Skip Navigation
Microblog Memes @lemmy.world
The Picard Maneuver @lemmy.world

Why does my treadmill want my email address?

329 comments

  • I was going to buy a really sweet drone. Then I watched the Getting Started video and there was an app and an account thing, and I realized the second they shut down the service, that drone would be a paperweight.

    I'm back to building my own because I'd like to use it for more than a year or two.

  • I wish we could go back to the 90s as far as this shit is concerned.

    Just take it out of the box and it works.

    Send in a registration post card if you feel like bothering, but thats about it.

    I am so sick of everything, especially shit that has no conceivable need to be online, not only demanding an internet connection, but demanding accounts and shit too.

    Cause you know why they do it. They want to track you, harvest your data, and monetize it. Its not about selling you a good product, its about selling you a good listening device.

    • honestly this shit is so ridiculous, a fucking toothbrush wanting wifi and an account so people can spy on you even more closely than they already do, hell when it comes to regular internet services i 80% of the time nope out when they ask for an account...how do people WANT TO buy this shit 😭?

    • Exactly - it's not about selling you a product, it's about you becoming a product they can then sell to third parties.

  • Meanwhile in my password manager: 727 login accounts 😶

    • 1510 in my Keepass and 1423 in my Bitwarden 😬

      That's collected over 15+ years though.

    • Mine has 40.

    • There kind of is an argument here that maybe more services should permit for use of public-private key authentication.

      Using one password with multiple services is a problem, because you have a shared secret with the other end, and if you use the same password with multiple services, that service or people who break into it could impersonate you elsewhere.

      But with public-private key encryption, you never hand out your private key. You only use it to sign a specific request sent you, so that risk doesn't exist. You can use a public-private key pair with one or multiple services.

      I mean, personally, I'd kind of rather have three physical keystore devices.

      One I carry with me. That stores the key or keypairs necessary to do the sort of things that I carry auth data with me -- my keys and the cards I carry in my wallet. Just means that I only need one device.

      The other I leave at home, stored securely. That authenticates to maybe more-critical stuff, things like a stockbroker, maybe -- stuff where I don't need day-to-day access, and don't want to worry about my credentials going missing.

      The last I keep in a safety deposit box in a bank. That has all my authentication stuff. That's to deal with catastrophic situations, like my house burns down or I get killed and need a way to pass authentication stuff. The bank makes me jump through a lot of hoops to get access to it, but it's there.

      I'd like the device to have a display and a keypad, so that I don't have to trust external input devices as to what it is that I'm authenticating (e.g. smartcard point-of-sale systems do this).

      I'd rather not use a smartphone for the first device. The smartphone is just too damned complicated and rapidly-changing for me to really want it to store my authentication data. I'd rather have it be a separate token, something that I can plug into a smartphone or point-of-sale terminal if I want to perform an authentication.

      There are crypto tokens that contain keystores -- powered or smartcard -- but they tend to not have a screen or keypad, to save on costs. I don't really feel like that's something that I need to save on, as long as I only have one.

      I'd like the device to optionally permit setting a passcode for a given key on it. That's not an ironclad form of security, but makes it harder than just pickpocketing someone's keys. And for some things, that I use all the time -- like my house -- I don't need to have a passcode.

      This has a number of benefits:

      • If you're mugged or something, you physically are unable to authorize to things that require the keys on the device at home or the device at the bank. In fact, you can credibly say that you can't do so. That counters coercion issues:

      • You don't need to trust POS terminals. Sketchy terminal? Not a problem.
      • You can keep a log of transactions on the device.
      • You don't have to worry about the latest clever smartphone attack compromising your credentials.
      • You can use the thing the same way with a smartphone or computer or point-of-sale terminal. That's something that we really don't have today -- most people don't have smartcard readers, and vendors generally don't have support for authentication for those.

      It has some downsides:

      • It's another device to carry.
      • It needs to be powered (though it could have very low power requirements, like a digital wristwatch, run for a year on a charge, unlike a smartphone, and could potentially charge off USB or similar). You wouldn't want your "keys" to lose power (though people who do stuff like smartphone payment already need to worry about this).
      • It costs something.

  • I don't want my computer to treat me like an idiot, it's my computer let me run whatever commands I want.

    • GNU/Linux

      • There are a few utilities even on the command-line side that will require confirmation (or passing --force or something like that) but it tends to be in cases where you almost certainly don't want to do what you're doing.

        And there are a very few that just don't let you do so at all.

        rm won't normally remove a file if you don't have permissions to do so, though if passed -f, will give you the permissions if you have the authority to do so.

        mkfs utilities ("create a new filesystem") typically require a force flag to overwrite a filesystem that's in use; normally when that happens, it means that someone's typed the wrong device file name and is about to blow away the contents of their drive.

        fsck, the filesystem checking and repair utility, will refuse to modify a mounted filesystem at all (which normally could be expected to corrupt a filesystem).

        That being said, I think that there's a serious problem on Windows dating back many years where programs throw way too many warnings up, where users constantly encounter confirmation prompts even when they are doing a pretty normal operation and in fact do want to do something. That's not just annoying. It also trains users to just whack "confirm" anytime something comes up, which makes it impossible for software to meaningfully warn when there is a serious problem.

        I'd also add that I kind of think that GUI software would benefit from a standard "confirmation API". It used to be the norm for software to throw a dialog box up for confirmation. Linux and Adnroid -- and I assume Windows and MacOS, but I'm out of date there -- have a notification API, where software can tell a notification manager that the user needs to see a message. That's nice, because then the notification manager can handle the notification in sophisticated ways; do things like text the user the notification, auto-dismiss notifications, filter some out, play a sound, refrain from playing sounds, etc. But AFAIK -- and I don't use a lot of GUI software these days -- they still use confirmation dialogs. I'd kind of prefer that they use something like the notification manager, so that one could set up the notification manager to auto-accept certain notifications, log notifications, and so forth. Another annoyance is that most dialog boxes are set up to have Enter and sometimes Space auto-accept. This is obnoxious, because one might be in the process of hitting Enter or Space when interacting with another window; if a confirmation dialog comes up, one can simply immediately inadvertently accept the confirmation. Having the notification manager handle confirmations would help avoid this. I'd personally rather have a dedicated key or key combination to confirm something that's used only for confirming things, and I'd rather have such confirmations processed in first-in-first-out order. With software throwing dialog boxes up, a confirmation can "jump in front" of another, using last-in-first-out order. Plus, it'd let me have the confirmations auto-accepted.

      • I use Trisquel GNU/Linux-libre, I'm just pointing how that it sucks that im seen as a boomer for wanting to actually use my computer rather then my computer using me.

  • This is something Matter was supposed to address at least for IoT stuff, but it's still not great.

329 comments