opqpo

Those kids should get a prestigious cybersecurity job instead of going to jail. They didn't harm anybody even though they had the power to do that.

Literally there is ZERO evidence in the report. It's all pathetic nonsense "ghostwriting" about a few accounts upvoting each other on some blogs. The company and others have every incentive to blame these pathetic "ghostwriting" nonsense on Russia to make profit and be relevant. They have no reason whatsover to not do that. Even mentioning that they have no clear evidence against Russia means bad for their business.

Do you have any evidence that FireEye’s report is accurate and not misleading? they are the ones that claimed it in the first place.

Also what do you think is more plausible for a western cybersecurity company's bottomline? make Russia innocent or make it guilty? especially when you got the backing of all western governments and media.

disagreeing is a pretty strong word. Nobody can disagree these days. I was flagged and labeled as a troll within a couple of minutes once I criticized a programming language. 30 years from now, people will remember this pathetic mob tyranny era when one couldn't even offend a programming language feelings!

Mediocrity, ugliness and verbosity is more well thought out than engineering.

Accessing nil pointers is well thought out. Not enforcing error checking is well thought out. Reflection and code generators instead of generics and traits is well thought out. Lack of conditional compilation is well thought out. No compilation with unused variables instead of warnings is well thought out. Doing your own sort, map, all essential array operations from scratch everytime is well thought out. Spamming err != nil instead of some operator like `?` is well thought out. iota instead of type-safe Enums is well thought out. No file-scope variables is well thought out. go mod is well thought out compared to npm and cargo.

I am not against forking or introducing modifications, etc.... You can fork it, change it for your use case as you wish, but just don't bundle my work, rebrand it and sell it as if it were yours. I meant my work to be free and I want the legal protection to keep my work free.

Never understood the obsession of OSI and their insisting to not give any protection for the authors. Authors whether individuals or companies should have the right to prevent their work from being rebranded or sold by others with minor or even major modification.

How many times did it happen that somebody or organization wrote a complex piece of software and published it and found that some companies just rebranded it and even sold their work with little to none added value while the original authors don't even have the right to complain because this is how open source works.

Something like BSL https://mariadb.com/bsl-faq-mariadb/ needs to get more reputation among solo developers and small companies with little to no funding that believe in the power of open source while believing also in having the right of earning a living from their own work.

Kubernetes is a godsend and it solves way more problems than it creates. But I just believe that this simple use case is an overkill. OP just wanted a remote DNS server that is addressable within the VPN. You don't really need Kubernetes or Kilo for that.

Never thought of that. But OP maybe liked tinkering with k3s and kilo. But I believe kilo's best use is for cluster to cluster connection. This use case is a bit artificial in my opinion.

Yes you can avoid that open source setup for a buggy and even slower userspace wireguard commercial implementation for only 10$/month per user

I believe that Kubernetes is artificially injected in your setup. You can just run your DNS server on the server and advertise it on the VPN address. You can still address it from anywhere in the VPN.

What does "trojans at ISPs" even mean? TLS works end-to-end and ISPs can do absolutely nothing to see the plaintext. It's unless the CAs at users-side are manually replaced with fake ones nothing can be done. I've never used Windows since I was a kid but I am sure this is pretty much impossible on Linux for example since adding CAs require root privilege.

I like how they are shamelessly trying to rebrand Jason Donenfeld's open source work in order to take the credit and the money for features with salesy names like "mobile IP" that were actually built inside WireGuard from day 1. In fact, I tried it a month ago and I had to uninstall it within the first 24 hours after I noticed the excessive logging eating my disk, the inexplicably high CPU usage, and low bandwidth compared to vanilla WireGuard, even pinging was failing randomly for some mysterious reason. WireGuard is a great piece of software, but Tailscale is a very bad, slow and buggy re-implementation of Zerotier over WireGuard.

Is there any way to create a bastion or proxy without messing with the client configs? something like a transparent proxy like nginx but talks SSH.

It may work for enterprises that want to search the web via an API. But for normal people? I extremely doubt it.

The only thing that is out of control is these MSM propagandists.

It is NOT. It's a paid commercial software where only the client is open source.

*was

That's why you should never start the Uber of X and the Amazon of Y. Because if the actual Uber and Amazon see that the market is good they will massacre you unless you're fortunate to get an offer from them.

What is their business model?

It's worse but still possible to swallow for me.

You can use it inside a browser. Look at https://github.com/grpc/grpc-web but I agree that something like playing with curl would be impossible. I once even stumbled upon a golang project that can generates a RESTful API from your gRPC API without any modifications on your server side.

I wish more people try gRPC outside of microservices. It is the perfect solution for building APIs.

What's wrong with using vanilla WireGuard? I tried this Tailscale recently and it was a huge mess. It destroyed WireGuard's throughput compared to the kernel implementation and the CPU usage was like I am running a game along with the excessive logging in systemd. One of the worst software I've ever installed on linux.

The certificate will still have to be verified at client-side with the CAs. I know my idea will break the entire TLS flow but I am just talking about the validity of it in the first place.

If we trust DNS to publish the SNI public keys, why don't we trust it for the website certificate in the first place?