• Kantai Collection (KanColle) – A Japanese browser game (later anime) where WWII-era Japanese warships are personified as girls. Hugely popular in Japan around 2013–2016. KanColle started mostly with Imperial Japanese Navy ships, but later added foreign ones. For France, the characters are:
  • Richelieu – Personification of the French battleship Richelieu.
  • Jean Bart – Battleship, Richelieu’s sister ship.
  • Commandant Teste – Seaplane tender.
• Azur Lane – A Chinese mobile game (also with anime and manga) that includes warships from multiple countries (Japan, USA, UK, Germany, etc.) as anime girls. This one has a more international cast compared to KanColle. Azur Lane has a whole French faction called Iris Libre (Free Iris, based on Free France) and Vichya Dominion (based on Vichy France). French shipgirls include:
  • Richelieu – Battleship, leader-type character (Free Iris).
  • Jean Bart – Battleship (Vichya Dominion, later joins Free Iris).
  • Le Malin – Destroyer.
  • Le Triomphant – Destroyer.
  • Algérie – Heavy cruiser.
  • Béarn – Aircraft carrier.
  • Saint Louis – Heavy cruiser.
  • Gascogne – Battleship (super prototype type).
  • Dupleix, Émile Bertin, Vauquelin, Kersaint, Forbin, Surcouf, etc. – Various cruisers, destroyers, and subs.

PS: those two games I had in mind immediately, but I used a bit of AI to put up those lists of characters, added some links manually (for other characters just swap the character name in query)

Yes, I believe smartphones and tablets are the first devices for almost everyone these days. Especially since kids are equipped with smartphone when parents send them to school.

This is interesting, I did a bit of research and it seems, none of this is legally enforceable unless the company has EU presence. Basically EU just saying “we will do everything we can, but we can’t really do anything if you don’t have any operations on our land”.

I struggle to understand, why do those sites block uk users? Are there really any “international regulations” that demand that if you don’t want to comply with whatever arbitrary rules some country set, you should stop serving users from that country?

Pixiv, Fanbox, DeviantArt, Tumblr, etc, are also widely used. Very few people only use a single platform. I think Twitter is top 1 for expanding your audience not only because how well their feed algorithm works, but maybe also because all those focused platforms are used more by artists and less by viewers (or used less often by viewers), while Twitter being general-purpose is the one where more people who like to watch/discover arts but are not artists themselves, are. But there are other factors, like Twitter comments being better than Pixiv or DeviantArt comments, etc. Finally, if we return to the context of this discussion, I don’t think any of those dedicated platforms in any way solve the problem of age verification and that is why I wouldn’t recommend migrating to them in this context even if they were otherwise good for art.

I had an impression it didn’t work great across instance boundaries. Like, algorithmic discoverability was very limited. I might be wrong and it might have changed since I last checked though. Also I had an impression that Mastodon doesn’t really have global cross-instance feed on the same level as Lemmy instances. And again, correct me if I’m wrong here.

It’s still the biggest art posting platform. And I’m not even sure where art posters should migrate to… I mean sure it would be nice to have them scattered across different fediverse instances, but it would be nice for us, not for them. The main thing they get from X is massive algorithmic reach. You hit like on a Miku art and another artist with their Miku art immediately slips into your feed, you like it even more and you decide to check their profile and you like their other works and you subscribe. This kind of easy and efficient advertisement is something that doesn’t exist anywhere else outside of few centralized systems.

Twitter / X started asking for age verification for adult content when browsed from EU. Works fine from Asia but you also need to set your account country to the one you’re browsing from.

Pretty much impossible within I2P

Only if you’re deliberately running an exit node (doing so requires special setup).

Owning a lot of Tor exit nodes doesn’t automatically deanonymize users. Exit nodes only see the traffic as it leaves Tor toward the clearnet, not the original sender. To actually identify someone, you’d need to match their traffic entering the network with the traffic exiting - a correlation attack - which requires visibility on both ends. The US doesn’t “own most exits” either; the network is run by many independent operators, and the Tor community actively monitors for malicious relays. Even if a law forced US exit operators to log everything, that alone wouldn’t deanonymize anyone unless combined with large-scale surveillance of entry traffic, which is extremely resource-intensive and not guaranteed to work. In practice, governments can make running exits legally risky, but they can’t just legislate Tor anonymity away.

In my experience, if you have anything but “Network: OK” status (for example, “Network: Firewalled”), it’s not working properly. If you’re behind a VPN, you need to port-forward and properly configure a port in I2P config/settings. Another sign that it’s misconfigured is 0 participating tunnels. This is how properly configured I2P network statistics looks like with high internet bandwidth:

• If the internet were fully controlled, you’d need mesh networks - DIY, decentralized networks using radios, local connections, or other alternative infrastructures. I don’t know all the details, but Yggdrasil is a promising modern project that functions as an alternative “internet” for mesh networks, while also working over the regular internet.

• Within the normal internet, the most resilient solution against heavy censorship is probably Shadowsocks. It’s widely used in mainland China because it can bypass full-scale DPI (deep packet inspection) by making traffic look like normal HTTPS. There are ways for authorities to detect it, and there are counter-methods, but it remains one of the most reliable tools for evading state-level traffic filtering.

• Next in line are Tor and I2P. Both are very resilient, and blocking them completely is difficult. It’s a continuous cat-and-mouse game: governments block some bridges or entry nodes, but new ones appear, allowing users to reconnect.

• Finally, regular VPNs are useful but generally less resilient. They’re the first target for legal restrictions and DPI filtering because their traffic patterns are easier to detect.

Overall, for deep censorship resistance, it’s a hierarchy: mesh networks > Shadowsocks > Tor/I2P > standard VPNs. You can ask chatbots about any of these and usually get accurate, practical advice because the technical principles are public knowledge.

• Tor is optimized for accessing the regular internet anonymously. It uses onion routing with a small number of long-lived relays, and you exit back to the clearnet through an exit node. Hidden services (now called onion services) exist, but they’re secondary to Tor’s main use case.
• I2P is designed primarily for internal services (called eepsites, torrents, chat, etc.) inside the I2P network itself. It doesn’t rely on exits the way Tor does. It uses garlic routing (a variant of onion routing with bundled messages), and every participant is both a client and a router, making it more peer-to-peer.

If you mean an HTTPS ban, it’s technically possible, but even mainland China and Russia haven’t gone that far. One major reason is that it would completely undermine basic internet security. It would instantly make man-in-the-middle attacks trivial, letting anyone sniff purchases, transactions, and more. Buying anything online - or using a credit card at all - would suddenly become extremely risky.

This is why it’s perfect time to get some tech literacy regarding tor, i2p, yggdrasil, and shadowsocks. It’s not perfect solution to use tech to circumvent restrictions that shouldn’t be there in the first place, but sometimes it really comes to that point and it’s really nice to have all systems ready!

torrents

Copilot is an extension for VS Code that does integration of their supported AI models with code editor (including creation/removal/moving/renaming/editing files, providing git-like diffs and per-file keep/discard choices for what AI made in response to your most recent prompt, etc). They only support fixed set of models and don’t provide any way to integrate free / opensource models. So the question is mainly what’s the alternative to Copilot for free models.

Is there anything as good as vscode-copilot for free models? I mean, integrating the process of querying models with actually generating and applying diffs to the files in project, etc.

I’m still on my free month of full-featured Copilot and I’m considering subscribing after it ends (10$/month). Mostly coding, bash scripting.