Amju Wolf @ amju_wolf @pawb.social

Posts

3

Comments

397

Joined

2 yr. ago

...with a fugly hole in the display no less.

If you are willing to host something yourself you might as well selfhost Nextcloud and use KeePass(XC) with it.

There are other ways your password database could leak. For example you could use a weak password, or it could leak in some way, and if you store it on a cloud service that also got compromised you'd be fucked without a compromised device.

But yeah, all these are much less likely.

You are only at a 1FA level if someone hacked your PW-Manager but in that instance you’re most likely fucked anyway

As long as you at least have actual, separate 2FA for access to your recovery email(s) you should be more or less fine.

Unless you mean that if your password manager is compromised it probably means that your device is compromised, which also means that you're probably also a victim to a session hijack for the recovery email(s), in which case you are truly fucked.

You can also have a multi-level approach where for "higher value" accounts you have a separate password database so the more valuable accounts aren't exposed as much as everything else... There are definitely options.

The “value metric” that pisses me off the most is per user pricing when the service doesn’t incur costs per user.

Even in cases where there is a cost per user (or there is at least a correlation in cost increase with number of users) the price is usually many orders of magnitude larger than the cost increase.

Nowadays many services just force MFA on you in some way, and stuff like SMS or email verification is shitty, insecure and inconvenient. TOTP is then the next best thing, and having it integrated with a password manager is fine as long as you are aware of the risks.

Elite's biggest issue is that it never really knew what kind of kame it wanted to be. An MMO? There aren't enough multiplayer features for that. A (mostly) single-player space experience? It's too shallow with no story, so it won't satisfy the RPG fans. A space "simulator" where you just have fun flying ships? It's probably closest to that, except you can't fly any ship you want, and in fact it takes dozens of hours of grind to be able to switch things out so they're fresh and you have more fun with the game again. And the simulation is very simplistic and not all that fun either, so it's not for hardcore simulation fans either.

And because of this approach it has a bad combination of features that not only won't fully satisfy either of the potential target groups; they also often work against each other. For example the multiplayer component is a dealbraker for me: I want a truly SP game where I can dictate how I play it - where I can mod it, or at least use cheats to find my own pacing, fly different ships on a whim, whatever. But the game simply won't allow that.

But it's also not a fully-fledged MMO where you could build (or at least own) systems/planets/bases whatever with your clan and compete against others for ... idk, something.

And, again, it's just not a story game that you could play from start to finish for the storytelling and worlbuilding.

Really sad, because the potential is there to have any (or perhaps even at least two) of those types of games.

That's not an unsolvable issue, and you can always handwave it away for simplicity with some lore. The ships are already magics, like any star ship, so you can just say that the motors and calibration compensate for different planets and whatnot so the ship is easy to use everywhere.

A loading screen lets you load different areas of the game discreetly and make the game performative. This is especially important as Starfield is a single player game, it’s not hosted on a server or anything so it can’t distribute resource load that way, its all happening client side on the player’s system. They would have to simulate the entire world on their PC alone or develop a way to stream the content out dynamically and seamlessly.

That's not how any of it works.

We have had level streaming in Unreal for like a decade. Sure it's more complex to do things this way, but in general the way it works is that when you approach some area (are some distance from a planet or part of a planet) the next chunk of the world loads in, together with any NPCs and logic and everything else - it's basically a self contained map, just seamlessly integrated with other maps. There is no meaningful performance hit if done correctly. You certainly don't simulate everything all the time.

Additionally, all the other games mentioned (NMS, Elite, Star Citizen) also have basically all of the processing on the client side. The servers don't help the clients in any way; they only store primitive states for gameplay purposes, but all the simulation and whatnot is done on the client. And they still manage to be better optimized.

Elite is still just cleverly hidden/styled loading screens. No Mans Sky (and apparently Star Citizen but I haven't played it) is even better and more seamless.

Which, as I understand it, is kinda the point of the bills too. As in, if there is documentation and it's reasonably easy to dis- and re-assemble, there can be a (bigger) market for spare parts.

Sure, for a few days a year it might get as bad as resistive heating. How horrible! So you don't get 3.2x total throughout the year, you get 3.1x. It's a non-issue.

There is nothing experimental about self-hosting Zigbee stuff. It's an open protocol, so as long as the devices follow it (at least somewhat correctly) you can work with it.

And the actual "hard work" has already been done by others - Zigbee2MQTT, for example, supports over 3000 devices, so the ground work of having device definitions with easy use has already been done. What Matter aims to do is to provide standards for devices so that they all have some minimal basic functionality, expose the same fields in the same way, etc. so you don't need a hand-maintained library like that. There isn't even really a reason to be skeptical; considering all this stuff already works well enough, it can only get better.

It can definitely be hard or tiring, but you wouldn't be an early adopter. It's like saying that switching to Linux now (or even 15 years ago) would make you an early adopter. It wouldn't; it already works, plenty people have done it, but that doesn't mean it won't get better with time or that it's easy or for everyone.

Or maybe Firefox should find a sustainable business model.

I love and use Firefox the software, but their nonprofit is questionable, their leadership is scummy, and their business plan is nonexistent. They could, for example, start by accepting donations towards the development of the actual browser, which is the core product of Mozilla.

The paradox being that if therr were "premium" smart TVs for people like us - with proper support, privacy, customization options and no crap like ads - we'd probably buy it, and pay a premium for it.

But that's just too much work for them and they probably don't even realize that kind of market exists.

32" is pretty small for a living room TV. I'd wager most are at least 40 or more.

It's insane to think that people think of that as cheap. It's ... adequate. Clearly enough to run a company that can support it and further develop it, all the infrastructure, etc. Somehow all other companies convinced us that it should be okay to pay $10 per month or more for the most basic of services, where until now their revenue per user was maybe $0.5/month for the biggest users (ad watchers).

$10 per year is what the vast majority of subscriptions should cost - they'd still make plenty of money, but it's just not enough for them when they know they can nickle and dime you for more.

I like how your rebuttals both say that supplements are both not able to give you vitamin D but also simultaneously a risk of overdose.

You do realize that you can be both chronically deficient of something while also acutely overdosing on it, right?

Pretty much all LED "lamps" are made of many separate LEDs. Nothing would prevent you from having a few UV LEDs in there.

Or if you are technically inclined you can buy Zigbee or Z-Wave stuff, get your own dongle for it and run Home Assistant on your home server, and do everything 100% locally and it can still be really "smart". You can also do anything with it. But it's definitely not for everyone.

Hopefully Thread/Matter will help with this, which is an initiative to make interoperable smart home ... stuff.