3 ms·
I think a definition of "Decentralized web" should mean is in order. The Internet/Arpa net was created to be decentralized from the beginning. If a node, or m
by ThinkBeat 5y ago
I think a definition of "Decentralized web" should mean is in order.
The Internet/Arpa net was created to be decentralized from the beginning.
If a node, or more nodes went down, the rest of the network may be able to keep working.
If you, and me and all of our friends and all of their friends all ran out own web servers from our homes
it would be decentralized to the extent that if a server disconnected all the others would keep working.
The information on that server may be gone for all time.
The prevalence of enormously centralized monopolies on the internet is not a result of its technical architecture, but rather narrow business interests and a terminal lack of regulation.
Now we are all using AOL v3. (Facebook, Twitter, Google etc).
It can get semantically confusing.
Facebook is massively distributed in some ways.
Datacenters in several locations, CDNs all over the place.
It would take a hell of a lot (I think) to blow Facebook off the web. Or maybe some servers centrally will shut everything off it they are blown up.
IPFS is weird when taking about decentralized.
Storage is not on one big server located somewhere.
It is distributed among many.
In theory if one goes down, the data is still available.
but only if the server/nodes that make up the IFPS network
are reliable and dependable.
If a lot of nodes are run from laptops that that are shutdown at the end of the day that would become a
significant problem.
Since more and more and more data would need to be stored
in IPFS the need for dependable storage servers will increase rapidly.
I would like a simpler approach more in line with the original architecture, lots and lots of independent small servers. The people who run them can mirror each other in some fashion to persist data. Like FTP used to be (and maybe still is), like NNTP used to be (and maybe still is)
- pwdisswordfish8 5y ago> Now we are all using AOL v3. (Facebook, Twitter, Google etc) ...GitHub (the Facebook of software development)...
- fleddr 5y ago"The prevalence of enormously centralized monopolies on the internet is not a result of its technical architecture, but rather narrow business interests and a terminal lack of regulation." Neither business interests or regulation have anything to do with something being centralized or not. If you would launch "Thinkbeat's blog", in the traditional sense of web publishing, it is 100% centralized. Regardless of its size or your good or bad intentions. You would be in full control of it. You could intentionally retire the site, and all information and access is lost, forever. You could accidentally do that by forgetting to pay a bill. Another scenario, you publish something fishy on your blog, and get banned from the blogging platform, or whichever dependency you rely on. Further, all content on the blog is 100% owned by you, and none of it by the users. So you would also own and control comments to the blog, and arbitrarily edit or delete them. You have 100% control on everything, and users none at all. A decentralized version of your blog would be radically different. You can't just retire it. You can't permanent delete or overwrite prior entries. Commenters may actually own their own comments. I know, a blog is not a good example and a poor use case, just saying that the decentralized web very much is a difference in technical architecture.