[SkyMarshal]

Should probably qualify this as "geographically decentralized".

As long as there are C-level executives, VPs, and the usual corporate heirarchy, they're not really decentralized in the sense that cryptocurrency itself is fully consensus-driven.

[fogof]

Indeed, I would say the article is mistitled. Or at least, it was titled in a way that tries to capitalize on the buzzwordiness of the term "decentralized" without really using the word in the sense in which it is usually used by the cryptocurrency community. A better title would have been "Coinbase now permanently Remote-first".

[CharlesW]

> As long as there are C-level executives, VPs, and the usual corporate heirarchy, they're not really decentralized in the sense that cryptocurrency itself is fully consensus-driven.

Is that the best litmus test? My impression is that Valve is decentralized and self-organizing company[1] even though Gabe still has a "President" title.

In any case, it's quite an achievement to lie twice in a relatively short title. It's also ironic considering that Coinbase's business model is built on centralizing cryptocurrency services.

[1] "Welcome to Flatland [PDF]": http://dl.pcgamer.com/Valve_Handbook_LowRes.pdf

[tessierashpool]

take a minute to look up the various "I was inside Valve and it had a hidden structure" war stories. that led to a lot of people reading "The Tyranny of Structurelessness" back in the day.

TLDR: "Welcome To Flatland" is Valve marketing and it doesn't hurt to take it with a grain of salt.

https://www.pcgamer.com/valves-flat-structure-contains-hidde...

https://medium.com/dunia-media/the-nightmare-of-valves-self-...

https://www.jofreeman.com/joreen/tyranny.htm

[bulek]

Important to note that Gabe gave the product rights that the fired team were developing. Said team went on to have a successful Kickstarter (250% successful!), and then failed in actually delivering results.

In hindsight, seems smart to pull the plug so far in advance.

I'm sure the complaints about informal power groups, careful politicking being required and gaurded groups are true... if only because that's true in every office I've worked in.

[owenmarshall]

> In hindsight, seems smart to pull the plug so far in advance.

Not knowing the product, this looks like it may be a valid interpretation; could another be that the project would be more likely to succeed with the commercial backing of a company like Valve?

[bulek]

Absolutely, let me walk back my language so that its clear I'm extrapolating from very limited data.

Edit: I can't edit it anymore. Was going to change it to "Based on this limited insight, it may have been the best option for both parties to part ways given the project teams complaints and valves lack of interest in the product"

[hypertele-Xii]

"I finally discovered who my boss was when she fired me."

[yjftsjthsd-h]

Surely that proves that it is decentralized; like so many decentralized services before it (IRC, email) the lack of the lack of formal structure meant that its structure was emergent instead (freenode, gmail).

[skybrian]

Sure. I think this is more a matter of Conway's law, or the old observation that when a company moves its headquarters, it usually makes the CEO's commute shorter.

They say that nobody on the executive team lives in San Francisco anymore, so that much seems likely to stick.

[phreack]

And even then, only barely, since all of the open jobs are US/Canada only. There's a bunch of countries even on the same timezones being skipped, so I'd say it's pretty geographically centralized too.

[Geee]

The word "distributed" is typically used to describe a system that consists of separated units. Decentralization is a more complex idea, but basically means that the authority of the system is distributed: https://en.wikipedia.org/wiki/Decentralization

[Melting_Harps]

> Should probably qualify this as "geographically decentralized".

What a loose use of the definition, which is oddly very much like Armstrong to attach his name and Coinbase to Bitcoin's virtues. And this is despite him actively helping push Bcash during USAF.

> As long as there are C-level executives, VPs, and the usual corporate heirarchy, they're not really decentralized in the sense that cryptocurrency itself is fully consensus-driven.

Also recall that these guys are ex Goldman Sachs with deep ties to VC, which includes YC and is definitely not decentralized in anyway. Why this is on the front page is showing it's ties to YC because otherwise the focus should be on how they keep shutting down during ATH, and price dips.

[CryptoPunk]

Bitcoin Cash stays true to Bitcoin's original vision of allowing every one on Earth to control their bitcoin with their own private keys, and transfer it without trusted third party intermediaries.

The Bitcoin Core cooption of Bitcoin - which was the cause of Bitcoin Cash being created - has led to Bitcoin being transformed into a currency that in any kind of future success scenario, only power-users (banks and exchanges) will directly transact with on-chain, with their own private keys. Armstrong's vision is much more pro-decentralization.

[SkyMarshal]

Bitcoin Cash foolishly believes they can wave a magic wand or “jazz hands” away the unavoidable technical constraints that inevitably led to Core’s roadmap.

Core accepts the technical realities and limitations of Bitcoin’s design and works within them. Bitcoin Cash simply wishes them away and ignores them, and that will eventually cause the whole system to fail.

They’ll learn the hard way what the wiser small block contingent have understood from the early days. Bitcoin Core’s layered architecture is the only realistically viable architecture given Bitcoin’s inherent design constraints.

[CryptoPunk]

>>Bitcoin Cash foolishly believes they can wave a magic wand or “jazz hands” away the unavoidable technical constraints that inevitably led to Core’s roadmap.

Nodes can handle far more than 7 transactions per second of throughput, even if you 4X the data size of that to account for propagation to multiple peers. 10 KB/s is nothing. Keeping nodes so light so that you can validate with a Raspberry Pi with a dialup connection, at the cost of preventing 100X more people from using Bitcoin, is a terrible tradeoff. There is no technological bottleneck preventing a 100X raising of the throughput limit: not Initial Blockchain Download, which is already being expedited with trusted third party set checkpoints, not computation, not storage and not bandwidth.

The vast majority of Bitcoin agreed with my sentiment, and that was the sentiment expressed by Satoshi any time he commented on it.

It was just irresponsible, overly timid leadership allowing a loud minority to sabotage Bitcoin's widely supported roadmap.

[SkyMarshal]

>Nodes can handle far more than 7 transactions per second of throughput, even if you 4X the data size of that to account for propagation to multiple peers. 10 KB/s is nothing. Keeping nodes so light so that you can validate with a Raspberry Pi with a dialup connection, at the cost of preventing 100X more people from using Bitcoin, is a terrible tradeoff.

That wasn’t the tradeoff. The tradeoff was that increasing transaction throughput would also increase the risk to loss of decentralization and thus effectively the systemic catastrophic failure of Bitcoin. Without decentralization it’s just technological trash.

The most vocal big blockers were business guys prone to the same kind of risk-blind short-term thinking that led to the Global Financial Crisis less than a decade before - improve some easily measured metric now at the expense of introducing greater medium to long term total failure risk into the system.

Part of the problem is, systemic risk is difficult to measure and quantify, whereas other metrics like ROI or TPS are easy to measure and quantify. And thus we get more of what we measure, and less of what we don’t (systemic resilience, safety factor, redundancy, etc.). Everyone has a very strong bias for optimizing measurable metrics while being blind to the unmeasurable but important, and that’s what was happening here.

Thankfully the Bitcoin engineering community understood this and resisted making this tradeoff.

>There is no technological bottleneck preventing a 100X raising of the throughput limit: not Initial Blockchain Download, which is already being expedited with trusted third party set checkpoints,

This is a fundamental and irreconcilable philosophical difference. You think third parties can be trusted indefinitely and safely integrated into a core role in a decentralized system. That kind of mindset is why Ethereum is being run almost entirely on Consensys’s centralized Infura database. But that mindset doesn’t fly in the Bitcoin community. Core and most others that prioritize decentralization and censorship resistance will, rightly, never agree.

https://nakamotoinstitute.org/trusted-third-parties/

> The vast majority of Bitcoin agreed with my sentiment, and that was the sentiment expressed by Satoshi any time he commented on it.

No, UASF proved that wasn’t true at all. The majority of people actually running Bitcoin nodes preferred Segwit and no block increase. The “loud minority” was the cartel of highly visible large businesses that wanted to increase the blocksize, not the “vast majority of Bitcoin”.

> It was just irresponsible, overly timid leadership allowing a loud minority to sabotage Bitcoin's widely supported roadmap.

No, it was wise, responsible, foresightful, and technologically savvy leadership that understood the potential butterfly effect here, small changes compounding the systemic catastrophic failure risk over time.

Even if that systemic risk to decentralization couldn’t be easily quantified, Core and small block advocates nevertheless acknowledged it, and integrated it into their analysis of the decision, and came to a very different conclusion from others who did not and who only focused on measurable metrics and short-term considerations.

It was the latter loud minority who proved to be irresponsible and almost forced Bitcoin into making the same category of error that Wall Street made in creating the GFC.

[CryptoPunk]

>>That wasn’t the tradeoff. The tradeoff was that increasing transaction throughput would also increase the risk to loss of decentralization and thus effectively the systemic catastrophic failure of Bitcoin. Without decentralization it’s just technological trash.

There's no reasonable case for there being a systemic risk to Bitcoin from node operators needing Netflix streaming levels of bandwidth.

>>This is a fundamental and irreconcilable philosophical difference. You think third parties can be trusted indefinitely and safely integrated into a core role in a decentralized system. That kind of mindset is why Ethereum is being run almost entirely on Consensys’s centralized Infura database.

Completely disingenuous characterization of my argument, complete with a disingenuous characterization of Ethereum's decentralization profile.

One more time: Bitcoin could remain fully decentralized, with tens of millions of people capable of running full nodes, with 100X greater throughput.

[UncleMeat]

> Bitcoin Core’s layered architecture is the only realistically viable architecture given Bitcoin’s inherent design constraints.

Then decentralization is dead and can never exist.

[SkyMarshal]

No, other protocols may have solved those design problems, like Mina (formerly Coda). This is just inherent to Bitcoin’s gen 1 design. https://minaprotocol.com/

[dheera]

Also, Coinbase isn't legally decentralized. If someone wanted to subpony them or serve them a court summons they still have a listed agent for service of process.

A truly decentralized company, in the ideological sense, would have none of that.

[riffraff]

I think the word is "subpoena"? Subpony is quite fun though.