Open source software generally both cooperates and competes. Handled well the benefits of the former offset the costs of the latter and the result is a gain for everyone. --- but software differentiating in consensus rules is the worst kind of competition: competition here can deprive users of the practical freedom to use their preferred software, and the fight risks leaving a salted earth in its wake.
So in the social UX layer, it's the ever-looming potential loss of "practical freedom to use their preferred software" that puts the criteria in 'consensus-critical.' If the data layer is borked by some catastrophe, the nodes are mostly useless for their users.
Not being
a hippie an altruist, I'd reverse or at least equate the assumptions w/r/t linking the benefits of cooperation and costs of competition. Dodo Island Effect, don't you know.
Then I'd invert the frame, asking if it is possible to compete at being the most cooperative ('frenemies')!
Microsoft was a pioneer of business strategy based on making incompatible extensions to formats, first leveraging their network effect and then-- after introducing incompatible changes-- using it against them, an approach they themselves called
embrace, extend, extinguish. Worse than zero sum, these kinds of moves can be tremendously damaging overall.
Bitcoin's creator described alternative implementations as a likely "menace to the network"-- words which I think were spoken with an early insight into the incredible difficulty in making distinct software actually consensus compatible even when that is your highest goal, an art our industry is still just learning. I wish we'd built mechanisms earlier on for better ways to enable diversity in the non-consensus parts without ending up with unintended diversity in the consensus parts. But we play the hand we're dealt. The potential harms from consensus disagreements from mistakes in re-implementation are tiny in comparison to those from adversarial implementations which intentionally push incompatible rules.
Very profound. Great catch on pointing out I'd mistakenly referred to as "zero-sum" contentious hard forks'
negative-sum game theoretics.
It's important to emphasize a Pyrrhic GavinCoin victory doesn't involve a smooth transfer of power from Evil Core to Heroic Classic.
There are Samson Options on the table. Many Gavinistas actually want a moment of maximal risk, either because they don't own many Bitcoins or like the narrative appeal of a TV Trope style
Cathartic Scream.
If hard-forkers insist on a first strike in the form of a "75/25" split, the rest of the network may respond with their Doomsday Machine, aka PoW change.
The sum productivity of such a situation quickly drops from zero to extremely negative. Perhaps that's just creative destruction at work.
But the recriminations will surely flow, as the Gavinistas survey the decimated Bitcoin ecosystem and exclaim "Look what you made me do!"