Feature Request: testnet

[k9ert]

It would be cool to play around with the contracts on testnet. So supporting testnet-addresses would be awesome.

[danrobinson]

We do hope to add support for creating testnet transactions, but it's non-trivial to implement (particularly since, unlike the rest of the playground, which is a static web app, it will require a backend server), so I don't have a firm deadline on when it will be available.

[k9ert]

I'm not 100% sure here but isn't that two different things? I'd probably be already happy to have testnet-address-support. With that i could potentially send testcoins to the generated addresses and even release them with the script without any additional integrated help from the ivy-playground.

So ivy should imho simply focus on creating the script and maybe signing transactions. Other tools can then do other things:

• https://www.bitaddress.org?testnet=true
• https://live.blockcypher.com/btc-testnet/pushtx/

Really helpfull probably would probably also to have testnet-support for https://coinb.in/#wallet or maybe other javascript based wallets operating on testnet but i could not find any (yet)

So having a "cross-tool" educational-tutorial would be even more helpful then making ivy automatically integrated with testnet. Would you agree?

[danrobinson]

Technically the SDK and playground already support testnet addresses—the Address (testnet) field generated in the playground is a real testnet address. But it would be pretty hard to spend from it.

You're right that tools like BlockCypher could handle the actual submission of transactions, but there are other steps—such as telling the library the UTXO ID of the contract, so it can properly generate signatures for the unlocking transaction—that make it a somewhat more complex flow.

Can you be more specific about the feature or set of features that you think would be useful?

[k9ert]

Yes, you're right, somehow got that wrong.

but there are other steps—such as telling the library the UTXO ID of the contract,
so it can properly generate signatures for the unlocking transaction—

One can do that manually on the blockchainexplorer. E.g. here is an UTX0 ID via the blockchain-testnet:
https://testnet.blockchain.info/tx/f3adf7b511bf48e4e2fb1a4fd21bc812e0cc93914a901a195db82176e086c51d?format=hex
Why not pasting that ID in a field in the "Transaction-detail-section" when unlocking a contract?

that make it a somewhat more complex flow.

Yes, indeed. But that's the point. Currently you can already choose whether you provide a public-key or a private-key (and the public one gets created). The second choice makes it somehow more complex, doesn't it?

I'm only partly know what i'm talking about but i also love to learn and teach things. I think if the one who's playing with ivy would be able to also interact with testnet, that would be awesome.