5 ms·
You skipped the option that's closest to consensus! SegWit2x has industry support and has a short enough window to lock-in before the Aug 1st UASF EDIT: Quick
by mhluongo 9y ago
You skipped the option that's closest to consensus! SegWit2x has industry support and has a short enough window to lock-in before the Aug 1st UASF
EDIT: Quick reference https://medium.com/@wintercooled/the-road-to-segwit-activation-uasf-segwit2x-and-segwit-signalling-explained-2ff00488b7cb https://medium.com/@wintercooled/the-road-to-segwit-activati...
- jpatokal 9y agoUnlike the UASF timebomb, SegWit2x won't activate unless there's 80%+ support, so it's not going to trigger a fork. (Well, technically it will, but if activated the non-SegWit2x rump will quickly be reduced to irrelevance.)
- wyager 9y agoThere's currently 86.5% support.
- dekz 9y agoThat's 86.5% miners signalling to signal support. Segwit2x is still yet to have a working and tested implementation. Would you support something which has not been fundamentally tested?
- awestroke 9y ago> Segwit2x is still yet to have a working ... https://github.com/btc1/bitcoin https://github.com/btc1/bitcoin > ... and tested implementation. It will go online on the testnet tomorrow
- spiralx 9y ago> It will go online on the testnet tomorrow It seems rather premature to call it working before it gets properly tested. And three weeks is a woefully short time to test something like this, fix any issues that are found, and then re-test.
- ilikechilis 9y agoThe testnet is already online since a few weeks now. But according to the segwit2x timeline this step should come tomorrow if they are still on track: July 14 - Agreement Participants Install and Test Milestone
- Taek 9y agoThough it has industry support, it does not have user support, because it does not have a working, tested implementation released. With less than 3 weeks until the fork triggers, I highly doubt they are going to get substantial user adoption by the time the fork triggers. One of the most amazing properties of Bitcoin is that it's completely user driven. When it comes to the fundamental consensus rules, no amount of hashrate or corporate agreements can force a change onto the userbase. It's a system that intentionally resists central control. Any changes need to start with userbase support, and I'm just not seeing that with Segwit2x. If I am wrong, I will find out soon enough.
- anopows 9y agoHow do you define user support? All the definitions I heard are able to be manipulated easily: -The/r/bitcoin consensus? Is mainly in existence because of censorship -User wallets online? Can be easily created with AWS instances -Bitcoin core? Who decides these bunch of people are in charge to represent all users and not a bunch of other developers? One obvious alternative is to let Bitcoin follow economic incentives: Bitcoin miners gain Bitcoin, so they have a vested interest that their income is and stays valuable. Thus why not let the miners decide. This is also the only place where you can't influence/manipulate Bitcoin easily.
- Taek 9y agoUser support is really easily measured in the event of a split and really hard to measure before the split. Quite simply, the coin price indicates support. After the split, there will be two versions of the Bitcoin network with independent prices. A non zero price indicates non-zero support, and as long as it's at least 25% of the original price, it'll probably be okay. A high price indicates high support. In the long term, the hashrate follows the price. Miners will mine the coin with the highest price, because they can't afford any other option. Electricity is expensive.
- anopows 9y agoSo you say mining of Bitcoin only follows the market value? Isn't it at least to a large degree the other way round? That market value is following based on what is the most mined chain? If there's a split and 80% of mining power switches to one branch. Then the 20% branch loses a lot of utility for at least some time. It now has five times slower transactions until difficulty gets adjusted(or you change the proof of work of bitcoin, which itself can be precarious). And now to a minor degree the minority chain has also less safety (less hashing power). Meanwhile why should anyone buy or send Bitcoins at the point of a split? There's a real chance that one branch will die and then you can lose the coins. The best strategy is probably to wait. So the market at least is in some way inhibited during a split. On the other hand I'd argue miners of the minority chain have a lot of pressure to change branches. Market is not yet a good indicator only hashing power and they'll lose money if they don't continue to mine on the more profitable chain. Their safest bet is to change chains. Also they know that other miners of the minority chain think probably similarly. That's why imo in Bitcoin history every fork/split was resolved very quickly and a branch "won".
- grwgreg 9y agohttps://en.bitcoin.it/wiki/Segwit_support https://en.bitcoin.it/wiki/Segwit_support None of the core developers support segwit2x.
- joosters 9y agoA big part of Bitcoin's entertaining in-fighting is due to the 'core' developers having a mindset that conflicts with many other people. Some speculate that this is because a large number of devs are paid by Blockstream, who have financial incentives to keep bitcoin congested so that their paid-for products (like their lightning implementations) have more value. I have no bitcoins, so I have no real preference in what happens. But watching politics destroy a 'decentralised' currency is good entertainment :)