Sorry, on demand is not a good way to state this, it’s just how my weird mind thinks of things. By “on demand”, I mean, like you are actively using it to store something or view something. If you’re not intentionally doing something with it, the drive should be completely idle. That’s more of a target than a requirement, though. It’s a way to keep storage drives tidy and not littered with temporary cache files, or databases used to store runtime state by various services. It’s just a strategy I like to take, to keep bulk storage separated from the applications and services that use it.

Even if a usb drive is intended to be permanently attached, it should still be treated as a temporary component. The reason is so that if something happens and the drive is disconnected, it limits the disruption to the system. You lose your media and documents until it’s reattached, of course, but the computer keeps chugging along happily.

If you use it for writing log files, then its loss can disrupt those services (and also prevent the problem from being reported). Also it’ll be constantly making noise, which can be annoying.

That’s my reasoning, anyway, you might prefer it done differently.

I’m not mixing up units, but let me better explain what I mean. The max speed is only in a best case scenario with a single sequential reader, and that speed drops dramatically when adding other simultaneous operations because the read head needs to seek to different locations. Random read speeds regularly test at less than 1MB/s, and even though multiple sequential streams wouldn’t be random, it’d still require plenty of seek time.

I did a little testing on a drive I have here just now to make sure I’m not completely full of shit. Single stream read was about 120MB/s and I was surprised how well it handled multiple read streams. My drive could handle roughly 9 sequential read streams from different locations on the drive while staying above 10MB/s, so while it wasn’t reaching its max speed, it wasn’t horrible, matched your expectations almost exactly. The real killer, though, was writing. If I added in a single write stream, the read speed dropped to about 1.5MB/s because it seemed to strongly prioritize writing over reading. Maybe some configuration could improve this? Interestingly, adding more readers improved this, but only up to about 4.5MB/s.

My results shouldn’t be taken seriously, it’s just one drive and me mucking around with dd, but I think still illustrative of what I was alluding to, that if you are using a single HDD for multiple things simultaneously, the performance can suffer badly. Actual performance will depend on its use, of course, and honestly the results are way better than I expected, so this isn’t likely a realistic concern at all unless you will be constantly writing large amounts of data to the drive.

Thanks for calling me out on this, these are really interesting results, I think.

2.5" is going to limit the storage size a lot. So if you need >4TB, the internal option isn’t going to work. A janky solution would be to route the SATA and power connector outside the mini-pc so that you can fit a larger drive, but that’s going to be pretty ugly, and still limit you to only a single drive.

A single hdd can be fast enough to serve UHD ripped 4k video, but it’s much closer to the limit than an ssd would be, and might not be a great experience if you will be doing anything else with that drive at that time. Having a cache drive or multiple hdds in an array (or even better, both) greatly improves this.

A usb enclosure would let you easily have multiple hdds, but as everyone will say, they are less reliable. My opinion is that while you should never use one as a boot/system drive, they’re fine for bulk storage for home use. Make sure you’re not writing logs or anything like that to it, it should be on demand use only, and you might have to reconnect it occasionally. Anecdotally, I’ve never had issues with usb enclosures, they’ve worked fine for me in the past and I continue to use one for backups, but maybe people with some horror stories would have very different views on this.

I had a setup similar to this for a year or two that ended with an hdd destroying itself one night. Probably because of the drive and not the usb enclosure I was using. Until then it worked fine, it’s definitely a viable route.

If you can swing a desktop pc case it’ll probably end up easier and cheaper and have some headroom for upgrades, that’s the route I went down after trying an escalating series of mini-pcs and running into their limits one too many times.

I have a few kindles, have upgraded over the years and have been able to use them all in the same manner:

With a new device I connect it to the internet and update the firmware to the latest version (the factory installed version has had a lot of missing functionality in my experience). Then I block it from my network, delete the AP entry and put it permanently into airplane mode.

When purchasing an ebook from Amazon you can download it for usb transfer and I organize it on my laptop with Calibre.

Calibre can also strip drm, but if you’re transferring it to the device you downloaded it for it isn’t necessary.

Amazon may at some point in the future change all of this, but the content I have already downloaded can not be revoked and is usable outside the Amazon ecosystem if the drm is removed.

The form of this kind of social media has got the same set of upsides and downsides as it does on Reddit. It won’t be exactly the same because the people are different, but the problems aren’t that different and the people aren’t that different either.

As a mostly lurker I find the experience pretty similar. I scroll through and find some interesting articles, bits of news, memes. It’s a slower pace, but I think in time it’ll grow faster. People migrate over occasionally, but there may be a critical mass moment when it’s big enough that lots of people start flooding over. Or it won’t and it’ll just fizzle out to nothing over time, who knows. For the moment it’s good enough for me to have replaced Reddit entirely.

As for things that are better: you get a lot more control over how you want to experience it. There’s no singular controller always dragging the experience down toward profitability. There are clients a-plenty, the api is open, you can control what parts of the network you see and which you don’t. It does take some effort, of course.

As for worse, because there’s no singular entity controlling the network, there’s going to be some very dark corners. You can block them (many will be blocked by individual server operators already), but they’re still there and they get to carry the Lemmy name and newcomers are most likely to experience it.

Just my thoughts on the subject, it’s been discussed a lot, I’m sure other people have quite different perspectives.