Hacker News new | past | comments | ask | show | jobs | submit login
$99 Mineserver: The Devil Is in the Details (cringely.com)
150 points by evo_9 on March 18, 2016 | hide | past | favorite | 91 comments



> First we needed a mobile app to administer the server so we hired an experienced mobile developer through guru.com. His credentials were great but maybe it should have been a tip-off when, right after we made the first payment, the developer moved from Europe to India... We were naive. The original development estimate was exceeded in the first week and we were up to more than 8X by the time we pulled the plug. Still the developer kept trying to charge us, eventually sending the project to arbitration, which we won.

This smacks of the client not knowing how to use a developer. Meaning: I can imagine they made a simple (probably incomplete) scope or product definition document, didn't realize where the complexity actually lay, found someone cheap, and then realized that no one had the ability to actually shepherd the project to completion. They knew nothing about their developer -- and she or he knew nothing about them. And they weren't paying that much, really. So it wasn't a huge priority in the developer's life. And maybe the dev was young or just kind of between things and something more important came up or whatever. It actually takes some thoughtfulness to be a professional freelancer. Can be harder sometimes than it looks.

Cheap overseas developer talent is fine. But. There's also a huge value to finding people who you can sit down with and talk to. Meet. Get to know a bit. Professional freelancers will absolutely help naive clients identify and avoid pitfalls from the start. We (professional freelancers) have already fallen into them all and generally have the scars and callouses to know what it really takes to produce good software in a freelance environment. Professionals know how much time and effort it actually takes to do things. They'll tell you if your "I need a Facebook clone next week" project has any chance of success. They'll help you with design and they'll help you understand the trade-offs between money, time, and functionality that all software projects must make.

Sounds like none of that happened here.

Anyway. This just touched a nerve for me. Maybe I'm totally off base.


I didn't read much past that first paragraph for basically the same reason. This project failed because nobody involved had a single clue as to how the real world works.

The worst part is that there are plenty of Minecraft server administration suites out on the web for download already. Hiring a rando "developer" (with of course a good resume, everybody in those code farms has 1,000 years of development experience in every language!) you've never met to do the one thing that might differentiate your product is just insane.

The craziest thing is that a Minecraft server manager is not a difficult thing to write. I tossed together a couple hundred line Perl script with a bunch of useful features in an afternoon and ran servers with it for years.


One would expect differently from Cringley of all person. He made a living about mocking people outsourcing and company being mismanaged.


Maybe he wanted his kids to learn the lesson first hand?


at the expense of real people paying real money?


> The craziest thing is that a Minecraft server manager is not a difficult thing to write. I tossed together a couple hundred line Perl script with a bunch of useful features in an afternoon and ran servers with it for years.

Just checking, but how easy would that have been for the intended audience? (Mums and dads who don't know much about computing or minecraft servers).


Why is that even a thing? Maybe people who don't know much about computing shouldn't be selling computing products.


From The Article:

> But $7500 in the hole is not much cost to start a technology business.

Unless I missed something, it doesn't sound like they paid the developer much at all.


I think $7500 is their net loss. They raised $35,452 on Kickstarter.

https://www.kickstarter.com/projects/583591444/mineservertm-...


They raised $35.5k on Kickstarter, and I'm assuming they have spent all of that plus $7.5k out of their own pocket before shipping a single unit.


7500 / 15 = 500. 500 units times $99 is $50K. With that kind of budget (that's everything, not just dev time) you can't hire a top-line developer.


7500 can get some good (although limited) development done around here (Portugal). Dunno how much time it would take whatever they needed though


Between the port forwarding and security and other kinds of hassles of running a server in your own house, I just used digitalocean hosting for a minecraft server for my kid a few years ago, $5 a month. Here's a script to install the Cuberite version mentioned in the article: https://github.com/cuberite/cuberite-ocean

Now there are free services like Gamocosm which shut down your digitalocean server when not using it, so that it might only cost $1 or so a month: https://gamocosm.com/ Meaning it would take 8 years before it cost as much as a Mineserver box.


I just used digitalocean hosting for a minecraft server for my kid a few years ago, $5 a month.

What are your ping times like?


Also curious, as I ran a $10 droplet to host a Terraria server and the ping-times were >200ms from Vancouver, BC to San Fran. Was basically unplayable, with usage spiking out. Now, it was simultaneously hosting Mumble and an (effectively blank) node site, but still.

For what it's worth, I also hosted a Starbound server, and so long as people weren't doing anything crazy and the population never exceeded ~4-5, things were fine.


Home internet should crush those ping times right, because thats how ISP optimize.


Great timing, thank you! I was just in the midst of preparing to set up a Minecraft server for my son and his buddies.


$5 plan on DO for Minecraft? That Cuberite software better be awesome, because I ended up buying an 8 core 16GB RAM HP rack server just to run a heavy-modded Forge server.


> heavy-modded Forge server

That's probably the problem. I've run a vanilla Minecraft server on a DO droplet just fine for a handful of people.


I've never heard of Gamocosm before but it looks very interesting...


The main thing that surprises me is how someone can be a tech journalist for years and then try launching a product using someone else's trademarks, and not expect to get a cease-and-desist.


Cringley is not a single journalist, it's a persona, kind of like the Dread Pirate Roberts.

https://en.wikipedia.org/wiki/Robert_X._Cringely


Yes, but Mark Stephens (this Cringely) is arguably the Cringely, in that he was the one who made the column popular, and he published a book (Accidental Empires: http://www.amazon.com/Accidental-Empires-Silicon-Millions-Co...) under the name. So there have been other Cringelys, but when people say "Robert X. Cringely," they almost always mean Mark Stephens.


What trademark? The word 'mine'?


Their tagline:

> Mineserver™ -- A $99 Home Minecraft Server by Mineserver LLC

They also used the most iconic symbol from from the game (the creeper face) as a logo. Yes, I'd say the risk that their trademark would be confused with the official product is quite obvious.


>> Mineserver™ -- A $99 Home Minecraft Server by Mineserver LLC

That's just a factual description.

>the creeper face

I don't know, the pixelated smile is a pretty clear riff off the iconic creeper frown, not a copy.


I'm disappointed that a respected technology journalist can misapply the term open source so completely. If he actually meant that java can be decompiled and therefor is 'open source' then why the hell did he capitalize the phrase?


Cringely is more unintentionally entertaining than respected.


> Minecraft, which is written in Java, is nominally Open Source

Minecraft is not open source by OSI's definition and doesn't even have the source code available. The Minecraft server is distributed as a compiled JAR file. What does the author talk about?


Probably that most mods are made by decompiling the minecraft server, modifying it and recompiling, or patching the server with the diffs resulting from that. So yes, people are looking at the source, modifying it and distribute their changes, and Mojang has rules to allow that. But it isn't "open source" in any real way.


It's a terrible way of indicating "you can hack on this".


they should have just used a true open-source minecraft: https://opensource.com/life/15/9/open-source-alternatives-mi...


"Cuberite is licensed under the Apache License V2"


Would someone like to explain the downvotes? Claiming that the use of Cuberite is not "true" open source is like someone in 2011 claiming that a webserver with SPDY can't be "true" open source because Chrome is proprietary.


I assume because Cuberite is only a server component, the examples in the linked article are all open-source minecraft-style standalone games or clients.

Most customers probably want Minecraft, with the mods for the original and such, and not something "minecraft-style", so I don't think that would have been a good option. Maybe a nice extra, but this project really doesn't look like it needs "nice extras" that add complexity.


"Nominally?" Pretty sure no.

"Minecraft, which is written in Java, is nominally Open Source, but there are some peculiar restrictions on distributing the code."


Mincraft is not open source. There's way too much entitlement in this post. "How dare you make it hard for me to not profit off of your success."

It's very clear in the terms of service that you cannot make money from Minecraft mods and that only Mojang is allowed to make money from the game. It has always been this way. It's been in the press many times and anyone wanting to make money from Minecraft should know this. Making money from the success of someone else's project should always involve a bit of research.

The ways that people have made money from Minecraft is by providing hosting and by streaming themselves playing Minecraft and making let's play videos. Another way is to host your Minecraft mod on adfly and make ad money from the downloads page. That's pretty much it.

They should have known this beforehand. I knew it and I don't make mods for Minecraft.


All they're really doing is providing another hosting solution. It happens to be personal hosting. So sure there are differences between this and a hosting services, but I don't really see a problem.


What's in the article sounds like way more than that and it's really the whole tone of the article that bugs me. Yes Mojang doesn't have to let you use its IP to help you market your thing. Yes Mojang can dictate the terms of their product, and no you don't get special treatment because you know people at Microsoft or have a widely read blog. Like really? The article is dripping with petulant self-entitlement. It's disgusting and I say that as someone who usually likes the author.


Minecraft certified hardware sounds like a win-win to me. Making that offer is not petulant self-entitlement. You're being ridiculous.


I think from the context he meant to say "Minecraft Server is nominally Open Source."


Which also isn't true. The server is under the same EULA as the client. It only is "open source" in so far that Mojang tolerates people decompiling it to make mods, and allows the resulting mods to be distributed so that players can integrate them with their games. Which is more forgiving than other vendors, but not "open source".


While I am not interested in the specific project, If/when I have kids, I would gladly pay 7500 dollars to have them excited and apply themselves to a project with real world business.

To have the realizations the struggles and successes of such a project, and have that under their belt before they graduate high school is a wonderful achievement.

What a great project!!


There is also ~32000 dollars of other people's money at play. I wouldn't write this project off as failed yet, but it sounds like an attitude of "it's a fun project for the kids and we don't really know what we are doing" that doesn't seem to appreciate this fact. (EDIT: on the other hand, he is backing it up with extra money, instead of trying to fail out of it cheaply somehow)

The comments on the article are interesting to read, with people arguing both sides.


Agreed, I was commenting on the losing 15 dollars per unit angle, but you are right that there is a burden of responsibility when using other's money. Another part of the great lessons here.


But it's not his money, right? Because it sounds like this is Kickstarter-funded...

"Fail with other people's money, kids." I suppose that is a powerful lesson...


They're shipping out units that cost more than what the kickstarter brought in. Presumably it's coming out of Cringley's pocket.


Good point.

Snarky comment rescinded!


It's great experience for the children.

I do wonder if he would have had that bit of "child labour" investigated by regulators if the project had progressed any further.


"had progressed"? It's still running, isn't it?


I have to question the phrase in the second to last paragraph, in the context of servers disappearing from the net:

"It’s the final bug, . . ."

Uhhhh.....seems optimistic at best especially considering their experience so far.


While still poor writing, in the paragraph a couple up where they start talking about it they say:

"What we hope is our final technical problem "


I think it's interesting that he continually says "we" throughout. As if his kids are really contacting developers and talking with suppliers and mojang lawyers etc. I'm sure they're doing something meaningful like installing the software on the boards and assembling them, but it's pretty clear who is doing the real work


A few years ago this desire to profit off of minecraft washed over a small part of the NYC VC scene with a few higher profile firms investigating some people who said they wanted to "solve" problems with minecraft. Anecdotally I found that everyone I talked to was doing so because their kid was into it and that they all looked at me cockeyed when I said "you know that doesn't sound like something a for-profit company would be ok with."


yeah, sounds like a typical month at a startup. don't worry, next month will be even worse, and you'll probably have even less money.

what's amazing is how few journalists actually have gone through this process, and how even fewer write about it.

the fact that he wrote a huge blog post about it means he's somehow surprised it's almost always this hard to do actual work on products and services -- which is why success is so vaunted and worshipped, not because it's fun and glamorous, but because it's basically getting just the shit kicked out of you 24/7.


"His e-mail address is right on the case and if something doesn’t work he can ssh, telnet, or VNC into the box to fix it."

Hopefully the telnet mention is just a stylistic phrase, rather than really meaning they have a telnet daemon enabled by default.


Based on the rest of the post, I think that hope might be unrealistic.


This might be Cringley's first time on the other side of a product launch, which makes it interesting to watch him eat his share of humble pie after countless articles opining on someone else's delay in shipping a product.


"[A] new fact has now become painfully clear to me: you don't say you have a working product unless you really have a working product"[1]

[1] https://en.wikipedia.org/wiki/Robert_X._Cringely#Controversy


He has a working product with this blog. It's so entertaining he should get into doing a satirical Silicon Valley rag.


The fact that these kids have been able to develop and distribute a product, even if they are not making money, is awesome. The experience and skills they gain with this project will be incredibly valuable for every future projects and jobs. It would be great if schools could offer this kind of hands-on experience with real world applications.


I think that you should have a call for programmers right in your kickstarter. Not saying it would have worked in this case necessarily, but having someone actively interested in the project is a great asset.


Another interesting project post-mortem from Robert Cringely was his ambitious attempt to design, build and fly an airplane in 30 days as part of a PBS documentary Plane Crazy.

http://www.pbs.org/cringely/pulpit/1998/pulpit_19980724_0005...


This is the first time I've heard of Cuberite. I love the concept, but this has to be a copyright violation, right?


No. It is independent work.


Cuberite is to Minecraft what Apache Harmony is to Java or ReactOS is to Windows.


Copyright of what, specifically?


The protocol that Minecraft developed for their client/server?


A network protocol doesn't really have copyright protection, and it seems unlikely Mojang has anything special in there patented. As long as it is cleanly reverse engineered there is not much legal attack surface.


There are lot of services out there that do free hosting for minecraft nowadays, for example: https://server.pro It also gives you a control panel to control your server, supports plugins and has a handy web file manager.


I'm hearing a lot of gold plating that shouldn't have been in 1.0. Wifi? Mobile app? A slick server that is incompatible with their admin panel? Dynamic DNS? Skip those for Mineserver 2.


Aren't wifi and a mobile app essential for their target market? The impression I get from this article, and from going and taking a look at the Kickstarter campaign, is that a big part of the audience is supposed to be people who don't know very much about computers and networking. A lot of these people have a home network that is entirely wifi, running off the wifi router built into their ISP provided cable modem.


If I added it up right, more than half of the backers selected a version without wifi. And are there really devices around that only provide WLAN and don't have Ethernet ports?


The mobile app could easily just discover the local server and then throw a webview of the admin website.


personally i understand that father is involved and wants to help. But a valuable lesson would be having the kids write the investor update, learn to face the failure.

I get that they are only 11, but if your going to have them launch a business they should learn the ups and the downs. Otherwise just post your prototype on Hack-a-day and have it be a fun family project.


I thought you could distribute OpenJDK freely.


I think he was talking about the Minecraft server software not being able to be distributed pre-compiled. Not the runtime environment.

I could have misread it though.


I'm pretty sure what they're referring to is a quirk of how Minecraft mods work.

The official Mojang Minecraft server code is closed-source, and is distributed as a .jar file. It does not natively support mods. Mods work by decompiling the Minecraft server .jar, patching it, and recompiling it.

The results of this process are non-redistributable. That's probably what the article is (confusedly) referring to.


The way most people get around this is by compiling a setup package that downloads the binaries.


The reason I ask is because there has been some stuff that has come to light over the distribution of the JVM although to my understanding the OpenJDK platform does not hold the same restrictions.


Sounds about right, kickstart an idea you had, then try and pay a bunch of consultants and hired coders to make it a reality.


Guess he should solder his own mobo, and pull himself off the ground from his bootstraps, like real startups do.

Or, like most businesses, contract out specialty work and put pieces together into something new.


Looks like they ran into some issues. But an amazing experience for the young boys. I'm not sure many would have that sort of life experience before their first job, let alone before college. Well done to the parents for helping their children to do this.


Minecraft codebase is rewritten in C++. New additions of minecraft will be C++


This is 100% false.

Minecraft has a 'Windows edition' that is written in C++, and a main version that is in Java. Both version will continue to exist, and be updated, AIUI.


I can't believe that they'll maintain both in perpetuity. The Windows edition will probably become the official version the moment that it's feature complete, and the Java version will become the "legacy version" that's never updated again.


Not in perpetuity, but until it is there the Java version will be further developed, and there has been no indication at all when this might be the case. I'd expect the console versions to be integrated before they touch the main one.


So, by "100% false" you mean, yes it was re-written in C++ and hopefully no they will not cut off future Java development, but who can say for sure?


For all the people with negative things to say, tell me, does your business ideas run without a hitch?


You're right. No one who has ever done a wrong thing is ever allowed to have a negative opinion about what someone else is up to. ;-)


What a shocking surprise Cringely has no earthly idea how tech companies actually get things done. He's becoming more irrelevent every day, thankfully.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: