Oddbean new post about | logout
 I think that’s a terrible idea, especially since I might also want to try older software on testnet. You’re making me rethink my position, I’m thinking we should revive testnet3 with a simple hard fork 
 if some people stay behind on the old one that's also fine, doesn't change anything 
 But isn’t that obviously a bad thing if we fragment into dozens of testnets? As a testnet user I value interacting on it with others, when playing with myself I’m already using regtest 
 Is it? I’m not really aware of much use of testnet to “interact with others”. Testnet lightning has been totally useless for years, and I’m not really sure what else there is? 
 I’m using testnet3 a lot for Lightning Network and Taproot Asset testing. I run an LNBits testnet node too and as of today, I’m fully on testnet3. My node has trouble keeping up but testnet4 won’t be as useful for me for a while. 
 That doesn’t mean you’re doing it with a broader community, though? There’s basically zero liquidity in testnet3 lightning (or where there is it’s entirely one-sided). Every time I’ve tried to use it I’ve had to open a direct channel and kinda the only destination is yalls anyway? 
 All my testing is done with others, both onchain and offchain, inside and outside LL, mostly asynchronously. I can also tell from my LNBits instance that others are using my node to send sats to and from their wallets/nodes. It’s true that testnet3 liquidity is bad, and since testnet3 has been attacked many nodes fell off the network. 
 Cool! I’m admittedly pretty surprised by that, but…cool! Most of the joint-testing I’ve seen done is on various signets (mutinynet is pretty common, signet og too, etc), and having multiple testnets that different groups use seems fine? The whole point of testnets is that they don’t have value, though… 
 LL will have to make the decision where to move testing infrastructure to, but from this conversation it doesn’t sound like testnet3/4 have a future or is run with expectations that are compatible with what we need in the long run. Having a network that allows for connectivity with people testing other implementations and services (LSPs, swap providers, payment processors) on the same network is important though. Paying a tiny amount for testing is fine though from my perspective. Test/signet coins are API tokens, aren’t they? 
 If you’re running a bunch of testing infra and mostly working with partners, I’d absolutely recommend your own signet! 
 And then convince competitors to join that same signet to test their infrastructure too? Why would they? Yet testing the compatibility of these infrastructures seem to be the primary benefit of a public testnet 
 Wait I understood from your above comment that you were looking for a testnet on which you can test your software with your partners. Where are you currently testing software with “competitors”?

In general it seems testnet fragmented a while ago. All the testing I’ve seen done for a year or so has all been on mutinynet (or more often mainnet lol) 🤷‍♂️. 
 Being able to perform swaps with other providers on existing infrastructure is a great benefit, no? I can’t run all swap services on my local network and also one day hope to be able to interact with RGB channels. Having to spin up a new node, acquire coins and open channels for each system i might interact with feels cumbersome and wasteful. I desire a “common” testnet Lightning Network 
 I agree that would be great! But sadly the last decade or so of Bitcoin testnet has demonstrated that people mostly aren’t willing to put on the effort required to do so (which is very substantial!).