficient…” They’ve gotten to maximum efficiency, Maria. Virtual machine overhead is effectively zero on hardware mediated virtualization ,which exists on all modern VT enabled Intel processors, and in ARM processors with h/w virtualization. That last 0.1% is hardware. 0.1% is well below the noise variance of running code in different temperature silicon, or the difference of cache line misses.
            The issue of multiple copies of the OS taking up RAM space is the same if one were to run those OSes on bare metal. Nonetheless several virtual machine products provide the means to mitigate and reduce the memory footprint.

          • Frank Corsi says:

            I have found that our customers are looking for a “LOW COST” solution with as much performance as they can get for that cost. But most of all our customers are more interested in how they can easily upgrade ram and cpu resources without having to move to another server. Or the ability to have automated control panel functionality over sending hosting company staff emails to have a sim restarted or to add land regions.

      •' Carlos Loff says:

        I will surely compare both hosting services and for me, and I guess for most future Grids clients, the most important will be a customizable pannel that quicly allows us to change and tune things on a more intuitive way and not having to wait for the Grid host to change stuff we request – Sp far there is only one Opensim service that offers full automation and intuitive tunning – Kitely – But they are to tight on land size limits and do not offer Grids service – Since Lisbon Grid wants to be very small in regions performance is always assured on the first hand but tuning and automation is very much required so there is no hurry for me and as I said there can be 2 Lisbons with different goals, let’s see howvthe campaign goes

  2.' Alex Ferraris says:

    AviWorlds is not loosing Lisbon. Carlos has specified that . But I must say I am concerned because hosting such a huge area does need power and lots of it. Specific tuning is something that can always be done.
    Specially now that we are rolling out our avi-labs SPARROW version which includes many fixes to the xengine and adds functionalities that no ohter grid or hosting can offer at the moment. Opensim as is can be a great solution but it is not better than Avi-Labs Sparrow version 1.0.
    Our search and classifieds alone are a powerful aditiion plus our marketplace 2.0 and many other features like CHANGING AVATAR NAMES and more…
    Avi-Labs is working ou our own user control panel which is almost done too.
    I am not trying to create any problems here but with that amount of money charged by Cloud Serve; and dont get me wrong. I never used Cloud Serve but I have spoken to Frank Corsi a few times on the phone and I totally respect there work and service.
    But business is business and as I said that amount of money can buy you CARLOS a 32 gig ram server, with 2 tera disk RAID 1 meaning it is MIRROED , 250 plus upload speeds, 8 core processor and more.
    Virtual Servers are not good for virtual worlds specially for a large scale region with the amount of prims and activities expected to exist on it. It will lag and it will lag alot. Just my opinion. Hey I may be wrong… 🙂

  3.' Carlos Loff says:

    I added a 1$ contribution so everyone can leave me just a tiny help and make it possible, to judge later how a dream becomes truth – Now folks – how hard can it be for 1.890 People around Opensim to give me a chance to show how I can make Opensim life more interesting on an Hypergrid basis ??? How hard is a 1$ risk ???

  4.' Alex Ferraris says:

    When a customers requirers a grid we supply the customer his/her own physical server. Running only his/her grid.
    The customer also has full access to the server and the consoles.
    Customer can make oars, start regions and even place more regions online.
    Full control of the robust, data base and all.
    Plus we can manage it for the customer if he or she wishes.

  5.' Carlos Loff says:

    A NEW GRID IS BORN – ONLY one land, but what a land – Use newest Singularity or Firestorm version 6,9 – LISBON GRID –

  6.' skylifegrid says:

    Well Said! … Customer Relation is Number 1 in any business success.. A 2 year Collage Degree in Hospitality taught me that!

  7.' Alex Ferraris says:

    Avi-Labs/AviWorlds grid offers a VIP customer service. Meaning we go ONE ON ONE with our customers and we are available 24/7 via facebook/skype/InWorld/Phone/Email and offline messages. We are always there for our customers!