Why has BitShares failed?
BitShares cannot afford to wait for the second coming of its messiah (i.e. Dan Larimer) once he is done with Steem. It needs to find a new leader, fix DPOS, and move on. Or slowly sink into oblivion.
This post is about the "why", not about the "if".
If you believe everything is fine and BitShares still has a bright future, please stop reading, as my reasoning will only annoy you and you'll probably perceive it as trolling or FUD.
First, I'll list things that IMO are not the reason:
The technology. Definitely not. Graphene is absolutely powerful, for me only Ripple can match its robustness and efficiency. Also, it's not the GUI - it's still not perfect but good enough when compared to similar projects.
The Larimers' (Stan & Dan) decisions. They've made a lot of mistakes and changed their plans and promises several times, hurting a lot of early investors, and, as a result, many of them dumped their shares. IMO this was inevitable as this is the very nature of projects that dare to enter uncharted territories: they need to pivot a lot before they have a chance to succeed.
The demand for the product. For a long time I thought this was the major weakness. Who needs a fully decentralized exchange (DEX) when scams like MtGox seem to be less and less frequent? Who needs Market-Pegged Assets (aka smart-coins) when projects like Tether or Uphold seem to be good enough for most people? Those are still valid issues, nevertheless I think a well implemented DEX offering reasonable liquidity for both UIAs and MPAs has a secure place on the market.
Lack of certain features, e.g. stealth, bonds, VM for smart-contracts. By comparing BitShares current capabilities to the existing features of other projects, it's clear to me that it's not the lack of features that prevented wider adoption.
The reason
IMO, the main reason lies in the very heart of the system - DPOS (Delegated Proof of Stake). Until recently I thought DPOS was a rock-solid foundation for solving the biggest problem in the crypto-space: the blockchain governance issue.
But unfortunately it turns out that DPOS has a major flaw. It's not a security flaw but an incentive scheme flaw. DPOS does not have an in-built mechanism to effectively encourage brave, long-term decisions. Instead, it favors lack of decisions or short-term decisions that give fairly immediate outcome.
The corporate world has a lot of internal flaws but its efficiency in executing large-scale projects is unquestionable and one of the crucial parts of it is the concept of board of directors headed by a CEO. The problem is that DPOS, in its current form, does not offer any solution that could replicate the role of the above structure. Besides, there is no incentive to vote, and the voting system does not even differentiate between long-term holders and short-term ones. As a result, BitShares has ended up in a fatal decision paralysis.
In a DAC or DAO there needs to be some entity which is elected by the shareholders but has the ability to make tough long-term decisions. The alternative is to use decentralized technology but retain centralized governance by the existence of a company that de facto controls the business strategy of the blockchain associated with it. This path is currently followed by Ripple, Ethereum, Lisk and... Steem. And it seems to work.
Conclusion
I guess some will argue that the situation with BitShares is not so bad: gradually stealth and other features will be implemented, liquidity on the DEX will slowly grow and everything will end well. My judgement is that this is not good enough, at least not in the crypto-space where there is so much competition and new capital flowing in. BitShares cannot afford to wait for the second coming of its messiah (i.e. Dan Larimer) once he is done with Steem. It needs to find a new leader, fix DPOS, and move on. Or slowly sink into oblivion.
This might be a minor quibble but I don't like it when people conflate DPOS, the blockchain consensus mechanism, with the other governance structures in BitShares.
I don't think DPOS (i.e. the system that elects the witnesses who produce blocks) has any major flaw, technical or incentive-based. The other governance structures (committee, workers, voting rights, funding allocation) is a different story.
I understand that is what you seem to mean, but I think it is worth stating that difference.
I agree that voting rights should be differentiated based on how long one commits to hold stake, similar to Steem.
However, the governance features in BitShares do technically allow mimicking something close to traditional corporate structures. The committee is the board of directors which is elected by the BTS stakeholders. A long-term worker proposal sending considerable funds to the committee account could be elected by the stakeholders. The committee would then redirect the funds (perhaps keeping a small portion to compensate them for their work) to another account that they choose (the CEO).
The main difference is that everything can rapidly change in BitShares in just one hour, whereas the timescale for change (replacing the board of directors or cutting off their share issuance authorization) is much slower in a traditional corporation. Maybe there needs to be hard-coded features for a little bit of added inertia into the system in order to restrain the (usually misguided) impulses of the stakeholders?
Hopefully a hard-forking code change like that isn't necessary. I think the stakeholders just need to restrain themselves from micromanaging and to give the benefit of the doubt to their elected committee for some reasonable amount of time. If a new competent leader can be found, they need to just approve a huge worker to the top (paying the committee account), elect a trusted and vested committee in charge of keeping the CEO in check (and finding an alternative if necessary), and do their own checks on the CEO's performance and committee's management on a longer time scale (quarters rather than weeks).
Yes, this is exactly what I meant and thank you for pointing it out.
I agree, the block production layer of DPOS works perfectly, it's the governance layer of DPOS that failed.
The problem is, by design the committee was not meant to work this way and with the current mindset of the shareholders there is no way to redefine the committee's role in the way you propose (which I like a lot). It seems to me that the system needs to collapse (or at least come to a complete stall) to make the shareholders realize the gravity of the situation. As of now, most of them seem to be confident that everything will be fine ("we don't need a leader"), just as the recent holders of the DAO are confident that collective investment decision making will work in practice.
I agree with your analysis, I was actually going to write similar article myself – but now that you did it, I don't have to.
Although I still think that DPOS is quite good system, better than most blockchain governance models. Problem is the quality of shareholders. There are too many BTS owners who don't like the current system, but instead of selling everything, they just try to stop all development.
Bitshares is like a startup company that sold too much shares to investors. Now they are not on the same page anymore. Too much conficting interests, too much ignorance, too much different views, etc. Because of this, nothing gets done. People start to abandon the community, developers are not interested, new money isn't flowing in...
My suggestion how to fix Bitshares is to launch a new chain and sharedrop it to people who have shown that they love Bitshares 2.0 system and willing to work for it. And of course Non-transferable BTS an BitBTS would solve a lot of problems, too.
I used to think this way as well, but on second thought I'd say the quality of voting people is nothing we can do about. The only way out is to design the incentive scheme in a such a way as to promote those who are better decision makers than the average shareholder. But I agree, it's a big challenge in a stake-based voting system, and I don't know how to do it. It seems to me that copying the corporate structure of power is the only way to go. On some level it saddens me to reach this conclusion but maybe this is the truth we need to face.
One way to solve low quality shareholder problem is obvious but not talked about (at least I haven't seen): let DAO have a right to confiscate shares from shareholders. If there was a bad situation where everything has stopped and nothing gets done, easiest solution would be, of course, to destroy the opposition. After the shares from opposition have been confiscated and redistributed/destroyed, DAO can work again.
In reality confiscation would be used very rarely, if ever. It would be enough if shareholders were aware of that possibility. Those who disagree would sell their shares before any bad conflict would arise.
I'd say the reason is spreading themselves too thin and not enough clear marketing. Note that all the projects that "succeeded" have more user friendly UI and focused on doing one specific thing well.