Welcome to the User Testnet | Hacker News Reader

Welcome to the User Testnet

unrealisedgains.substack.com

32 pointswhitxyz46 comments

in the short-term while proof-of-work phases out (limiting transaction fees) and the volume of unmined Bitcoin continues to diminish (hopefully bringing price stability to BTC and the market as a whole)

Neither of these are correct BTW. PoW/PoS has nothing to do with transaction throughput and BTC will never be stable.

Getting burnt by scammers is as good a reason as any to avoid crypto altogether. Even if testnet currencies aren’t material and don’t represent real losses, getting scammed or worse (i.e. having your identity exposed) will disincentivise users in the onboarding process from continuing their dive down the rabbit hole.

If real crypto can't be made secure (see earlier "almost got scammed" article) then it's unrealistic to give people a secure learning environment. Arguably the testnet should be even scammier than the real cryptoverse so that people can either toughen up or give up before real money is involved.

I think thats actually a solid idea for a crypto wallet on-boarding: simulate each scam that's likely to be tried against you, let you fall for it so you can see how the scheme works, thereby teaching you how crypto works, one exploit at a time.

Similarly an onboarding on common user errors would be useful too. An estimated 10x more cryptocurrency is lost to user error than scams. (source in my other post on this thread)

> BTC will never be stable.

Care to elaborate?

When changes in sentiment and macro factors meet a fixed supply there must be boom/bust cycles.

The writing style is hard to follow. Did anyone who read the article find interesting points?

Overall, would not recommend wasting your time with this article. Dunno why it was posted or written. Kind of a stream of consciousness thing, doesn't work well.

If you want good crypto content check out the Zero Knowledge podcast.

I'm surprised the author makes no mention of centralized chains over testnets, since the importance of "skin in the game" as a learning strategy is emphasized.

To learn Ethereum, getting around the Binance Smart Chain or Polygon ecosystems can be done for very little cost, all the while involving tokens with actual market value (BNB and custodied ETH respectively), and mostly up to date EVM implementations.

Importantly, this will also confront users with a multitude of scams. I don't expect the amount and ingenuity of scams to die down, so learning to operate in an adversarial environment should be an integral part of any crypto education[1]. It would be dangerous to land into the real world of crypto with a false sense of security conferred by an otherwise realistic experience in value-less testnets. Technical expertise might deflect some scams, but a lot of them play more into social engineering (sometimes to an extreme degree[2]), which you will only encounter in live environments.

[1]: This applies to cryptography as well.

[2]: https://twitter.com/thomasg_eth/status/1492663192404779013

User error is an order of magnitude more dangerous than scams.

Estimates suggest 4 million bitcoin are lost due to user error [1]. Estimates also suggest $14B per year of cryptocurrency is lost to scams. [2]

At $40k per bitcoin this suggests 350,000 bitcoin are lost each year to scams and theft, ignoring all other cryptocurrencies. This would take 11 years to add up to 4 million lost to user error.

Testnets are good for overcoming user error. Please learn on testnets to your heart's content.

1 https://fortune.com/2017/11/25/lost-bitcoins/

2 https://www.cnbc.com/2022/01/06/crypto-scammers-took-a-recor...

The "lost bitcoins" figure is misleading, since most of these happened during the early days of Bitcoin when their value was low and mining outputs high. Little care was taken about backing up wallets, many have simply been abandoned. The new users of 2021 did not handle thousands of Bitcoins.

A relevant comparison should rely on surveys realized in recent years, where crypto holdings represent significant monetary value for most participants and where tooling has improved to be slightly safer.

As such, I'm not convinced that your conclusion is right. And if we add "lost gambling on cashgrab investments" to the scam category, I'm fairly sure it is wrong. From experience prowling many crypto developer communities, everyone in the field discusses investments at some point, so as a user you cannot avoid building an understanding of those mechanics as well.

Good point let's factor that in. Let's say 4% per hear or 1,500 Bitcoins are lost per day as of 2020 (crappy sources but let's use them as an example) (1)(2). That puts us at 547,000 Bitcoin lost per year compared to 350,000 from scams across all cryptocurrencies.

(1) https://coinquora.com/bitcoin-supply-loses-1500-coins-per-da...

(2) https://static1.squarespace.com/static/5d580747908cdc0001e67...

This piece cites the exact same article you linked in your previous reply as the source used to come up with their number. Same issue.

To be clear the paper claims they developed their own methodology. And that this methodology lines up with the previous source: "This figure is consistent with research conducted by Ratliff in 2014 (n. 5) and Chainalysis in 2017.6"

Whether you want to believe them is another issue, may simply be that they took the source and built a curve around it.

Happy for any other sources or surveys you care to share.

It sounds like you think scams and user error are not equivalent. Let's say user error losses are 150 bitcoin per day instead. That would suggest scams are 10X losses from user error.

Learning to operate in an adversarial environment should be an integral part of any crypto education.

So you need a testnet to debug your attacks?

If you're trying to craft attacks you should execute them on a private network and not the public testnets, as people monitor the testnets and will copy your attacks if they see you exploiting the corresponding testnet contract of a mainnet dApp. I suppose that means the testnets are already a kind of adversarial environment.

Ahhh I was hoping this was referring to crypto as in cryptography. Oh well, does anyone have good low risk ways to learn to properly write crypto code?

I feel obligated to say the lowest risk way is to not write it at all. Don’t roll your own crypto.

Sadly the article is not about Crypto like in RSA but Crypto like in BTC...

“Don’t roll your own crypto” is great sense for production software. It is not good advice for people who are interested in learning to write good crypto, which we will continue to need. For those people: roll as much as you can! Roll your own Feistel network. Roll your own AES implementation, and then roll your own side channel attack to break it. Roll your own timing attack. And then eventually, maybe you’ll be the one to roll the first draft of the built-in implementation of $WELL_KNOWN_PRIMITIVE in $NEW_MAJOR_LANGUAGE. :)

See “I feel obligated to say…”

In other words, I didn’t have any advice so I didn’t try to give any.

Oh. Well, I did have advice, so I gave it. :)

MysteryTwisterC3 is an excellent way to learn RSA, ECC, and so on through the process of cryptanalysis. https://mysterytwister.org/home/welcome/ To learn how to implement stuff correctly, I recommend trying to work through an algorithm (say SHA-1 for example), and try to find people that explain it. Once you're ready to implement it yourself, go through the list of flaws in common implementations (side-channel, power analysis, diff, linear, etc) and try to prove whether your implementation is vulnerable or not vulnerable. Once you've done this more than once, you can do it for a lot of things.

It turns out one of the best books on cryptography is in a bundle if you like to read, Applied Cryptography: https://www.humblebundle.com/books/cybersecurity-2022-wiley-...

Schneier has answered this question also: https://www.schneier.com/crypto-gram/archives/1999/1015.html...

https://cryptoisnotcryptocurrency.com

cryptopals is a good way to learn about practical crypto exploits

Out of curiosity— can you think of any high-risk ways to learn to write crypto code? Do it while driving? Start throwing untested ideas into production in a large database of personal information? Do it for an international drug cartel?

Write your own (encryption algorithm, password hash mechanism, bitcoin wallet) take your pick. Rolling your own crypto is high-risk in itself and the risk goes up the more value in what you're protecting with it.

Using it to protect drug cartel communications seems about as high risk as you can go... I bet there would be some serious repercussions when they got busted because of a failure in your code...

+1 for https://cryptopals.com/

I was curious about how prevalent the use of crypto is as referring to cryptocurrency.

I had to get to the fifth page of Google results for "crypto" to get to this non-cryptocurrency gem: https://www.rottentomatoes.com/m/crypto

You might want to check out: https://github.com/cryptosubtlety/intuitive-advanced-cryptog...

> A Look at Low-Risk Ways to Learn Crypto

1. Don't. There are no non-scam use-cases for blockchain that don't have better non-blockchain solutions.

2. If you must, focus on implementing a toy blockchain in $LANG_OF_CHOICE. At least that way you get programming practice.

Classic HN comment. There are a bunch of legit use-cases, that's why crypto market cap is 2 billions.

https://www.youtube.com/watch?v=4UvkmbugpvU

2 trillions

Yes, my bad. 2 trillion. The value of companies that are doing business in the domain, e.g. Kraken, FTX should also be added to that value.

And some NFT collections have billion dollar marketcaps as well.

>There are no non-scam use-cases for blockchain that don't have better non-blockchain solutions.

This is a classic. I can assure you that cryptocurrency has a use case, although you might not be interested in third world stories. You might be writing from a country where you are allowed to buy any currency you want. Maybe if you want to have euros, dollars, swiss francs, you name it, you can. In a bank account! It must be great living in a place [where people trust banks](https://en.wikipedia.org/wiki/Corralito). You probably are not forced to earn and save money in a currency that loses it's value by 50% every year.

You are giving an opinion (and it's even plain wrong) without realizing how privileged and out of touch you are with the reality of a lot of people.

What about spending a few bucks to buy some and then trying to use it?

You would learn way too fast that it lacks any purpose ! Better spend time and money learning all the way to shuffled tokens online before learning the final lesson: most people dont recognize the value you spent on them.

Most people don’t recognise the value of almost everything you buy. The important thing is that somebody does.

That's how I started learning. I was researching online anonymity and Monero was the 'gateway drug' (yeah, not Doge or even Bitcoin) into the whole cryptocurrency ecosystem. There's a lot to learn just setting up an account on an exchange, and in respect to anonymity, many on-ramp gaps have been closed fairly recently (as a result of the explosion in popularity in early 2021) so those seeking such things are increasingly pushed to less-trustworthy fringe services (no judgement on the right- or wrong-ness of any aspect of this) where you may just be kissing your fiat goodbye - but then those seeking anonymity are probably OK with a certain level of risk.

Literally a few bucks doesn't buy a single transaction fee now. The article mentions that sacrificing gas fees of up to US$100 to buy a $221 image of a bee or having to purchase a minimum amount of governance tokens to properly participate in a DAO of your choosing isn't a good experience for newcomers. If you think fees will never be fixed, fine, but otherwise it's giving people a false impression of crypto.

Depends on what cryptocurrency you want to experiment with. Bitcoin Cash looks to have $0.0007 transaction fees [1]. Seems like $5 would get you some basic experience. Another option is to get some free tokens from a testnet faucet, then everything you play with is free.

[1] https://blockchair.com/bitcoin-cash/charts/median-transactio...

You're not going to be learning much about smart contracts, DeFi, NFTs, etc. on obscure chains though.

Depends on what you want to learn. Rocketpool has a testnet setup so you can play with ETH staking: https://docs.rocketpool.net/guides/testnet/overview.html

Get some free testnet ETH and try it out!

I haven't looked into minting NFT on testnet. You can write your own contracts on testnet tho but that's a heavier lift.

Disclosure: I own crypto and think ETH will be <$0.05 USD for a transfer within 4 years.

Depends, a non technical friend of mine asked me about polygon because their favorite artist was mining NFT's. The coinbase -> metamask -> mint process was less than .25cents

I guess this made it to the front page of HN because the mods are sleeping?

I know what you are thinking. Another cryptocurrency post is on HN once again and it is difficult to ignore it.

Whilst it is there, someone here can’t help themselves and scream the same thing over again including ‘scam’ ‘ponzi’ ‘pyramid’, ‘vaporware’, ‘planet-incinerating’, ‘proof-of-waste’, ‘piece-of-sh*t’, ‘casino slot machine’, etc.

It gets quite predictable and really unsurprising. If for every crypto-related thread on HN, any mention of those words was worth $100, well… that would be a lot of money for me on that prediction then, since they love to hate anything crypto.

So predictable, its profitable.

The tide is noticeably turning in the last month or so though.

Reply on news.ycombinator.com