douglasg14b @ douglasg14b @lemmy.world

Posts

11

Comments

930

Joined

2 yr. ago

The sad part is is that you're right.

And the reason that it's sad is that most of the individual veneers on proprietary projects deeply about a project itself and have the same goals as they do with open source software, which is just to make something that's useful and do cool shit.

Yep, the business itself can force them not take care of problems or force them to go in directions that are counter to their core motivations.

Ever is this data dump so I can mirror it?

Yep, just like electron or Tauri. A web view wrapped in a native application.

These are very common these days, it's the same use case and value proposition. Mainly because it's just easier to develop UIs with web technologies that look the same everywhere, never without the app.

You do know that a pwa can be packaged up in an app container and you won't even be able to tell the difference?

It doesn't actually have to operate like a pwa, and require native pwa sport.

There are tons of apps that you use that are just well packaged PWAs, packaged as an app store app, and you don't even know about it.

PWAs only suck on when they suck, just like everything else.

Did I say that it did?

No?

Then why the rhetorical question for something that I never stated?

Now that we're past that, I'm not sure if I think it's okay, but I at least recognize that it's normalized within society. And has been for like 70+ years now. The problem happens with how the data is used, and particularly abused.

If you walk into my store, you expect that I am monitoring you. You expect that you are on camera and that your shopping patterns, like all foot traffic, are probably being analyzed and aggregated. What you buy is tracked, at least in aggregate, by default really, that's just volume tracking and prediction.

Suffice to say that broad customer behavior analysis has been a thing for a couple generations now, at least.

When you go to a website, why would you think that it is not keeping track of where you go and what you click on in the same manner?

Now that I've stated that I do want to say that the real problems that we experience come in with how this data is misused out of what it's scope should be. And that we should have strong regulatory agencies forcing compliance of how this data is used and enforcing the right to privacy for people that want it removed.

Oh, you get the benefit of explicit scanning?

We get the beauty of every file that's modified being scanned before the write "completes". It's an absolute joy starting a build and watching ~80% of the available compute be consumed by antivirus software.

Or, you know, normal filesystem caching as part of your tool's workflow.

Or dependency installing and unpacking....

Or anything actually that touches a lot of files.

No, no they are not.

Bad ones? Yeah, just like that.

I got the feeling that these replies are written by ChatGPT?

I build software and can confirm this.

This is pretty run-of-the-mill analytics and user session recording. There's nothing surprising here.

Usually it's not actual screen recording but rather user action diff recording (Which effectively acts like recording the application except that it only records things that changed so that the recording is much cheaper to store)

This is extremely effective for tracking down bugs, solving user support issues with software, or watching session recordings to figure out if users are using the software in unexpected ways.

Pretty much any error tracking analytic software worth it's salt does that these days!

Proof that it's already too late ☝️

It kind of is though?

Essentially no protections for operators from bots, which means no protection for communities from automated infiltration and astroturfing.

The only thing stopping it is that the cost/benefit isn't there yet with how small Lemmy is. But that's slowly changing.

A strategy you could employ is to buy the battery early.

Keep it at about 30-50% change, and refrigerated. You're talking < 2-3% capacity loss per year, probably much lower since the temperature is so low.

Meaning the battery is effectively new when you go to swap it.

Parsing commas is hard, right? No, not really.

• young
• poor
• women

Parsing commas is hard, clearly, since you still missed the lack of parallel structure. “Young, poor, women” reads as a list of adjectives modifying a single noun, "women".

🤦

Young poor women are the primary recipient of Medicaid?

Found the Chrome user

Using a new IDE is always a painful undertaking TBF.

I switch from visual studio to rider in order to better support my co-workers on macs. And I have never looked back, it's just too damn good.

Though, the settings for exceptions and when to break are never right for me. While VS has it right, right out of the box.

Yeah, location based as in you and the objects in the game are based on real world coordinates. The "grid" for the game is overlayed onto the real world

Same ingress lost its appeal after a while. The gameplay loop was shallow and repetitive. It was based around rather fast gameplay loops, that would resolve, and then you rinse and repeat.

I made some cool friends though, it was cool to meet people at capture points.

I'm aiming for much MUCH more depth here. Fundamentally different from ingress of similar games, aside from being location based. More industry and exploration, with a more typical loop around economy, growth, and advancement.

Great question, and one I've struggled with.

I'm a big privacy advocate, and my personal devices and home network reflect that. Which really brings me to a difficult crossroads here.

I don't have a good answer for you right now, the best I have are the problems I'm trying to balance:

• Anticheat: How do detect and build better detection for location spoofing? This, intrinsically, requires the recording of directly associated location data. And the collection of mass anonymized data in order to determine "what looks normal", to spot abnormal (spoofing) behavior. How can I balance this against privacy concerns? It's a rough one for sure.
  • This is the toughest one here. Likely I'll need a combination of data retention periods and anonymization. At the very least sensitive data is separated from the rest of the game data, and is encrypted at rest. Likely there are clever protocols and solutions already out there I just don't know about yet that can improve protections here.
• Audit Logs: When a player performs an action that interacts with a location-based feature, where they where when that action was performed it is stored alongside the audit log of that action. This ties in closely with Anticheat, and also enables pattern matching to try and find oddities (exploits, cheating, bugs, and other problems).
  • Right now these stay around forever, and can be used to simulate the global game state at any point in the past (really REALLY useful for debugging problems, especially when you don't have a good repro). Eventually such state should make granular rollbacks possible in case of exploits or rampant cheating. (A game where you have to physically go somewhere to capture a mine means rollbacks have a crazy high cost, making them granular is pretty important)
• Analytics and Telemetry: Location data isn't in use here right now. And I don't see how it would be while also respecting privacy.

Selling the data: 😂😂😂 I'd rather light my servers on fire than stoop to that level.