Call it a demo of our micro-payment service? Try paying with the Taler wallet, the gratis demo currency "KUDOS" will work. Plus, we had a real problem with DDoS from bots given that generating the cgit is expensive. Now, https://git.taler.net/ (static version) has no paywall, and git:// is also completely gratis.
You can backup tokens and restore them to another wallet (and we are in the process of making that easier). We call that sharing, and it is not a transaction as it is simply granting access to the same money to both wallets.
There is no backdoor. There may be bugs (point those out, we'll fix!), and there is the very clear and documented statement that the receiver of funds is subject to legitimization based on the legal requirements under which the exchange operates. That is not a backdoor. Backdoor implies secrecy, we have no secrets other than our private keys. ;-)
The privacy properties for P2P transfers between consumers are identical to those with merchants: the exchange can require the receiver to provide information about themselves, but the payer remains anonymous just like for purchases at a merchant. And yes, it involves spending (burning, if you prefer) the old tokens (we prefer digital coins) and issuing (blindly signing) fresh coins. Afterwards, the fresh digital coins are indistinguishable from those issued during a regular withdrawal from a bank account.
On your point #2, just a small hint: At https://taler-ops.ch/ we legally operate GNU Taler in Switzerland already. Now, Switzerland is a bit easier (lower capital requirements, no DORA), but actually quite similar in terms of KYC/AML as FATF provides a common reference frame.
"It's maybe private in that the Vintner does not need to know who you are, but you can bet the control freak tyrants in the government will know that you bought that wine." -- actually no. Because the digital cash you "extracted" from the exchange was blindly signed, the exchange/central bank would NOT learn that you bought wine. Only that someone (anyone who previously withdrew a sufficient amount of digital cash) bought wine. Please do read up on cryptography, how to do this is known since 1989.
It's only partially open source. Some server-side code remains proprietary, and the client-side will depend on proprietary code of Google and Apple and they do not plan to support platforms that are actually Free Software. The law overall is badly written. For example, articles 12 and 26 effectively say that "The source is shared with public, except if it is proprietary or insecure." Or take Article 4: "The government may operate systems that protect the privacy of the identity subjects."
We do need to expire cryptographic keys eventually, there is no sane alternative to this. Now, note that "regularly online" here could mean once every X years depending on how the system is configured.
GNU Taler (https://taler.net/); privacy-preserving payment system. Written in C/Java/TypeScript/Kotlin/Rust/Postgresql/etc.; needs include coding (especially integration into other applications) documentation (review, proof-reading), testing (incl. benchmark, UX), packaging, translation (Weblate/gettext); any help welcome ;-).
On this Internet if you don't trust third parties to route all of your traffic through and peers are behind NAT or CGNATs. Or imagine a mixed network with some peers on https://en.wikipedia.org/wiki/Wireless_ad_hoc_network (s), possibly with some connected to the Internet. The key assumption we have (for the random walk to work) is that it must be a small-world network (https://en.wikipedia.org/wiki/Small-world_network) and network topologies surprisingly often are.
I'm personally primarily busy with GNU Taler, but still in touch with the more active developers. Major work is still happening on the transport subsystem with the goal of improving NAT traversal. Some work is being done to make the code run on Android. A messenger application (incl. text and voice) works --- if your network/NAT are happy, which most of the time they are not (see my first point). There is also fun cryptographic work happening on Reclaim:ID (SSI), and improvements to the GNU Name System (registrar implementation, automatic scalable import of large DNS zones). Plus some work on automatically testing the system using Linux network namespaces. But yes, quite regular breaking changes and still too frequent breakage on the transport layer means that the active network is tiny. OTOH, until NAT traversal reliably works (incl. our self-imposed restriction to not simply assume that Google runs a TURN server/ICE environment to relay our for us), GNUnet is really only suitable for developers anyway. So normal users sadly still cannot use it. But, development is still quite active.
Yes, the random walk relates to the "NAT problem". Basically, due to NAT, peers may not succeed when trying to connect to arbitrary other peers (especially if you assume no TURN servers or other infrastructure to facilitate). The same situation may also arise in mobile ad-hoc networks, where your wireless signals simply don't get everywhere. As a result, the usual greedy routing will get stuck in a local minimal. By prefixing the greedy routing with a random walk, you basically randomize the starting position, and then end up at a random local minima. Replicate at enough local minima (O(sqrt(n)) and do enough (O(sqrt(n)) lookups and the birthday paradox will make it increasingly likely for you to succeed. Without the random walk, both the putter and the getter would greedily route always to their local minimum, and if they differ, never find each other's data.
NGI funded us to work on the GNU Name System (https://www.rfc-editor.org/rfc/rfc9498.html), that includes the Go implementation, updates to the C implementation, and the specification. Plus NGI funds work on https://taler.net/. Another example for a project they funded is https://cryptpad.fr/. https://nlnet.nl/project/ has an extensive list. If you actually care about privacy on the Internet and FLOSS, I'd be surprised if you did not find some project there that you know or use.
If you are working on improving privacy on the Internet with FLOSS, you can still apply for funding at https://nlnet.nl/propose.
Indeed, I just wanted to point out that the "game changing" desire of entropyie to fully automate payments can still be done with GNU Taler. So you are right that the scammer in the default deployment won't succeed, but additionally entropyie could still totally get what entropyie wants as a feature on their own system (and conceivably make this reasonably secure, too).
Not quite. GNU Taler's reference implementation requires explicit user consent for payment. But as GNU Taler is Free Software, you are free to modify the code on your system (and share your with others) to not require an explicit confirmation if you prefer that. Or even better, to add an option to do that on some "trusted" sites which you make configurable for each user via settings. I don't know if the risk is worth the UX improvement, but a key contribution of GNU Taler is that it is a Free Software payment system where you actually have the full four freedoms. Use them!
Eh, you realize that this very work on the GNU Name System prompted IETF to create the ".alt" zone for this purpose already, minus the $200k fee? Registration is open at https://gana.gnunet.org/dot-alt/dot_alt.html
You are wrong on many details here. First of all, GNU isn't even a legal entity. The only legal entity is the FSF. Second, only some GNU packages have a copyright assignment to the FSF, not all. Some packages use a DCO, and some have way more creative solutions (see https://www.gnunet.org/static/pdf/copyright.pdf for example). Finally, not all GNU packages are necessarily GPL-x-or-later. Yes, that's generally a good default (and sometimes AGPL-or-later or LGPL-or-later are good, too), but not a strict rule.