Hacker Newsnew | past | comments | ask | show | jobs | submit | fpoling's commentslogin

They use the datasenter for model training, not to serve online users. Presumably even if it will be offline for a week or even a month it will not be a total disaster as long as they have, for example, offsite tape backups.

You can setup a separated account with a long password on MacOS and remove your user account from accounts that can unlock FileVault. Then you can change your account to use a short password. You can also change various settings regarding how long Mac has to sleep before requiring to unlock FileVault.

I didn’t understand how a user that cannot unlock FileVault helps. Can you please elaborate on this setup? Thanks.

With that setup on boot or after a long sleep one first must log in into an account with longer password. Then one logs out of that and switches to the primary account with a short password.

If one kilogram of stuff consumes just 100Wt, then in one month it consumes about 300 MJ. So as long as things works for a year or more energy cost to put them into orbit becomes irrelevant.

To keep things in orbit ion thrusters work nicely and require just inert gases to keep them functioning. Plus on a low Earth orbit there are suggestions that a ramjet that capture few atoms of atmosphere and accelerates them could work.

Radiative cooling scales by 4th power temperature. So if one can design electronics to run at, say, 100 C, then calling would be much less problematic.

But radiation is the real problem. Dealing with that would require entirely different architecture/design.


But they hosted the repo on Microsoft-run GitHub ...

The public-facing mirror :-)

I have found that using Cursor to write in Rust what I previously would write as a shell or Python or jq script was rather helpful.

The datasets are big and having the scripts written in the performant language to process them saves non-trivial amounts of time, like waiting just 10 minutes versus an hour.

Initial code style in the scripts was rather ugly with a lot of repeated code. But with enough prompting that I reuse the generated code became sufficiently readable and reasonable to quickly check that it is indeed doing what was required and can be manually altered.

But prompting it to do non-trivial changes to existing code base was a time sink. It took too much time to explain/correct the output. And critically the prompts cannot be reused.


Same though lately discovered some rough edges in rust with LLM. Sticking a working app into a from scratch container image seems particularly problematic even if you give it the hint that it needs to static link

I have lived in Spain for the last two years and observed the luck of maintenance in a lot of things.

For example, people typically pay for house/apartment insurance. But insurance companies never send a person to check for things like leaking pipes or whatever. Rather they simply wait until an accident happens and dispatch an emergency crew and cover a lot of damage that could be easily prevented. Then people tolerate non-trivial damage to homes/apartments like leaky roof not reporting it to insurance companies for weeks.

Then with cars people often do not follow the maintenance schedule and insurance companies do not ask for that. Typically people drive until damage happens due to a minor accident or maintenance are forced by state required technical inspection once in few years. The car companies even offer free maintenance checks as a part of guarantee but people skip even that.

Yet when someone spends efforts to complain, thinks do gets done. For example there a city service to remove graffiti on public areas. If one files a complain, they react and remove the graffiti. However sometimes one needs to send a complain twice.


I think you are describing how the entire world works. I have lived in 3 western European countries through my life, and they all work this way.

Never I had the pipes in my home inspected, even now that I live in areas where it freezes regularly.

Never has anyone (not even my insurance) forced me to follow any particular maintenance schedule (albeit I'm quite sure somewhere in the fine print it will read that if the accident is because of poor maintenance they'll just ignore the claim).

Here the city service to remove Graffiti is almost overnight, and works better than many other public services...


I lived in a building in Norway that had phone lines from like 1924. As it was declared as historic it was very difficult to dig newer cables. So ISP tried to use the old cables. At best one can get 10 MBit/s with a lot of packet loss. Eventually everyone switched to mobile internet. With 5g and stationary antenna it was possible to get 300 MBit/s with no packet loss and sub 10ms ping to local servers.


With Bitlocker it is still possible to have single password-based key. But enabling that requires to enter a few commands on the command line.


And you can be sure it didn’t add a ‘recovery’ key, how?


Using the same CLI, which shows all the alternative "protectors".


Again, that is a lot of trust since it could trivially just… not show it. Which is already the default for most FDE systems for intermediate/system managed keys.


It could also just pretend to encrypt your drive with a null key and not do anything, either.

You need some implicit trust in a system to use it. And at worst, you can probably reverse engineer the (unencrypted) BitLocker metadata that preboot authentication reads.


No, that would be trivial to verify with any other operating system.

Key ring contents (and what is done with them) are typically much harder to verify as they’re encrypted.


It requires the Pro edition of Windows too.


There are stories about bodies of Christian monks that did not decompose for a log time after the dearth. Modern take on it attributes it to the climate in caves where the body was put after the dearth. But another important part was diet. Often the well-preserved bodies were of those who had eaten only rough bread and water for months and years before the dearth.

So I suspect both of the factors are at the play here as well.


I read an article about this years ago and came away with the impression that they were effectively mummifying themselves before death. From the Wikipedia page on sokushinbutsu:

In medieval Japan, this tradition developed a process for sokushinbutsu, which a monk completed over about 3,000 days. It involved a strict diet called mokujiki (literally, 'eating a tree'). The monk abstained from any cereals and relied on pine needles, resins, and seeds found in the mountains, which would eliminate all fat in the body. Increasing rates of fasting and meditation would lead to starvation. The monks would slowly reduce then stop liquid intake, thus dehydrating the body and shrinking all organs.

https://en.wikipedia.org/wiki/Sokushinbutsu


I think you’re referring to the phenomenon named “incorruptibility”:

https://en.wikipedia.org/wiki/Incorruptibility

Recent case, died 2019, found not decomposed in 2023, had not been embalmed, unsealed wood coffin:

https://www.ncregister.com/cna/incorrupt-body-of-sister-wilh...


In "The Brothers Karamazov" by Fyodor Dostoevsky describes sort-of the opposite process. A body of monk who had been considered a living saint started to smell on the second day after his dearth. That made everybody to realize that he was the opposite of the saint. And then another monk mentioned that eating sweets and cakes could not lead to sainthood.


The peasant child died and didn't stink though. So, I guess he didn't eat a lot sweets I suppose that aligns with his wretched upbringing...


Assuming no supernatural forces, can this sweets adversion be explained by the fat vs muscle ratio in the body?


Also related: Mystery of the Tibetan Mummy (https://www.youtube.com/watch?v=gCkB9daCVZ0)


Even Wikipedia article [1] on the rotation curves mentioned that those could be explained if one stops using Newtonian mechanics and rather try to use the full equations of General Relativity. Basically it turned out on the galaxy scale one could not use Newtonian approximation. GR equations are highly non-linear and even tiny non-Newtonian effects accumulates.

Then there are various articles pointing out that taking into account that galaxies are filled with plasma can also explain observed results.

[1] - https://en.wikipedia.org/wiki/Galaxy_rotation_curve


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: