Oddbean

Poll: How do you guys feel about governments funding freedom tech? Governments (US/EU) have funded the following projects: Briar Peertube WireGuard Signal Prosody (XMPP) Dino (XMPP) SearX (search) Stubby (DNS) ipfs-search.com QubesOS (support for disabilities) At least in some part Source: nlnet.nl Clearly these things make our lives better. But it was paid for with stolen money. Do the ends justify the means? Comment below.

Government grants are still their largest source of funding, if I'm reading their reports right. I'm not shitting on TOR or anything the tech is solid, I still use it all the time, but it's good to know where the funding comes from. Here is their sponsor page and most recent fiscal reports. https://www.torproject.org/about/sponsors/ https://blog.torproject.org/transparency-openness-and-our-2021-and-2022-financials/

"their tech is solid" and out of date just read a good paper on bloom filter based encryption and circa 2018 already there was a proposal for a one-shot protocol for circuit establishment via tor afaik not reached us plebs it's a performance improvement, sure, but i was already jaded on the tor project in 2007 for the way they regarded ideas about using a monetary incentive to improve the size of the network today been doing a lot of research into it and i get it now that with old techniques they literally are limited in network node count could be fixed with multiple networks with other directories running, at minimum circuit establishment speed can clearly be dramatically increased to one outbound message now, where it used to be a 6 cycle to and fro through the new channel, at minimum i had been developing an idea based on HORNET - the actual original inspiration for LN, btw, client side onion relaying - but the problem with the design is it's not spam resistant i pretty much guarantee that next year lightning is gonna be creaking because of this vulnerability not sure what to say other than these many points... gonna be following up with a lot more research in this holiday week and maybe i'll manage to figure out a possible path to agoric, path-blinded relay routing definitely feels like a dawning moment right now

Tor also, in the early days. i don't think it matters that much, this isn't exactly a long list. qubes doesn't work on most hardware. wireguard is ridiculously elementary in what it does, literally just elliptic curve keys and ECDH, same stuff used here in nostr. signal was probably developed by intels. it was originally called "red phone" and was the first chat system binding to mobile phone numbers and using the mobile network (SMS) as a back channel to pass shared secrets, iirc - or something like that, maybe it was just for a rendezvous.

i built the majority of the implementation of a replacement for Tor that uses the HORNET/LN style client-side onion messages. https://github.com/indra-labs/indra i was in the middle of changing part of the design of the protocol to make variable length routing headers and still need to design a connectionless UDP protocol (tricky because of congestion issues) but my funding ran out... i've been wanting to build an agoric, monetised version of Tor since 2007 at least. hopefully some day soon i'll get to finish the work. there's a lot i can write about it but that was the most important stuff, with those two changes made and some funding to launch an initial caretaker org/foundation to return the initial funds to the investors the rest is designed to be a freestanding economy.

Spooks, liars, profiteers, war criminals and child molesters need to communicate privately and securely like other people. More or less like other people, anyway. And Solar Winds and similar débâcles should of convinced them of the need for critical tools be to open source.

So long as it is open source and transparent, I welcome it. It is true Governments steal our money, supposedly so they can do "good things" for us. I won't complain when they finally, actually live up to those words.