Discussion Ethically Online

 
  • Like
Reactions: Nebulys
Speaking of phones, I guess Jolla is trying to get back in to the phone business. I suppose technically they never left because of the SailfishOS and the related community phones, but it's been a while since they had a "real" commercial phone in the market.


Quite curious about it.
If the camera is good and the phone has lengthy support, I could be into it.

RE: Motorola, I actually liked my previous Motorola phone quite a bit but the camera was a bit eh, well I guess more like their post processing of pictures was a bit eh.
And it only got like 1 full Android version update and maybe 3 years or security updates, which sucks. I think they've improved on that now, but I think they still only offer lengthy support for the most expensive phones.
But other than that I really liked the phone, it was as close to vanilla android as you can get these days with a few somewhat unique features too. Karate chop motion would turn on the torch! Coolest feature ever.
 
  • Toucan
Reactions: lashman

Coding Computer Science GIF by CC0 Studios
 
  • Toucan
Reactions: Nebulys and lashman
Using AI vibe coding to make your personal webpage, online CV or an application only you use is one thing.

Using it in an enterprise product with millions of users is another entirely.
 
  • This!
Reactions: lashman
It's a shame, but in their defence they are governed by Swiss law. If they say no, they get shut down.

How do you have an email provider not at the mercy of any state or government? It's impossible. They have to be accountable to someone.

I think the lesson here is if you have an anonymous Proton account and you want it to stay that way, you don't pay for it with a credit card. Proton take crypto, and will also take an envelope of money off you if you post it to them.
 
It's a shame, but in their defence they are governed by Swiss law. If they say no, they get shut down.

How do you have an email provider not at the mercy of any state or government? It's impossible. They have to be accountable to someone.

I think the lesson here is if you have an anonymous Proton account and you want it to stay that way, you don't pay for it with a credit card. Proton take crypto, and will also take an envelope of money off you if you post it to them.
The sad truth is that if the government wants your ass there's no way of stopping them.
 
  • Sad
Reactions: lashman
The sad truth is that if the government wants your ass there's no way of stopping them.
I think there's a cost of them getting your ass.

So you need to have that fine balance of:

1) Not pissing them off too much.

2) Being obscure and anonymous enough that it's not worth the effort of them chasing you.

If you get the balance between 1 or 2 wrong, they will get you.
 
  • Like
Reactions: NarohDethan
Regarding Proton situation, I think all mail providers needs to provide stuff if required by law, I follow tutamail on mastodon and they do the whole typical fediverse spiel where they constantly be shitting on others and saying how great they are with privacy, for then to say they ONLY provide email accounts to governments 25% of all requests. Doesn't sound very encrypted and private to me.
 
  • This!
  • Like
Reactions: Nebulys and lashman
Regarding Proton situation, I think all mail providers needs to provide stuff if required by law, I follow tutamail on mastodon and they do the whole typical fediverse spiel where they constantly be shitting on others and saying how great they are with privacy, for then to say they ONLY provide email accounts to governments 25% of all requests. Doesn't sound very encrypted and private to me.
There will be bits in the chain that can be more or less private.

If you store your passwords or 2FAs on Apple or Google or whatever, the feds can get them to hand that information over. It may be encrypted, but if they have your password (from a leak, for example) your whole stack is compromised.

As I said, in this case the person they caught was foolish enough to tie themselves to the account through a credit card. You can’t pay for something using a mainstream payment method and expect to keep anonymity.

And at the end of the day, there’s nothing to suggest the user’s email or data stored on the account has been breached, just that their identity was revealed through their payment method.

I still reckon Proton and Tuta are better than most other alternatives, but they’re only as secure as you are not sloppy.
 
Using AI vibe coding to make your personal webpage, online CV or an application only you use is one thing.

Using it in an enterprise product with millions of users is another entirely.
So I was thinking about this and decided I'd try and vibe code something.

I designed (and got Kagi Assistant using Claude to code) a notepad and calculator combo app I've called Figur. I haven't decided if I'm going to publish it yet. I may do if I can get around to it. Though I imagine the code in the backend is very embarrassing.

It runs in docker and lets me make documents (called 'sheets') that consist of calculations, variables and Markdown.

When a 'line' is added you can pick between text or calculation. If you choose text, you get one line to write Markdown. If you choose Calculation you get a table with three columns: VARIABLE | VALUE | CALCULATION.

The VARIABLE is the name of the calculation, the VALUE is the equation and the CALCULATION is the final figure that's stored in the variable.

You can pull the value from variables and use them to do further calculations (e.g. VAR1 + VAR2)

It can do regular numbers, decimals and currency (£ only for now, but if I do ever publish it I'll think of a way to include more).

Then I can use Markdown text and headings to break up the calculations and annotate it.

A helpful syntax guide on the right is there if you can't remember what the right markdown or number operators are. You can close this if it's annoying you.

Figur.webp

It supports multiple sheets (the left menu lets me add and delete them).

It also works splendidly on mobile and can be used as a PWA.

And lastly, it saves each sheet on the server it's running on to a folder called /saves/ in JSON format, so if you host it on your server and hide it behind a reverse proxy or Cloudflare tunnel (complete with authentication on the front end) you can access it on any device, anywhere in the world and edit your sheets.

When I switch to Linux and Android it's going to replace Soulver, which admittedly is a far more complicated and accomplished app, but I severely underutilise it and this app basically covers all the things I used it for.

I'm pretty happy with the outcome, for what is basically an afternoon of writing up my needs, bug testing and iterating based on things I hadn't thought of when I first prompted the AI.
 
Last edited:
How are you all so good with getting AI to do so nice things, I have to always wrestle with them so much to get something usable lol
It can depend on the model, and how you communicate what you want the app to do.

If you want it to look a specific way, you need to be hooking Claude Code up to Figma. Now that's some black magic.

Cant wait to hear your impressions with the OS!
So far the only incompatible app that I'll miss is eBay. It won't even download. Otherwise, I'm in the process of logging in now and it's been smooth sailing so far.

I was worried GrapheneOS would be a nightmare, but it's actually incredibly mature and usable. I'm sure I'll encounter issues, but I haven't yet.
 
  • Like
Reactions: lashman
It can depend on the model, and how you communicate what you want the app to do.

If you want it to look a specific way, you need to be hooking Claude Code up to Figma. Now that's some black magic.
Ive been sticking with the free model of chatgpt lol, but it suits me, mostly I just need to get some direction and help but not to actually do what Im trying to build. And yeah bug hunting of course
 
  • Like
Reactions: Nebulys


It feels like Proton are being intentionally misleading in their statements. They know that most of their customers aren't familiar with how legal process actually works, so are happy to spread half-truths.

Under US law, a US law enforcement agency (LEA) typically has to apply for a subpoena or search warrant with a US court. The court is then responsible for deciding if the legal bar for search a request has been met, then either grants or denies it.

The problem is, if a company has no real US footprint (no US corporate entity, offices, servers, etc.), then a US court typically doesn't have the jurisdiction to compel the company to hand over customer data (except in some rare circumstances). Even if the court approved the warrant anyway, it wouldn't really be legally binding.

Which is why the Mutual Legal Assistance Treaty (MLAT) exists. MLAT enables law enforcement agencies in one company to send requests for information to law enforcement agencies in another. Switzerland has such a treaty with the US. This means that the FBI can request that Swiss authorities hand over a Swiss company's data on their behalf.

Any country requesting information held by a company in a foreign jurisdiction would typically do so via MLAT. Which means from Proton's perspective, the legal request would appear to originate from their local law enforcement, not the FBI. Which they clearly understand based on their Reddit post.

Saying "we don't respond to legal requests from anywhere other than Swiss authorities" seems very intentionally worded to give the impression that the company does not cooperate with foreign law enforcement. But since it'd be the Swiss authorities handling any such requests, they'd have to comply, since as they admitted, they have to comply with local laws.

There is, however, some useful (but more nuanced) information here:

Firstly, MLAT requests are handled by local law enforcement according to local law. So if there is a difference between the law of the sending and recipient country, that might mean the MLAT request is denied. That probably doesn't mean much, because if you're on the FBI's radar, the chances are you did something that is also massively illegal in Switzerland too.

Secondly, they are 100% correct in saying that no other service provider is going to do any better. They're all beholden to local laws, and the ones that think they're not tend to get their doors blown off by SWAT like CyberBunker did. The only exception is if the company resides in a country which does not cooperate with US law enforcement (which Proton does not).

But the part that's extremely disingenuous is that the "we only respond to requests from the Swiss authorities". That statement is likely intended to imply they don't cooperate with law enforcement in any other countries, which is simply not true. Switzerland has MLAT agreements with over 30 counties.

People really need to understand that no company is going to shield you from the FBI (or any reputable law enforcement agency). They'll use misleading statements to make it sounds like they don't cooperate with law enforcement, but they do. They have to.
 
How are you all so good with getting AI to do so nice things, I have to always wrestle with them so much to get something usable lol
As a software engineer that integrates AI and my hobby is AI, every so often I earnestly try 'vibe coding' on different levels, because the place I work also uses LLMs for a lot of uses, and actually where I'm work they're preparing for a bigger push to get more devs into agentic coding, including tying in an agentic IDE workflow with an internal RAG search system.

My experience lately (last few weeks) is that the current state of the art models can agentically makes code that will compile and usually work ok, but I've witnessed (and spent a long time cleaning up) agents ripping out performant code and replacing it with objectively worse code that an intern would understand was unnecessary to remove, create inefficient code. Since it's agentic and we have thorough testing it doesn't output broken code, but first try even the top models will output some nonsense syntax that won't compile...

For some work items it's worse than nothing, and in the best case it's only a little faster than doing it by hand. Worst cases, I've done some sunk cost stuff where I vibe coded some new functionality and in the end it really did take longer to clean up the mess than if I'd just done it by hand. This is all within the last month.


I don't think the coding skill of the models has actually improved much at all in the last 6 months, just the agentic flows are a lot better, the thinking component is better, and therefore you can fire and forget and get something that you might be able to forget about, but you shouldn't.

My 2c
 
  • Like
Reactions: Nebulys and lashman
As a software engineer that integrates AI and my hobby is AI, every so often I earnestly try 'vibe coding' on different levels, because the place I work also uses LLMs for a lot of uses, and actually where I'm work they're preparing for a bigger push to get more devs into agentic coding, including tying in an agentic IDE workflow with an internal RAG search system.

My experience lately (last few weeks) is that the current state of the art models can agentically makes code that will compile and usually work ok, but I've witnessed (and spent a long time cleaning up) agents ripping out performant code and replacing it with objectively worse code that an intern would understand was unnecessary to remove, create inefficient code. Since it's agentic and we have thorough testing it doesn't output broken code, but first try even the top models will output some nonsense syntax that won't compile...

For some work items it's worse than nothing, and in the best case it's only a little faster than doing it by hand. Worst cases, I've done some sunk cost stuff where I vibe coded some new functionality and in the end it really did take longer to clean up the mess than if I'd just done it by hand. This is all within the last month.


I don't think the coding skill of the models has actually improved much at all in the last 6 months, just the agentic flows are a lot better, the thinking component is better, and therefore you can fire and forget and get something that you might be able to forget about, but you shouldn't.

My 2c
Yeah Ive noticed how bad it can get with code. For me since I dont really know coding it kind of is essential for bigger concept coding things. I cant imagine it would have gone faster learning to code properly. Also I do learn while doing it like this. I guess in that sense its good I dont have more advanced AI doing more complete and more usable stuff. But yeah in those moments where one doesnt even know where to begin in a project it can be nice to have some guidance and examples to start with.
 
  • Like
Reactions: lashman