Having fled from the impending demise of NetherByte… which was still up and running the last I checked… and its “$22.50 for six months” pricing to find refuge at MCPro Hosting, which has a better reputation, but charges about that much a month if you add on the ability to do server backups, and for less RAM, the whole “buy or rent” question has surfaced in my head again.
At what point is it worth just buying some hardware and hosting the server myself? Visions of Intel NUC boxes float through my head, but the cost even at that end puts the return on the investment a bit too far out in the future. If I could just put together something that would handle our group, wasn’t a complete pain in the ass to manage, and had a ROI point of about 12 months, I would be very interested.
On to this fertile mental pasture… and remember, fertilizer is traditionally most shit… lands a post about the Mineserver Kickstarter campaign.
Mineserver, according to the campaign, is a hardware and software package that gives you a headless server that you can plug into your network, administer through a web interface, can be made accessible/discoverable outside your network (so your friends can play), and even has an Android/iOS admin app that allows parents to control access from their ever present phones and tablets.
For this, the three primaries in this operation Channing, Cole, and Fallon (ages 13, 11, and 9 if I have the names in the right order) want only $99 for a Mineserver capable of hosting 20 player, or $199 for a Mineserver Pro, which is billed as being able to host 50 players and still keep its cool. Less if you order early.
Pull the other one, right?
The tale is more plausible when you bring their father into the picture, Mark Stephens, more commonly known as Robert X. Cringely. A long time staple of Silicon Valley, his column in InfoWorld was a must-read though his primary claim to fame is his book Accidental Empires, a history of Silicon Valley and the early tech industry, very much a must read in my cranky old opinion (along with Rick Chapman’s In Search of Stupidity, which fills in some of the missing lore), which was turned into the PBS documentary Triumph of the Nerds. His blog, I, Cringely, is a regular read of mine and is linked somewhere down in my blogroll.
Anyway, Cringely and his tech connections and knowledge and backing of the whole venture makes everything more plausible. The kids have clearly had access to the right sources and mentoring from the right people in order to put this sort of project together. This gives the project credibility.
Still, I look at it and I have a few doubts. In this sort of venture it seems to me a good plan to emphasize your strengths and obscure your weaknesses.
The strengths they are running with are cost, ease of administration via their custom software, security and safety for your kids, and server speed.
However, on speed, they are focused almost entirely network speed because the Mineserver will be plugged into your local router. (Though there is a WiFi option for people who want the box to sit somewhere else.) That is a speed boost for people in your house, maybe not so much for anybody remote.
Things they have not brought into the picture include any details about the admin software, the discoverability aspect, the Linux distro, the Minecraft server version, the long term viability when it comes to updates and support for ongoing Minecraft development, and most important to me, any hardware specs whatsoever.
The last to me is doubly vexing. First, as I have learned fairly quickly that, at least for hosting services, saying a config will support X players is often hopelessly optimistic. I refer back to MCPro Hosting where, during their setup I told them I wanted to be able to host 20 players for vanilla Minecraft and they immediately recommended a 30 player option where we are constantly at edge of processor and RAM usage with four players in-game. So when they say a Mineserver can accommodate 20 players, whose measure are they using?
Second, hardware isn’t something this project should be competing on, yet when asked point blank about specs, Cringley has declined to answer because he says he doesn’t want to project to be reverse engineered. (Comment on his blog post.) But the secret sauce on this burger is the software, the stuff that they clearly see as the strong part of their pitch. Hardware is a commodity and ought to warrant two lines at the bottom of the page with basic specs simple to prove that the platform has the moxie to do what they say it does. Doubly so because whenever I show the Kickstarter to anybody in tech, the first question they ask when they see the hardware is, “Oh, is that run on a Raspberry Pi?”
Screen grab from the project video
I hope it isn’t a Raspberry Pi, or if it is, that they have been able to really optimize their software as I am not sure that would run anything beyond 10 players very well. Also, Raspberry Pi as a server has been tried and talked about before.
Still, the doubts I express might just be mine. As somebody who works in enterprise software and frets about such details professionally, I tend to have a skewed outlook. For somebody who wants a home server this may very well be an ideal solution.
The project itself looks like a slam dunk to fund. They opted for just a three week campaign and here, a couple days in they are just inches from their funding goal of $15,000. (The joy of having a father people listen to, something my daughter will never experience.) That will get them cases to kick off production, as everything else is reported to be done, so that they can start shipping out units before Christmas. That would have to be some sort of short turn-around record for a Kickstarter project more complicated than potato salad.
It looks cool, sounds cool, and I want to believe, all the more so because of the enthusiasm of the kids in their project video.
What do you think? Worth a go or not? Certainly something I will keep my eye on.
Mineserver Kickstarter page
I also wonder what the guy who did the Mineserver software distro thinks about the project. So few good names to choose from.
Addendum: The project passed its goal somewhere between when I wrote this and when it posted, so congratulations to the team. Now where will thing go with stretch goals and such? I hope they stay focused where ever they head.