konha

I don’t know. Depending on how much time passes between now and the moment you try to recover the key I bet at least one of your friends will have misplaced or lost that piece of paper.

The free version is a loss leader to build mindshare amongst people who are (later) in a position to buy licenses.

> It's not rare, it happens constantly in enterprise software, project managemment software, anything where you have collaboration

The number of revoked tokens compared to all active tokens should still be tiny in those systems, wouldn’t you agree?

> Everyone else wants immediate revocation of rights, not waiting for a token to expire.

With a revocation list you can still have that. Once you propagated your revocation to all relying parties the token effectively expires early.

So? Most of your code is worthless to anyone except yourself.

Yes. GDPR allows for delays when complying with deletion requests. You should ideally document it and factor the delay into any deadlines you might be bound to.

You’d need to make sure the process is somewhat predictable, like running the vacuum on a set schedule so you know for sure what maximum amount of time a deletion request will take.

Arq works with file system level snapshots. So I guess at one point all the files have to be present for the backup work. Not ideal for scenarios where the cloud data exceeds the local storage.

Seconded. Check out both books by the author, they’re great!

https://lets-go.alexedwards.net/ https://lets-go-further.alexedwards.net/

Just a note: Make sure whatever you use to backup iCloud can handle files that are not currently synced to your device. Arq (https://www.arqbackup.com) has a setting to "materialize" these files before backup for example.

> What’s the barrier to entry for starting up a similar business?

The merchant of record model.

Dragnet searches are controversial in many societies, not just the US.

There’s a balance between individuals rights to privacy and what makes law enforcement easier.

One argument against broad surveillance measures like this one is that surveillance infrastructure is easy to implement and hard to get rid of again. You might be fine with the laws that are enforced with it today, but you might not be with what it’s used for in the future.

From the article:

> Emergency legislation to exonerate wrongfully convicted Post Office workers has completed its journey through Scottish Parliament and each will now receive initial compensation of £600,000, with the ability to claim more as financial redress for their suffering.

There is a chrome extension for everything. OPs point is (and I agree) that these things you only need once in a while are incredibly useful when they are only a search away. Installing a chrome extension for a few uses every other week is way too much effort.

They can be. Depends on how the are implemented.

Passkeys can:

- Replace the whole login (including discovery of the user id)

- Just replace the password, after a user specified a user id

- Be used as a second factor just like TOTP

They are definitely more phishing resistant for what it’s worth, even if just used for MFA. TOTP codes can be copied manually by an unsuspecting user.

It is. “Vorratsgesellschaft”

Depending on how openly and where this happens you might be able to get the platform they are selling your software on or their payment processor to take them down.

That's for legitimate platforms and processors of course. If it's on some darknet bulletin board - tough luck. Not worth the time.

Goes both ways actually.

I cannot count the times someone felt the need to bend over backwards to accommodate me because of something I didn’t want to do/eat/drink/whatever when I‘d been perfectly fine without any special treatment and moving on with whatever we were doing.

> How do they merge?

Manually. When you merge your code back into a common branch you look at the existing migrations and the new migrations and make sure the numbers make sense (after the merge). You can avoid stepping on each other’s toes by using a timestamp for the number in your filename. You still have to make sure (when you merge) the order makes sense.

> How does the migration tool know which migration to apply first?

Migrations are ordered by the (ascending) number in the filename. (Or some other scheme, but this is common)

> But having to check the latest migration and check that with my local db seems a little bit error-prone and cumbersome, no?

The number in the database should never be higher than the highest numbered file in source control. So you can increment without looking.

> How are teams of size >2 handling this?

Directory with .sql files starting with a number. Each file contains a set of changes to the db. The db has a table with the number that was applied last. To migrate your db you check if you have a file with a number that is higher than the one in the db. Then you apply that file to your db. That’s it.

Sounds like you are working in a way that is not intended by your tool / framework.

Why not use ssh certificates at that point?

Right. But for validation they can still be useful if you need a way to prevent (or at least hinder) users to create lots of accounts.

Passkeys have a fallback flow where they show a qr code you can scan from the screen of the device you want to log in to. Requires bluetooth though to prove you are "near" the device. I guess that's also disabled on these hypothetical locked down devices?

> Copy and paste and externally using any devices to connect with the system is prohibited.

Like a keyboard? In which case: How do you enter your password?

Is that a common scenario? Where you can use a (presumably) usb attached keyboard but you cannot insert your security key?

It seems to me people’s expectations vary a lot.

For me ChatGPT has replaced a lot of Google searches and skimming stackoverflow just to get some details for a small obstacle in my workflow.

Even if answers are not a hundred percent correct all of the time, it saves so much time to get to “good enough”.

I use it to generate shell scripts, SQL queries and general boilerplate like configuration files.

A couple of days ago I went from “I’d like to have a local DNS server running to test one aspect of my code” to a running docker container of CoreDNS with the configuration I needed for my tests in a matter of minutes. The same task would have taken forever had I googled for the information myself and stitched together snippets from the docs until it worked the way I wanted.

I run into the message limit occasionally, but 50 messages per 3 hours is more than enough most of the time.

> Also in Germany, if you change your name voluntarily […], this is considered to be a strong sign that you deeply hate your parents

Huh? I’m German and I have never heard of this.

Not sure if you are being serious, but that is terribly bad advice.

For starters google “piercing the corporate veil”.

So? Don‘t know what kind of homework you‘re thinking about. But using Google to do research doesn’t exactly sound like a problem to me.

Have students work with ChatGPT and let them spot / mark where the AI is hallucinating or wrong? Might translate well to general competence wrt working with any kind of source online.

That sounds incredibly backwards to me.

How about accepting new realities and changing homework assignments instead?

> Should be GDPR compliant since they don't use cookies or local storage.

That’s not how it works. If there’s personal data being transferred to the US, you are in violation according to the Schrems II ruling. If you only collect non-PII, you should be fine. Make sure though that your definition of PII matches the regulator‘s definition.

That’s how it works in Germany. Consumers see final prices and an itemized invoice shows how much tax is included in the price. Advertising prices to consumers that do not include taxes is against consumer protection laws over here.