And instead of using a secondary ISP (what a #VPN provider actually is), you should be using a proper #Anonymizing overlay network, like #Tor instead.
HideMyAss was about as clear an indictment of the whole model as you could get. Operators will not go to jail for you, so they must be unable to have enough data to get you in trouble instead.
Such overlay networks only address *some* of the routing issues though, not overall unreliability.
@ad_on_is The truth is that the #internet should be considered useful as a WAN routing layer, but nothing should actually be using it directly as it is unreliable both because of malicious actors and because of gross neglect by many operators.
Support for #AsynchronousCommunication (which #Usenet demonstrated) is an essential property. #Anonymization, #P2P routing & key-addressing are also minimum requirements to hinder trivial censorship. #Mixnet operation is even better.
"The 'net interprets censorship as damage and routes around it" was a comment about #Usenet, a #federated / #P2P system with gossiped (https://en.wikipedia.org/wiki/Gossip_protocol) message exchange which wasn't particularly picky about its transport layer (indeed you could load a spool on a floppy and mail it), not the internet.
@kde@floss.social@kde@lemmy.kde.social From the article it seems more of a #Qt win for now (though it does mention patches for many others), but in any case that's neat.
Now it all just needs sane ways to interface from #CommonLisp.
@sailor_sega_saturn@TinyTimmyTokyo Eh, no guarantee (or any reason to believe really) a simulation would be even focused in any way on humanity (no anthropocentrism needed).
Similarly for superintelligence, few reasons for it to care.
Cryogenics is a better bet and as you say it's quite unlikely unfortunately.
@cecilkorik @ad_on_is Encryption yes, VPNs not really.
And instead of using a secondary ISP (what a #VPN provider actually is), you should be using a proper #Anonymizing overlay network, like #Tor instead.
HideMyAss was about as clear an indictment of the whole model as you could get. Operators will not go to jail for you, so they must be unable to have enough data to get you in trouble instead.
Such overlay networks only address *some* of the routing issues though, not overall unreliability.
@corrupts_absolutely @ad_on_is For static sites, you may be able to use the web.archive.org to bypass that.
For dynamic sites, you're out of luck unfortunately.
@ad_on_is The truth is that the #internet should be considered useful as a WAN routing layer, but nothing should actually be using it directly as it is unreliable both because of malicious actors and because of gross neglect by many operators.
Support for #AsynchronousCommunication (which #Usenet demonstrated) is an essential property. #Anonymization, #P2P routing & key-addressing are also minimum requirements to hinder trivial censorship. #Mixnet operation is even better.
@ad_on_is The problem you're hitting is that the #clearnet / #Internet in general weren't adequately designed to handle malicious #infrastructure operators.
"The 'net interprets censorship as damage and routes around it" was a comment about #Usenet, a #federated / #P2P system with gossiped (https://en.wikipedia.org/wiki/Gossip_protocol) message exchange which wasn't particularly picky about its transport layer (indeed you could load a spool on a floppy and mail it), not the internet.
@supervent @nn4x This goodness, absolutely this.
Why is anyone still using the #clearnet for this?
#I2P is absolutely one of the better options.
@nyl Personally, I'd be going *out of my way* to disable information-leaking features like that.
#Proprietary suggestion & AI engine? No thanks.
@kde@floss.social @kde@lemmy.kde.social From the article it seems more of a #Qt win for now (though it does mention patches for many others), but in any case that's neat.
Now it all just needs sane ways to interface from #CommonLisp.
@Salty @sailor_sega_saturn @TinyTimmyTokyo Right up until the secret gets liberated and one gets buried in concrete for trying to hide it, anyway.
No guarantees of any utopia anyway though.
@sailor_sega_saturn @TinyTimmyTokyo Eh, no guarantee (or any reason to believe really) a simulation would be even focused in any way on humanity (no anthropocentrism needed).
Similarly for superintelligence, few reasons for it to care.
Cryogenics is a better bet and as you say it's quite unlikely unfortunately.