Slashdot Mirror
If You Can Decentralize the Internet, Mozilla Has $2 Million For You (cnet.com)
Mozilla and the National Science Foundation want a new internet. And they want it to be free and accessible for everybody. From a report: They'll pay $2 million for it. On Wednesday, the two organizations issued a call to action for "big ideas that decentralize the web" as part of the "Wireless Innovation for a Networked Society" challenges. The challenges include getting the internet to communities off the grid, with proposals like a backpack with a computer and Wi-Fi router inside.
-Can't do it, it's too late
-Need more money
-Brendan Eich! ahahgfhahadgdaha!
Linux, you magnificent bastard, I read the fucking manual!
https://freenetproject.org/ Those guys are already trying to do it. It is fully decentralized and private. It is very slow, and consumes huge bandwidth, but it works. The real concern here is the lack of choice when it comes to ISPs. They control the last mile, which almost everyone MUST lean on if they want to be on the internet. Break up the monopolies/duopolies and most the problems Mozilla wants to solve evaporates.
The Internet was designed to be distributed so that it had no central point of attack/vulnerability. Was NOBODY paying attention for the last 20 years while money-grubbing businesses jockeyed for control, thus creating the very problem that it was designed to circumvent??!!
HOW FUCKING STUPID DO YOU HAVE TO BE??!!
slashdot: A failed experiment.
If they want to decentralize the web, DNS and the SSL racket has to change. Domains have been completely compromised by both business interests, particularly the .com domains which have been squatted to hell and back, and government interests that can take away those names just because your politically inconvenient (See: Torrent sites).. And the SSL racket has to go, why the hell should we have to pay huge sums of cash to companys that *clearly* can not guarantee the integrity of the trust chain for certificates and have let us down again and again.
To my thinking, whatever must come next must be decentralized and let *US* choose who we trust and who we don't, both for domains, and for encryption.
Excuse the Unicode crap in my posts. That's an apostrophe, and slashdot is busted.
The problem with this mentality is it defeats the entire purpose of decentralization and non-censorship. If you want a fully free and uncensored internet, you will always have to put up with sites/opinions/ideas you don't like. That is part of FULLY free speech. The left and the right both cry foul about censorship when their ideas are being squashed, but are very will to squash others ideas they don't agree with. If you want an open internet, you get 4chan (and worse) in the mix.
Mozilla isn't trying to decentralize the internet. The challenge with the money involved is either to deploy access to places that have none OR deploy BETTER access to places that have lousy access.
NEITHER OF THOSE IS "DECENTRALIZATION."
Incidentally, whatever happened to Internet2? The very high speed internet that colleges & such institutions were working on?
Anyway, my suggestion: for such a thing, deprecate IPv4 and use only IPv6, and that too, using a 96:32 split instead of 64:64. And make this hierarchichal, so that the uppermost blocks drill down from IANA -> RIR -> Nation -> Organizations. And instead of having provider independent IP addresses, which tends to break that, encourage them to use multicast addresses to group the IP addresses that they have in different domains, be it nations or complete regions.
I'm not sure how to do it, but I know it'll involve middle-out compression.
https://en.wikipedia.org/wiki/...
Freifunk (German for: "free radio") is a non-commercial open grassroots initiative to support free computer networks in the German region. The main goals of Freifunk are to build a large scale free wireless Wi-Fi network that is decentralized, owned by those who run it and to support local communication.
The initiative counts about 400 local communities with over 41,000 access points. Freifunk uses mesh technology to bring up ad hoc networks by interconnecting multiple Wireless LANs
Now it's all black cock for white girl and no baby-making since its all in the ass anyway.
SSL is now completely free via let's encrypt.
Let's Encrypt requires a fully qualified domain name (FQDN) under a well-known top-level domain (TLD), not an IP address in RFC 1918 space or a name under a made-up TLD such as .local or .internal. So do all other CAs whose root certificates are included in Mozilla NSS, as a FQDN is one of the Baseline Requirements adopted by the CA/Browser Forum.
Domains are cheap.
Cheap enough for every head of household to buy and to continue to renew in perpetuity? Because buying a domain is the only way to get a certificate for hosts on your LAN that visitors' devices will trust, and a certificate is the only way you're going to satisfy the "Secure Contexts" requirement for recently introduced JavaScript APIs.
Free ones are available.
Namely?
If you're referring to subdomains offered by dynamic DNS providers, these providers have to be on Mozilla's Public Suffix List (PSL). If a domain isn't already on the PSL, and 20 other users of subdomains under the same domain have obtained certificates in the past week, Let's Encrypt will deny you a certificate, citing its rate limit policy. If a domain is on the PSL, each subdomain gets its own separate rate limiting bucket of 20 certificates per subdomain per week. In addition, submissions to the PSL must be made by the dynamic DNS provider as a pull request through GitHub.com, and use of GitHub.com requires running proprietary software written in JavaScript on your computer.
http://librarybox.us/
US military has been working on this for a long time and a lot of the research is freely available. For example, multi-hop wireless networking has known scalability limits (you can't connect everybody to everybody because the system slows to a halt, a few hundred nodes is probably the limit) so you need a certain percentage of basestations that are better connected via microwave links or wire lines. Store-and-forward helps with disconnection but is very slow and how do you secure such a persistent connection or decide when it should be dropped? How to only partially trust a newly discovered remote service? (See papers on mobile agents and execution on servers that are assumed to compromised.) . I haven't looked at this area for a while, but I think the state of the art is basically what the US uses in warfare, which is small, local multi-hop networks connected to mobile basestations that connect to satellites with encryption on everything and some jamming resistance. How to distribute encryption keys is always a question; you can cooperatively generate a shared key with a remote connection, but how much do you trust them? How do you trust de-centralized services? Reputation systems can help trust be self-healing. At the least some sandboxing is required at multiple levels of both the networking systems and operating systems, including the ability to reset a system back to a known state (e.g., boot from CD-ROM). There are a lot of difficult questions, and ultimately everything relies on people, so it will never be perfectly secure, but some kind of verifiably-fairly-trustworthy-most-of-the-time system that can self-heal-eventually can probably be cobbled together. Opening up long range radio bands to open-access-networking would be very useful for decentralization. Notice that in the US citizens have little access to open and hackable long-range-wide-area radio links that can carry any kind of content. Cellular and Ham radio is fairly tightly controlled as are high power emissions in any band. Ultra Wide Band might be useful for solving the problem of sharing wide area radio links without bad actors destroying the shared resource. Bouncing signals (laser or radio) off the moon is limited to when the moon is above the horizon (and not obscured by trees or buildings) and is tricky to do. Maybe a grid of big metal reflectors in orbit that everyone could bounce signals off of could work more predictably than bouncing signals off the ionosphere?
How's life in the hypocrite lane?