Bitmain Blog

I think you missed the key issue. Mining is largely centralized under the control of Bitmain, both via pools and indirectly via miners they have sold (which in the past have had backdoors). They profit from using an exploit called ASICBOOST.

Segwit will disable ASICBOOST as a side effect of improving the protocol. Therefore, BitMain has been launching FUD campaigns and numerous attempts to propose alternative software to stop segwit (XT, BU, BTC1, BTCABC etc.)

Segwit2X is a compromise-- but Bitmain does not like it because it activates segwit.

So BitMain has announced they plan to do a forcible Hard Fork of Bitcoin, splitting off into another chain, that they will privately mine-- and they intend to neutralize segwit on it.

They have publicly complained that segwit offers lower fees for transactions that use less resources.

The entire conflict (including the "scaling problems" which have recently been proven to be due to spam driving up transaction costs) is Bitmain attempting to increase and protect their near monopoly.

Please be aware that their fork, which will have nearly unlimited block size will increase their monopoly. Due to the great firewall of china, large blocks have trouble getting transmitted, so whenever a chinese miner finds a block, when the blocks are very big, they have a significant competitive advantage in finding the next block. Thus simply activating a spam campaign (which they have been doing the past 3 months or so already) will allow them…

Also add that the largest manufacturer and user of ASICs is most likely covertly using something called AsicBoost [1], a patented [2] chip optimization for Bitcoin's proof of work. They (Jihan Wu/Bitmain/Antpool) have stated that the capability is manufactured into their chips but they are currently not using it [3] for the greater good of Bitcoin, though this would cost them millions in mining profits.

Activating Segwit as a soft fork would prevent the ability to use AsicBoost and therefore a lot of people suspect that most of this debate is actually just a delay tactic by Bitmain to maintain their mining dominance with covert AsicBoost. They likely have no intention to support Bitcoin Unlimited (they don't seem to be running the software).

Segwit recently activated on Litecoin only after the threat of a User Activated Soft Fork (much easier to execute on Litecoin). The Segwit activation threshold on Litecoin was 75% of the hashpower (compared to 95% on Bitcoin). At one point Segwit had achieved 75% of the hashpower until suddenly 3 mining pools increased their hashpower by about 300% in 1 day to block Segwit (the same day that Jihan Wu, CEO of Bitmain declared that the Litecoin miners to be shipped needed "additional firmware testing"). These pools were Antpool, LTC.top, and LTC1BTC [4], all of which controlled a much smaller fraction pre-segwit. For this reason, it's likely that Jihan Wu controls both Antpool and BTC.top and therefore about 30% of the hashpower on…

Well, that's a remarkably uninformative announcement. Here's an attempt at a neutral tl;dr from a Bitcoin amateur.

Bitcoin is currently suffering from significant scaling problems, which lead to high transaction fees. Numerous proposals to fix the scaling issue have been proposed, the two main camps being "increase the block size" and "muddle through by discarding less useful data" (aka Segregated Witness/SegWit). However, any changes require consensus from the miners who create Bitcoins and process transactions, and because it's not in their best incentive to do anything to reduce those transaction fees, no change has received majority consensus.

In an attempt to break this deadlock, there is a "Bitcoin Improvement Proposal #148" (BIP148) that proposes a User-Activated Soft Fork (UASF) taking effect on August 1, 2017. Basically, everybody who agrees to this proposal wants SegWit to happen and (here's the key part) commits to discarding all confirmations that do not flag support for SegWit from this date onward. If successful, this will fork Bitcoin, because whether a transaction succeeded or not is going to depend on which side of the network you believe.

However, BIP148's odds of success look low, as many of the largest miners out there led by Bitmain have stated that they will trigger a User-Activated Hard Fork (UAHF) if needed to stop it. Specifically, if UASF appears successful, instead of complying with SegWit, they'll start mining BTC with large blocks…

The days are counting down to the "Segwit2X" rollout, the idea supported in the "New York Agreement" (NYA)[0].

There is a contingency plan in place should the Core-supported User Activated Soft Fork become activated.[1]

Segwit2X has working code, has been tested in beta, and is now in RC.[2]

Without commenting on the merits of the different approaches, the current situation is thrilling to watch as a spectator. To call it a "Civil War" is not an exaggeration.

[0] https://medium.com/@DCGco/bitcoin-scaling-agreement-at-consensus-2017-133521fe9a77

[1] https://blog.bitmain.com/en/uahf-contingency-plan-uasf-bip148/

[2] https://lists.linuxfoundation.org/pipermail/bitcoin-segwit2x/2017-July/000145.html

Bitcoin network is going to split at end of summer, driving uncertainty in its future value. The SegWit proposal does not have universal support.
Some mining groups are advocating a hard fork, some advocate soft fork, many even want it to remain as is.
https://blog.bitmain.com/en/uahf-contingency-plan-uasf-bip148/
https://bitcoinmagazine.com/articles/why-viabtc-rejects-segwit-soft-fork-in-favor-of-block-size-hard-fork-interview-with-haipo-yang-1479409475/
http://www.uasf.co/
http://www.coindesk.com/bip-148-segwit2x-bitcoin-scaling-compromise-might-not-easy/
https://bitcoinmagazine.com/articles/beginners-guide-surviving-coin-split/

the only way your money is at risk with Coinbase is if BTC itself does lose out - the so called 'wipeout' scenario described here: https://blog.bitmain.com/en/uahf-contingency-plan-uasf-bip148

'wipeout' is a total longshot and we ought to be able to see it coming. either way it's not really a problem unless you have transactions on both versions to reconcile, and you can avoid that by simply not spending coin for a couple days. Coinbase can't just freeze everyones coin so having to reconcile with a fork after it achieves majority hashrate would be a real nightmare.