Quote from: aplistir on February 10, 2020, 03:15:25 PM

--snip--
Wow. I didn't know the block reward was already THAT low in testnet Undecided

Do you have an address for donating extra coins? I should have a little less than 4 coins somewhere.

Testnet coins being hard to get defeats the whole purpose of testnet. Don't you think it would be a good time to reset the whole testnet (or just start a new one on a side), so that anyone can play with it as much as they want. Or are we "testing" what will happen when block rewards end altogether?

What would be needed to start a new testnet and make it public? Just one machine mining it and keeping it alive? And a faucet for the coins. And maybe a block explorer for it.
Running a new testnet and a faucet would be cheaper than mining the "old" testnet.

I don't thing anyone would have any reason to oppose having a sizeable pre-mine for the faucet, say 1M coins. (since they are worthless)

Yup, the block reward is THAT low Sad

To be honest, i've already taught the same thing a couple of times: maybe it's time to reset the testnet again... It's a TESTnet afterall, people need tBTC for tests, and it's a bummer if they can't just CPU mine some tBTC. I have no idear why people insist on running old ASIC's on the testnet to begin with. The stash of tBTC i currently have was CPU mined a long time ago, but nowadays is simply impossible. That's why i need all those adds/affiliate links/begging for tips/donating cpu power/..., i simply need to either rent some hashrate from miningrigrentals or buy an old asic (an S7 or something) if i want to replenish my faucet, and i don't want to pay these costs out of my own pocket.

As for a donation address for tBTC: sure

This address was generated by my testnet node, the one running the testnet hot wallet, so all donations go directly into funding the faucet: 2NBsursJUfmepDu3WT7GBRLrL6AdHSPC9V4

Quote from: Saint-loup on February 10, 2020, 03:41:33 PM

--snip--
Maybe bitcointalk should launch a campaign to ask people to send back their tBTC to faucets... Undecided

50 000 sats by claim I think it's the same amount you could earn on "mainnet" faucets few years ago Embarrassed

Yeah, i realise 50.000 is a tad low... It should be sufficient for a couple basic tests tough... I dont think you need more tBTC if you're trying out creating a testnet paper wallet and sweeping it in testnet electrum, or setting up a testnet lightning wallet, or testing out some basic json-rpc calls on your testnet node... However, i realise that if you want to test a new mixing algorithm, 50.000 tSat is just dust.

I do plan on increasing the claim amount in the future, after i see if this model is self-sustainable.

A campaign to urge devs to send the tBTC back to one of the working testnet faucets would be great... I do think most people will have lost the funds, or they'll just ignore such questions... But if only a few devs would do this, it would probably make a huge difference.

Quote from: BrewMaster on February 10, 2020, 04:42:24 PM

--snip--
i was thinking the same when i found out about the current reward, but you have to realize that testnet difficulty drops to minimum quite often and miners can potential burst mine lots of blocks to get a lot of reward in total.

additionally there currently are at least half a dozen working testnet faucets that pay from 100k to 1 mil satoshi per claims that are ranging from 15 min to 12 hours. these amounts are more than enough for majority of tests that developers want to perform.
for anything that requires any higher amounts (which are extremely rare cases) they can always run their own testnet using a self contained network (eg. testnet in a box) and run their tests.

Half a dozen seems about right, I found 6 (including mine), that's why i created this one, testnet faucets are important, they give people a no-risk tool to develop and learn.

The testnet difficulty is peculiar. It drops to 1 if no block has been found for 20 minutes. However, if the difficulty retargets after the the diff has been dropped to 1, it'll stay 1 for 2016 blocks, than it'll double for 2016 blocks, then it'll double again,...
However, if you're cpu mining, and you're competing against an ASIC, this just doesn't matter. The asic will find thousands of blocks (difficulty 1) before you've found only 1 (with your CPU). Sure, mining moves faster, but you still need a decent hashrate if you want to mine some tBTC.

The regtest is great, but it's not really a network. It's fine for debugging your json-rpc api calls, but most of the time i used it, it did not suffice for "real world" simulations, so you actually do need the test network for that...