aborsy

The battery powered device is a client to a remote exit node. Normal VPN basically

I would like less battery usage with exit node.

Yes , you are supposed to do. That doesn’t change the fact that he has no insight into technology, and is a bandwagon person.

If I’m a mathematician, I can write public posts on math. Extensive posts about politics, AI, crypto, … are not useful without expertise in those domains

Mr. Tao is an excellent politician. Lots of awards and texts, yet no major problem solved.

It seems now that NS is solved he is mobilizing the community to convince taxpayers continue to pay even though AI may do a better job in his work.

Also, his opinion of AI has continually changed in the past years, after the capabilities were demonstrated.

You should compare battery consumption of TS exit node with something like WG app which low

Agree!

Questions: can new research like this be done using publicly available models?

Or will access to internal frontier models provide a big boost?

The best backup solution that I found is ZFS send (even if it’s not strictly a back up tool).

ZFS send raw encrypted stream from laptop to a server automatically. Encrypted compressed incremental. No need to verify: if snapshot exists at destination, there is no error. If a problem is encountered, restore from redundancy. Errors can be detected and corrected.

Restic has also been working great. The problem is that, it has no redundancy to correct errors. The assumption is that, server holding repository will have redundancy, but then I can replicate directly to that server. I worry that at some point, there will be an error in repository. I may loose 5 years of snapshots (restic has some functionality to remove involved snapshots and rebuild the index, but may not succeed).

Kopia has error correction, with Reed Solomon.

Borg2 has perpetually remained in beta. Looking forward to test it.

I stay away from sync: problems with permissions, git repo, conflict, incomplete sync status in syncthing etc.

Disable password authentication and ignore the noise.

Find, fzf, vim, ssh, …

SFTP is just simpler and better, if available.

Thanks!

GNOME Remote Login runs an RDP server, and RDP indeed isn’t fast enough, although is TLS-encrypted.

I have used Remmina over SSH and Tailscale (and generally don’t trust remote desktop tools), which is mentioned in various places by my operating system provider.

How does RustDesk compare with that?

A limited form of it with different syntax comes with Docker Desktop. The sbx tool is not available for non-Ubuntu distributions.

Docker supports unprivileged containers too.

The agent is per installed. If there is an update to the agent, the VM notifies you on a first run and updates it if you approve.

Other than the login problem, it’s a decent option.

The problem with oral exam is that it’s done serially student by student, which can be infeasible with large classes. The grading is also highly subjective.

Written exams are also not best suited to project-based classes.

What’s the best way to grade students in classes with coding projects?

Nobel was awarded due to project management work. There are many thousands of scientists with work much much better than what Demis has done.

I hope so!

No. If software is small enough, it can be proven.

There is formal verification of the protocol and aspects of code:

https://www.wireguard.com/formal-verification/

The code is small enough that can be reviewed.

It does too many things, and the product has got too complex. I saw a year ago they were looking for someone just to help with complexity.

I use it but feel uncomfortable, that it has large attack surface and LLMs will find exploits in it.

Without taillock it makes no sense. Anyone on their coordination servers will be able to connect to your network.

France has a similar program called CIFRE and it is bad.

An academic pairs with a buddy in industry. This relationship usually remains fixed in their careers. The representative in the company hires an engineer, makes a CIFRE contract and asks the academic to register the student and takes care of paperwork. There is no open call, competitive submission of proposals, a committee and a process. I don’t know how government is OK with this. The company sends the information about the recruitment to the government and gets tax waiver. Therefore, the company hires an engineer paid by taxpayers, nearly free for the company. A small amount of money is sent to the academic for conference travel etc. In many cases, student remains in the company and there is no need for academic other than for periodic paperwork. Importantly, the student works on tasks assigned by the company, which in many cases is not really research. It could be product improvement, mechanical work, client work, or some R&D or short term research. It’s not science or long term research, and no classes are involved. Next year, next student, it will be the same pair.

The program attracts students because they want jobs (in many cases their contracts is extended after 3 years of PhD, and they are actually incentivized to work well during that 3 years and be a good employee). Companies like it because of essentially free engineers and government rapport. Academics are content because they can say they have a student (even if they don’t see the student much if at all sometimes), cite collaboration with industry, and get some money for travel. The academic cannot complain, or require student spend more time in university, because the company would say, we will replace you next time with another academic who will be happy to sign off to paperwork and not to cause issues. The three parties are happy, and the taxpayer looses, because of a badly set up program. Maybe the American version will work, not the French version!

The impact on job satisfaction for those with jobs will be negative too. People will be trapped in their positions, fearing to leave or move around.

A new breed of academics has appeared whose jobs is to put their names in every paper possible. Literally, their job is to work on frameworks to buy co-authorship.

They do this in various ways, like establishing paper pipelines, collecting rents on labs and committees, focusing on money layer, using their profiles and citation count to help with acceptance of papers of other people , etc. You talk to them and they can’t explain their papers beyond a superficial introduction.

They collect huge citations, travel and give talk on the winner horses, collect credit, which feeds back into this fraudulent scheme. A scientist used to be a scientist not long ago, not a credit collector.

I wonder if Google could invent a new metric to expose them (weak ratio of first authorship, etc).

Docker sandboxes do that. They take quite bit of resources though, almost same as a standard VM.

I need to mount directories or remove after creation easily, like I do with containers.

Gateway would be good too.

Firewall needs more functionality.

Yeah, for non-X86 devices, getting to U-boot with pressing a combination of PINs in particular order and conditions and releasing at right time is a pain.

I think I wasted $100,000 in salary for $100 more in device cost, in setting up an OpenWRT router.

Apart from installation and upgrades, the OS itself is nice, very flexible and capable.

Indeed, I was referring to Attended Sysupgrade in luci or CLI.

How about OPNSense on open hardware of your choice, and passing messy wireless to separate AP?

OpenWRT is very good, but the installation and upgrades are not easy. There is a zoo of images for different hardware, installation options and tools. It has to run on small devices, so there are limitations. The documentation on Wiki is scattered and could be improved.

I had to search forums for weeks for a custom package installation for my router. Right now I have been trying to upgrade to the latest version via LUCI for a while, and it stucks. Probably have to wait for few weeks, go through CLI and maybe search forums again.

I just thought I am paying a hefty time price for a bit more expensive x86 mini pc and AP.

Docker has introduced sandboxes for this purpose.

Terrible website, filled with advertisements and constant pop ups.