>I moved my code from GitHub to a self-hosted Forgejo
It was more about the symbolism. If the goal is decentralization, “I moved to a personal forge I control” is the post's core idea. But framing it as “leaving GitHub for Forgejo” inevitably creates a new flag to gather around.
That may be useful and even necessary, but it also shows that decentralization movements often produce new centers, names, and identities.
If the Forgejo core team announces tomorrow that their 'philosophy has changed' and overhauls the architecture, tens of thousands of 'independent' home servers will grumble, but they'll inevitably run git pull to update anyway. And eventually, they might migrate again. That cycle is inevitable, and I have no intention of denying it.
The issue in the context of this post is that it rejects dependency on GitHub while presenting Forgejo as the escape. If your defense is 'why reinvent the wheel,' you are essentially proving my point: we are just choosing a new center to depend on.
Beyond that, we might just be talking past each other at this point. I don't think you are wrong, and I have no intention of twisting your words. Your underlying point is likely this: GitHub's service quality has degraded, and with their aggressive AI strategies driving users away, it is strange to view the act of leaving negatively. I fully respect our difference of opinion there.
I honestly just got a bit annoyed earlier because you framed your reply as if I hadn't even read the article. Anyway, let's leave it at that. There is no reason to misrepresent your views, and no reason for us to argue further.
This is a hilarious - and empirically wrong - thing to say considering Forgejo's raison d'être[1] and lineage.
If users just grumbled and ran git pull anyway, Forgejo would not exist, and neither would its fork-parent project (Gitea); everyone would be using Gogs under duress. Forgejo's past shows that people care about the philosophy, which is why this particular fork of a fork is thriving.
People may converge on a new node like forgejo, so from that point of view things are not decentralizing, the dependency graph stays roughly the same shape, but it’s a step in the right direction. It’s marginally better than github hosting everything.
That’s what I should have said originally instead of being an asshole.
There is no difference between what you say people cry for and what you say they actually want.
What do you think, what is the business for GitHub in providing limitless private and public repo hosting?
Others recognize there's some kind of more comfortable middle ground where decentralization means the same as a town/city/state type of social good that is independent and capable of working without larger centralized structures. Having to work towards it, pay money into it, etc, are expected but because the work that goes into maintaining the infrastructure has a clear line of derivation (taxes clearly go to X, Y, Z) would be a benefit.
It's typically the first class tho that dominates all conversations regarding decentralization, and that class includes the Epstein billionaires who just dont want laws to apply anywhere they want to do unethical, immoral and whatever. eg, money is the only law.
https://archive.ph/2024.08.23-032320/https://vitalik.eth.lim...
It doesn't necessarily mean we should immediately suspect them of raising a flag for its own sake.
https://archive.ph/2024.08.23-032320/https://vitalik.eth.lim...
Epstein files remind me of the first line from《삼국지연의》or Pan-Asianists who raise the Universalism flag..
>The Himalayas divide, only to accentuate, two mighty civilisations, the Chinese with its communism of Confucius, and the Indian with its individualism of the Vedas. But not even the snowy barriers can interrupt for one moment that broad expanse of love for the Ultimate and Universal, which is the common thought-inheritance of every Asiatic race, enabling them to produce all the great religions of the world, and distinguishing them from those maritime peoples of the Mediterranean and the Baltic, who love to dwell on the Particular, and to search out the means, not the end, of life.
I spent about two hours reading your writings. Your arguments, such as those regarding the Taiwanese digital democracy movement and how to break the endless cycle of centralization and decentralization, are quite profound. If I were to summarize what you are trying to tell me, it seems to boil down to: "Your previous point was too generalized, and the attempt itself [at building alternatives] should not be dismissed."
Also, your Romance of the Three Kingdoms analogy confirms we are definitely both East Asians: "The empire, long divided, must unite; long united, must divide."
Overall, if I have understood your work correctly, your core message is this: while historical attempts at universalism or a 'new order' may eventually turn into new centralized powers, we shouldn't look at them only through that cynical lens. Instead, we must simply remain constantly vigilant.
I hope I understood you correctly.
No need to learn nuance, you seem to be doing better than most of the natives here just by being yourself (with the help of modern tools)
I just wanted to get more people to read the vitalik.eth articles along with me. Hope you got something out of 2hours! I had been nibbling at them and steeping in their diagrams for weeks, your centralisation comments provoked me (in a good way)
Vitaliy Buterin and Balaji Srinivasan are both based in SEAsia, Audrey Tang is in Taiwan, Glen Weyl from California is the only Westerner--- but UCBerkeley is like 80% East Asians? Surely they also know about the lore.
I don't know if "staying vigilant" is the mood I was going for. Maybe "being aware of emerging centralising entities and be prepared to engage fruitfully with them"?
These days, "emerging centralising entities" might even be old programming languages like APL that have become popular again..
In the context of the links above, such entities need to provide an "exit option" in order to gain wider traction. For example, APL is clearly an escape from C-like syntax, but there might be new directions that that such "array languages" can take from Python or even natural language: