3 ms·

The Bitcoin Core team does have a roadmap: https://bitcoin.org/en/bitcoin-core/capacity-increases-faq https://bitcoin.org/en/bitcoin-core/capacity-increases-faq

HappyTypist 11y ago
But it's arguable if it's a good one. It's main proposal, "soft fork segwit" isn't really a soft fork. It reduces the security of every single existing node to "Accept every transaction miner says, don't do valifation". That's because existing nodes don't know SegWit, and would by default reject these transactions as designed. So the core devs are proposing to reuse "Anyone can spend" op codes to "SegWit", just so old nodes won't complain. That means a malicious miner can spend Bitcoin they don't own and nodes will happily accept it. Why do they want soft fork SegWit? Because if they hard fork, they might as well as simply increase the block size. And some "core developers", with a strong correlation to those employed by a company named Blockstream Inc, absolutely oppose it. What does Blockstream do? They're making a "layer 3" transaction network called "Lightning" that bypasses the Bitcoin network and charges it's own set of fees.
- TD-Linux 11y ago
  Well, I was mostly arguing the point that there wasn't strong leadership or roadmap. Disagreeing with the leadership doesn't mean it isn't strong :) FWIW what you describe is pretty much the definition of a soft fork though - disallowing previously allowed transactions. Old nodes can't fully validate the new transaction types, but they can't spend them either, so they aren't really at risk. Also, the discussion was about Bitcoin Core, not Blockstream, and the Lightning Network is not on the Core roadmap page. Though from reading the lightning network whitepaper at https://lightning.network/ https://lightning.network/ I don't see anything inherently wrong with it.
  - HappyTypist 11y ago
    That's a misnomer. Old clients cannot validate the new transaction types at all and will blindly think they're valid. This is the equivalent of your browser accepting any single SSL certificate that says "X.999" irregaddless of if it's signed. Instead of your browser saying "I don't know how to validate this certificate, please upgrade me so I can". -- Bitcoin core has for better or worse grown synonymous with Blockstream. A majority of commit access holders are employed by them, and unsurprisingly they tend to all vote in a bloc. The roadmap does mention the lightning network, just as a reference to "off chain solutions". The two main issues with lightning are: 1) centralisation pressure among hubs: The number of jobs will eventually converge to a small number of central nodes. 2) Lightning is being developed by a single company in essentially-private, in contradiction to how Bitcoin development has been done since its inception. They refuse to accept outside contributions and only pay lip service.
- ikeboy 11y ago
  Are you claiming that "Wallets are strongly encouraged to upgrade but can continue to operate without modification as the deployment does not break backwards compatibility." is simply false? edit: I just read the discussion https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-December/012026.html https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015... about this, anyone waiting for a few confirmations is not really vulnerable.
- makomk 11y ago
  Errm, that's exactly the definition of a soft fork: old nodes can't fully validate the new transactions and can potentially accept transactions that aren't valid. The P2SH soft fork a few years back worked in exactly the same way, reusing existing opcodes in a way which meant non-upgraded nodes would allow anyone to spend those transactions. If that isn't a soft fork then nothing is.

HappyTypist 11y ago

But it's arguable if it's a good one. It's main proposal, "soft fork segwit" isn't really a soft fork. It reduces the security of every single existing node to "Accept every transaction miner says, don't do valifation". That's because existing nodes don't know SegWit, and would by default reject these transactions as designed. So the core devs are proposing to reuse "Anyone can spend" op codes to "SegWit", just so old nodes won't complain. That means a malicious miner can spend Bitcoin they don't own and nodes will happily accept it. Why do they want soft fork SegWit? Because if they hard fork, they might as well as simply increase the block size. And some "core developers", with a strong correlation to those employed by a company named Blockstream Inc, absolutely oppose it. What does Blockstream do? They're making a "layer 3" transaction network called "Lightning" that bypasses the Bitcoin network and charges it's own set of fees.

TD-Linux 11y ago
Well, I was mostly arguing the point that there wasn't strong leadership or roadmap. Disagreeing with the leadership doesn't mean it isn't strong :) FWIW what you describe is pretty much the definition of a soft fork though - disallowing previously allowed transactions. Old nodes can't fully validate the new transaction types, but they can't spend them either, so they aren't really at risk. Also, the discussion was about Bitcoin Core, not Blockstream, and the Lightning Network is not on the Core roadmap page. Though from reading the lightning network whitepaper at https://lightning.network/ https://lightning.network/ I don't see anything inherently wrong with it.
- HappyTypist 11y ago
  That's a misnomer. Old clients cannot validate the new transaction types at all and will blindly think they're valid. This is the equivalent of your browser accepting any single SSL certificate that says "X.999" irregaddless of if it's signed. Instead of your browser saying "I don't know how to validate this certificate, please upgrade me so I can". -- Bitcoin core has for better or worse grown synonymous with Blockstream. A majority of commit access holders are employed by them, and unsurprisingly they tend to all vote in a bloc. The roadmap does mention the lightning network, just as a reference to "off chain solutions". The two main issues with lightning are: 1) centralisation pressure among hubs: The number of jobs will eventually converge to a small number of central nodes. 2) Lightning is being developed by a single company in essentially-private, in contradiction to how Bitcoin development has been done since its inception. They refuse to accept outside contributions and only pay lip service.
ikeboy 11y ago
Are you claiming that "Wallets are strongly encouraged to upgrade but can continue to operate without modification as the deployment does not break backwards compatibility." is simply false? edit: I just read the discussion https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-December/012026.html https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015... about this, anyone waiting for a few confirmations is not really vulnerable.
makomk 11y ago
Errm, that's exactly the definition of a soft fork: old nodes can't fully validate the new transactions and can potentially accept transactions that aren't valid. The P2SH soft fork a few years back worked in exactly the same way, reusing existing opcodes in a way which meant non-upgraded nodes would allow anyone to spend those transactions. If that isn't a soft fork then nothing is.