you can get a mini pc for about 1000 dollars that you can SSH into for claude sessions

Why does the western world care about Chinese CPUs, let them play in their sandbox with their own systems

we are working on it, the thousands of gig workers tuning frontier models

You need to spell and grammar check your writing

lemme guess, you are anti-israel as well

the vast majority of things are on you, if you had a working phone, it would just be install an app, login, and it works

it is better than having to carry keys around

For an office door, why not a key card or lanyard badge or whatever? A code? I feel like you could use something easier than an app too.

You're in a rush in the morning and you leave home without your badge.

Happens to me at least once a month. But I almost never don't have my phone.

Key cards/badges don’t have GPS in them and can’t keep tabs on you outside the office.

it's sold as harder to copy (and easier to revoke when you're doing layoffs - which shouldn't have been a really challenge for card-based systems, they're just old and didn't keep up.)

Neither of these are challenges at all for semi-modern card systems.

Yeah, that article was silly. Yes, setting up a new phone is a giant annoying exercise in yak shaving, but that’s to be expected when transferring to any device.

That said, I’ve had to unlock a door with a dedicated phone app, and it’s always way too many taps even if the app hasn’t randomly logged you out because your session is a month old or because your phone restarted to store an update. (Even more if you prefer to keep Bluetooth off -which among other things saves on battery.)

how does that solution work if the table that is dropped has foreign key constraints?

That's why I said:

> disabling foreign key checks on large operations

And you have to know that, according to your business logic, what you're doing is safe.

Did your symptoms ever improve?

what happens if you suffer catastrophic failure, you'd lose up to one hour of data?

The system had RAID, so disks could be replaced without taking anything down.

We never had a catastrophic failure, but yeah, and hour of lost data wouldn't be the end of the world. For regular maintenance we would just coordinate with the QA team to pause testing temporarily. (Testing was only around-the-clock near a product release.)