title: People Die, but Long Live GitHub
description: Some Thoughts on Electronic Archives Based on GitHub
date: 2023-03-17 19:15:03
I don't know if anyone has noticed, but Joe Armstrong has been busy migrating his blog to TiddlyWiki in recent months. I have been following his Twitter for a long time, but I didn't think much about it before. It wasn't until I heard that the master passed away on April 20th that I realized his previous actions were a precautionary measure.
TiddlyWiki is a single-file wiki system, but that's not important. What's important is, where do you store your information? If you want to store information that can be accessed by people 100 years from now, what should you do?
- Facebook, Twitter, Weibo? Let alone 100 years, I doubt if they will still exist in 20 years;
- Google Cloud, Amazon, Alibaba Cloud + personal domain? They may exist longer. But no one can guarantee that accidents won't happen, such as the previous Tencent Cloud incident, or what if the server is hacked and the files are deleted. Also, how do you ensure that the information can always be accessed? Domain names expire. At some point, machines running old versions of operating systems may be forced offline or upgraded. Can your service still run properly on the new version?
- Dropbox, Google Drive, Baidu Cloud. Similarly, it depends on whether you believe these companies will still exist in a hundred years. Oh no, even if they do, the services may have been shut down long ago (looking at you, Google);
- Wiki. Wiki is great, but it is not suitable for storing personal information and can be edited or deleted.
- Decentralized storage, such as blockchain. To be honest, I have limited knowledge of blockchain, but intuitively, I doubt if it can help us achieve our goals. Let's compare it to another decentralized example: BitTorrent. When you want to download some old animations or movies, it is common to encounter slow download speeds because the "seeds" are dead. And that's just after a few years. Of course, maybe blockchain has some magical solution to this problem? Welcome for friends who know more to supplement.
So, what other choices do we have? After thinking about it, there is only GitHub.
GitHub has become one of the most important infrastructures on the Internet, with too many people and too many things directly or indirectly relying on GitHub. Unless humans no longer need open source code in the future (which is obviously impossible), I can't think of any possibility of GitHub shutting down. For GitHub, existing for 100 years is nothing, and even 500 years is not impossible. This is my prediction, it may not be accurate, but I am quite confident.
Anyway, for our goal of 100 years, GitHub is fully capable. In addition to the persistence of the service itself, GitHub has two unique advantages:
- Git. Git can save all history.
- Fork. Even if a hacker hacks into an account and deletes a repository, can they delete all the forks?
In conclusion, at present, I believe that GitHub is the only way to store information and make it accessible on a scale of a hundred years. There may be option 2, option 3 in the future, but GitHub will still exist as option 1. I believe that Joe's use of GitHub to host his blog was not a sudden idea. He must have considered all existing storage methods after understanding his own physical condition, and then realized that only GitHub could meet his needs.
People always want to leave some kind of trace to prove that they have lived, but in fact, 99.999% of ordinary people are forgotten by history—this used to be the case. We are in the early stage of the information age, and also in the early stage of human civilization. From now on, more and more things will be digitized. Since someone has realized the uniqueness of GitHub, as time goes by, more and more people will realize it too. What will happen then? Naturally, more and more people will move their information to GitHub, relying on GitHub to achieve the "immortality" that was once beyond reach. People die twice, first in the flesh, and second when they are forgotten. I can't remember who said this, but now we can avoid the second death. As long as GitHub supports it, there will definitely be people doing this, and I am one of them at least. In a few decades or centuries, GitHub will become the largest digital cemetery in the world, with most registered users deceased. However, personal homepages, projects, and commit histories will still tell the stories of what they did in their lifetime—just like Joe's blog. Although this is a somewhat creepy inference, from another perspective, it proves the great progress of humanity: fighting against death is the eternal theme of human civilization, and we have achieved a phased victory. Now it is articles, photos, videos, and perhaps even models trained with personal habits as input. What about the future? Will there be genetic information, or even complete copies of consciousness? With stable storage, there is so much we can do. The counterexample is some Memorial Websites now, which put the information of the deceased on their own websites for family and friends to mourn. This is not useless, but in my opinion, it is somewhat reckless to use such services—even if they are confident, they cannot escape the risk of shutting down one day. Can we still expect them to manage this data properly at that time? Impossible.
Since it is inevitable for GitHub to become a digital cemetery, my only hope for them is to maintain some kind of moral obligation. I can totally imagine that one day they will introduce a policy to archive all accounts without any activity within twenty years and take down all GitHub Pages. That would be truly terrifying. I pray that day will never come.
Original article link: People Die, but Long Live GitHub - laike9m's blog