[gsaines]

Hi Sillysaurus, thanks a lot for the feedback. We've been talking to gamedevs and VCs and so far nobody has raised a flag about opensourcing everything, which is what we're thinking we're going to do. In the words of one VC we spoke to "it's all about the branding and community, who cares who has the code?" It feels risky, but we really think that it's consistent with the intention of the game to make it easy for people to be both players and makers. In a perfect world we'd get so good at educating players they could go through our campaign and contribute to the CodeCombat github repo as their first open source project. That's a ways off in the future, but it's definitely one of our dreams. :)

So here's a question for you: we spoke with a YC alum that has open sourced 90% of his codebase and recommended keeping a small subset of it proprietary. He said that when he was out fundraising it defused a lot of tension when VCs and angels would ask about his open source policy and he could say "well, I keep some of my hardest algorithms in a closed repo." As a game, we have the art, which I agree is a pretty substantial barrier to copying us (many tens of thousands of dollars in our case), but would you still recommend 100% open source vs 90 or 95%?

[sillysaurus2]

we spoke with a YC alum that has open sourced 90% of his codebase and recommended keeping a small subset of it proprietary. He said that when he was out fundraising it defused a lot of tension when VCs and angels would ask about his open source policy and he could say "well, I keep some of my hardest algorithms in a closed repo." As a game, we have the art, which I agree is a pretty substantial barrier to copying us (many tens of thousands of dollars in our case), but would you still recommend 100% open source vs 90 or 95%?

Well, you have users beating down your door. Fundraising is never easy, but you'll have a much easier time with it due to your popularity. And if Sequoia are smart, they will invest in you, because you've discovered an entirely new game design. Historically, whenever someone has invented a game design that turns out to be immensely popular (Pong, Doom, Minecraft, etc), that's almost always resulted in the company becoming a hugely successful game studio. So if you focus on growing your community and on keeping the game fun, then it's quite plausible your studio will be hugely successful too.

Additionally, if you release your code 100% open source, you will get an insane amount of goodwill from the gamedev community. Right now the community is starved for examples of open-source games that actually work. Right now newbie programmers can go find some code snippets for various little tasks, little parts of a game's engine, like how to draw a sprite. But a tangible, cohesive product? Something that newbie programmers can git clone and actually run, and which actually functions? That's almost unheard of. (Quake and Doom engines are open source, but they don't provide any art assets, so newbie programmers can't download the code and do anything useful with it.) And that's a very good position to be in, because it's unheard of in the same way Rails was unheard of before it became popular. Meaning, if you do it, then you'll earn the instant loyalty of thousands of indie gamedevs. The wonderful PR you'll receive will be such a huge benefit to your company that it seems totally unjustified for anyone to be concerned that someone will steal your idea by stealing your code. (For what it's worth, I know of exactly zero times it's happened in practice. Code theft is a very overblown concern, especially in a hits-based industry like game development.)

You'll also enjoy many intangible benefits derived from open sourcing a fully functional game. For example, after Rails became popular, 37signals probably found it much easier to find and hire talent, since they were a sexy company to work for. I would bet you'll wind up in the same position. In fact, just thinking about you open sourcing this game makes me want to go work for you. If newbie gamedevs can download it, run it, learn how it works, tweak it, and become inspired to make thier own games... well, that's what gamedev is all about. It's why kids want to be gamedevs rather than other kinds of devs. So you'll be a huge boon to the gamedev community, which means talented people will want to work for you because of it.

[gsaines]

That's a very useful 2c, we haven't worked in the game dev industry, so it's extremely useful to get the opinions of folks that have that experience.

If we have our druthers, we'll release 100% of it and include the game art under a noncompetitive license so that anyone (and here we're hoping that includes teachers, professors, tutors, etc) can clone and release their own versions with different content. In a perfect world we would also be able to sync the content created like that back to our servers so that 1) we can give it more attention from the larger community and 2) let others benefit from their experience.

One of our primary interests is in releasing a game that CAN be cloned and is functional from the ground up because that's something we would have loved to have around. We didn't know it was rare for such things to exist, only that we thought it would be a cool thing to do.

Again, we really appreciate the advice, we're still learning this game dev stuff. :)

[mbesto]

For what it's worth - I'm in the process of reading the biography of John Carmack and John Romero of id software (interesting read...recommend it!) and apparently Carmack is a pretty big proponent for open source software: http://en.wikipedia.org/wiki/John_D._Carmack#Free_software

[gsaines]

Thanks for the link mbesto, will check that out.

[Cyranix]

Don't forget that in addition to code and art, you have data. Even if someone clones your codebase and invests time in recreating visual assets, they won't be able to draw on the player experiences you've been collecting. With data, you're able to identify trends and test new approaches smartly in order to maintain the edge; your would-be competitors wouldn't have much more to go on than their own intuition.

I won't claim to know what he thinks exactly -- maybe he's the alum you refer to? -- but I wonder if Samuel Clay (Newsblur, YC S12) would agree with that sentiment, given that he has published https://github.com/samuelclay/NewsBlur for all to see. I haven't committed to my own startup idea yet, but my plan for it does involve open-sourcing the code while keeping a custom corpus and its associated NLP models private.

[gsaines]

Hey Cyranix, yeah the db will definitely be an asset, especially when it comes time to test the business model of recruiting. Sam Altman's comment on stage about recruiters being sorts sleazy is definitely true, and so casts some doubt on our business model research, but we really do want to try the viral method before putting up a paywall.

We did in fact speak with Sam at startup school, pretty small world I guess. He not only had good data about what it took to get devs interested but what kind of conversation rates we could expect from interest -> pull request -> valuable input. To be fair, Sam does keep 10% of his codebase proprietary so it's not 100% out there.

[conesus]

I know I'm quite late to the party, but I actually keep 100% of the source code open. The only parts I keep in a secret repo are tied to keys, hostnames and addresses, and certificates. But, just having a secret repo where sensitive information (including proprietary algorithms) was enough to satisfy the few investors I talked to.

Mind you, my next project is closed source, as I plan to open-source reusable pieces that will actually help people instead of a massive codebase that few have the means to run.