InWorldz Blog

Where your Dreams are our Vision!News about the InWorldz Virtual World

Tech Blog

«

Aug
30

The hard thing about hosting things

Over the past 6 years, we at InWorldz have put a lot of blood, sweat, and tears into our little virtual world. We’ve done our best as a small company to provide an unparalleled level of service and quality to you, our customers.

During this time it has become increasingly obvious that the biggest “innovation” to the 3d hosting landscape has been a race to try to out-cheap the next guy by finding the lowest priced servers from the lowest rated providers and hosting as much as possible on them. Eventually when enough people join, services begin to get slow, crash, and fail and the host doesn’t have the knowledge or competency to fix the issues.

3d hosting is not web hosting. It is far more resource intensive, and yet the vast majority of people are still trying to price it cheaper than even the cheapest generic web hosts, and making almost no money doing it. The hardware we have at InWorldz to host 1500 regions could probably be used to host 10,000 websites that get moderate traffic. Hosting websites and hosting living, breathing virtual worlds are two entirely different business models. It is easy to power empty regions on just about anything nowadays, but add scripts or visitors, and the game changes quite a bit. It would be extremely odd to have a website with no visitors using a lot of resources, but regions with no visitors using a lot of resources is a very common occurrence in 3d. Unfortunately the only source of press in our small ecosystem mistakenly covers the cut throat pricing models as if they mean 3d hosting has come into its own, when not even huge companies like amazon can offer CPU cores as cheap as some of the people in the 3d hosting space are.

The problem with the very cheap prices goes beyond just the eventual quality of service degradation and crashes that are always seen if the hosts pick up real traffic. Very simple economics paint an interesting picture.

Even if a grid has 500 regions operating with no server costs, charging $10/mo the grid is pulling in $5000/mo. To put that into perspective, someone working 40 hours per week at the lowest available minimum wage in the U.S. right now earns $1,256/mo, or more than 1/4 of the grid’s total revenue. This doesn’t count the many states that have instituted higher minimums already.

This means, assuming the cost for servers that will be required as the business grows stays at or below 25% of total revenues using the cheapest hosts available, it will take 800 region sales just to pay a single software developer full time below the national average. This doesn’t include the owners of the company getting paid, nor any support personnel. With the small size of the market vs the number of grids that exist out there due to heavy fractionalization, don’t expect to have your particular complex bug fixed any time soon.

Having a “staff” of people that you’re paying in favors or regions doesn’t scale out when hard work needs to be done. People’s time is worth money. Everyone has bills to pay. Developing script engines, physics engines, and scalable data components is complex and expensive work where the people involved make top dollar for their contributions. If you don’t pay them what they’re worth, you will lose key people to other ventures and more lucrative opportunities. In the end, this is and will continue to be the biggest loss to the current breed of 3d virtual worlds.

With paid software developers, we’ve been able to develop and release a performant script engine in under a year, and a highly realistic and performant physics engine in even less time. Both work well, and we have no need to replace either. Those same goals have been on the agenda and in development under other software models for much, much longer. When people get paid enough to live, they can concentrate on just a few objectives and accomplish them efficiently.

This is a constant reality under any development model open or closed. The most modern, successful, and active open source projects have people working on them full time from all walks of life usually employed by companies with a commercial interest in the software, or by investors that want to see the software used in some way for commercial purposes. Without the funds to back that development, the projects would suffer and lose software experts because people need to eat, and food isn’t free.

InWorldz pricing is designed to specifically tackle the realities of an always on hosting platform in a small market that requires a lot of innovation. We keep region prices affordable, around the cost of a family night out, but still at a level where we can pay people to fix the bugs that are important to you, and implement the features you absolutely must have. We want to see a future where scale out components can continue to be developed so that grid crashes and data loss don’t have to be inevitable and mar the the reputation of virtual worlds. We want to see a future where everyone involved can continue to concentrate on working virtual world technologies, and not be forced to move on to something else to pay the rent. Finally, we want to see the pace of software development increase so that virtual worlds can be viable solutions for a VR feature, not decrease into obscurity.

Other business models exist beyond hosting, but remember that whatever actually has paying customers will be the project that gets the most development attention. If it’s not hosting, then the simulation software, the thing you actually interact with the most, becomes a secondary concern.

Site Navigation

search engine optimization