> To me it's like the limitations of first game consoles (1kB usable RAM for the NES). Having great games work on it needed quite crazy coding skills and several hacks to use less space.

That’s not a fair comparison, in my opinion. A billion NES consoles can operate in parallel, without affecting the performance of other consoles. It scales perfectly. For a shared blockchain, however, performance deteriorates for you when someone else plays. And it doesn’t even have to be the same game. I mean, if a single game like CryptoKitties can reduce performance and increase fees, how will it look when there’s 10 games out there, or 100?

> I always find it odd that such a tech-craving population as HN readers always seem to miss the bigger picture.

I’m not saying this can’t be solved in the future. I’m talking about the present motivation for building blockchain-based games. Why can’t we critique technology just because it has the potential to improve in the future?

You're totally right — it doesn't make sense for every DApp to run on the mainchain.

There's an inherent tradeoff between performance and decentralization. The major blockchains like Ethereum focus on decentralization above all else, because it's extremely important for a cryptocurrency where there's a real financial incentive for someone to attack the network. Performance is important, but shouldn't come at the cost of decentralization. There's a ton of smart people working on scaling Ethereum, and they will scale, but not to the level required to run WoW and Twitter simultaneously on the mainnet.

Which is why we think scalability will be achieved through sidechains. A sidechain can play by its own set of rules. Something like an online game or community on the scale of Hacker News needs to optimize for realtime performance, but can make reasonable tradeoffs for security — on a blockchain data integrity is already ensured by cryptographic signing, and you can't "double-spend" a comment or post, so there's little financial incentive to attack the sidechain to delete a historic comment.

I'd say the motivation for learning to build blockchain-based games now is because the infrastructure for this tech is coming very soon. We're in the process of building out one game as a demo to showcase Loom Network, and are talking to a couple other game studios who want to build games on our platform.

Things like CryptoKitties may just be a novelty right now, but they're paving the way for new types of game economies that that we've never seen before.

I think the main motivation for a developer who believes that ETH will solve the scalability challenges at some point is to master these technologies early on. I have lots of doubts about DApps but I can see why it's good to learn them now.