ISMETA

joined 1 year ago
[–] ISMETA 25 points 11 months ago
[–] ISMETA 19 points 11 months ago (1 children)

PHP 8 makes it possible to rescue the princess but your 83 legacy princesses are all still PHP 5.

[–] ISMETA 3 points 1 year ago

I wonder if there are any technical reasons with maybe ActivityPub that old posts might have to eventually be archived?

But I enjoy the idea of never archiving or at least keeping posts open for a long time like 10 years. I'm not sure why but there is something special aboaut old threads being revived or kept alive. I had this one comment on Reddit where I purchased a product and months and years later people would still respond asking if it was still working or if I'd still recommend it. This I is like "The internet never forgets" but in a good way.

[–] ISMETA 27 points 1 year ago (2 children)

I hope somebody has a better answer for you but I think it's just that the world has decided that one has to use two hands to operate a Smartphone now. I'd love a phone with a 5" screen but it looks like we are a (vocal) minority that's not profitable to serve?

[–] ISMETA 2 points 1 year ago

NT Cassettes were about the size of a full size SD card and maybe twice as thick and that was in 1992. Imagine what 31 years of research could have done to that technology!

[–] ISMETA 3 points 1 year ago* (last edited 1 year ago) (4 children)

You could probably carry some apps on individual tapes with you? If it was all miniaturized enough and we can cache some things. Digital photos and videos would work too as that obviously has been done many years ago.

[–] ISMETA 3 points 1 year ago (6 children)

Do we really need our modern data storage for mobile phones? Mobile phones for sure would be very very different than our modern smart phones but mobile phone networks don't sound impossible. Of course the internet would have to work very differently too, but maybe routing and forwarding could be done just with everything from RAM?

[–] ISMETA 3 points 1 year ago (1 children)

That's of course true, i was just wondering how far we could push it.

[–] ISMETA 8 points 1 year ago

I'm incredibly unqualified to even think about how one might get faster random access times but i was imagining sci-fi solutions like looped tape so that you are always at most 1/2 of the tape length away from the point you want to reach, or the tape equivalent of multi actuator hard drives where there's be multiple independent tapes in one cassette in a sort of RAID style thing but maybe instead of (only) striping data could be stored on multiple tapes in different places to always have one tape that is at a position close to the data you want. Or a system where the same tape has multiple read heads applied to it in distant places.

[–] ISMETA 3 points 1 year ago

I'm aware of the enterprise backup solutions as mentioned in the main post. Still 45 TB are very impressive. But I was just wondering how diverse and powerful tape could be if it was the only viable storage solution. I'm assuming high speed rewinding and seeking and miniaturization would be something that the industry would have put a lot of effort into in that case, but for backup solutions those properties are less important.

44
submitted 1 year ago* (last edited 1 year ago) by ISMETA to c/[email protected]
 

Let's do a thought experiment.

Let's assume we just never got hard drives to work all that well, head crashes are common and large storage capacities are only possible in servers with incredibly expensive anti vibration setups and what not and there's no way they'd ever work in portable devices. And optical media just didn't work out. Maybe somehow we didn't discover the science in time and the companies working on it just failed or bad management decisions killed off the research into it before anything useful came off of it. And flash storage just never came down in price.

How far could we have pushed cassettes and tape if all the effort that went into other technologies had to be put into cassettes because there simply wasn't a good alternative for data storage. What are the limits of how fast we could move tape in a cassette? How miniaturized would the technology be by now in 2023? I know that there are contemporary tape backup systems with large capacities but there hasn't been any efforts into high speed seek times or making cassettes a viable tiny medium for use as removable media on PDAs or mobile phones.

If we could pack data densely enough and move the tape quickly enough how possible would modern computer tasks like high resolution digital video or image editing be? Assuming that we still had the high speed processors and non-persistent memory of the present but just no other data storage medium aside from magnetic tape.

For inspiration consider that in 1992 we had NT Cassettes that are about as big as an SD card an could store nearly a Gigabyte and in the enterprise LTO-9 tape (released in 2021) stores around 18 TB. So it doesn't sound impossible to have tiny cassettes with a lot of storage if we spend the last 3 decades working on it.

 

I'm aware that the proper way to handle backups for a NAS is to have a second TrueNAS machine on preferably a different continental plate or at least a different city, and sending ZFS snapshots to it.

But well I don't exactly have a second flat nor the budget for a second NAS and cloud storage is a continuous expense that I'd rather not deal with. So I was thinking I'd just backup the ~1.3TB of not re-downloadable data on a bunch of USB hard drives but then I noticed how similar the prices for 2TB m.2 SSDs and for 2TB USB hard drives are.

So my plan is:

  • Get 1 Icy Dock NVMe m.2 enclosure.
  • Get 3 of the cheapest brand name 2TB m.2 NVMe SSDs.
  • Setup the SSDs with BTRFS and LUKS.
  • Backup my ~1.3TB of important data to each of the SSDs. Put each SSD into a sturdy box give one to my parents, one to a good friend that lives many hours away and put one on a shelf in my room. The drive at my friends place would give me at least some of my data back in case the entire city was leveled by a Kaiju.
  • I'd rsync my NAS to the drive at home every week to protect against the NAS kicking the bucket
  • Since I visit my parents frequently I could just occasionally swap out their drive with mine to keep that backup somewhat up to date. Which gives me protection from a house fire. The drive at my friends place I should probably be able to update like every year or two. Due to LUKS I wouldn't have to worry about the drives being stolen and with BTRFS I could run a scrub when I get one of the drives that haven't been update in a while to see if the backup was corrupted. The scrubs are a reason why I think SSDs would be better than HDDs because then the scrubs would be fast. I'm choosing BTRFS over ZFS for the external drives here because then I can just read the backup drives from any normal Linux machine and don't need to setup ZFS on Linux first.

So how dumb is that plan? Am I better of doing the same thing with HDDs instead?

[–] ISMETA 5 points 1 year ago (3 children)

Sounds good but there isn't any consumer equipment that can handle 2GB/s. Even 10 Gigabit Ethernet switches are super expensive and I don't think we have anything that can do more than 10Gb/s in the consumer Networking space at all .

[–] ISMETA 2 points 1 year ago

As somebody who likes using the terminal I too have mostly stopped using dd and use gnome disks instead. Getting the rightdd flags to get the best performance and progress indicators is a challenge to Google every time.

 

I downloaded Street­Complete to contribute to OSM and on https://f-droid.org/en/packages/de.westnordost.streetcomplete/ it warns you that "This app promotes or depends entirely on a non-free network service". I'm not super concerned about that and will still try out Street­Complete but i was just wondering what non-free network service this is. Does anybody here know?

 

The 2-Player Starter Set was announced for the 27 October 2023 a while ago. What do you think the two decks will be? If you got to decide, the decks what would you make it?

 

After only 3 short years my Google Pixel 4A is EOL. It still works perfectly fine, but since our entire lives run through our phones I don't feel safe trusting my bank accounts and so on to a phone without security updates. I feel like I wasn't quite aware how short the supported time would be when i bought it.

So I gotta get a new one. I don't really have a lot of requirements. I just need the basics: A camera, headphone jack, USB-C charging and F-Droid (so app side loading). I just watch some YouTube, use lemmy and messaging apps, 3fa, email and so on and if there is something to take a video of photo of I like doing that too. I don't need NFC or high refresh rate or 3 cameras or wireless charging and I don't play any phone games so pretty much any phone should work, right?

I was thinking of getting something a bit sustainable and long lasting so I looked at Fairphone but the Fairphone 4 is only supported until 2026 which is not really long and also it's really expensive for just another 3 years. Looks like there will be a Fairphone 5 soonish? But since the 4 didn't have a headphone jack the 5 probably won't either?

So since it looks like a ~6 or 7 year lifespan is just not something that's available I've been thinking why not go cheap? Thus I've considered getting the Samsung Galaxy A14 (non-5G). Are there any significant differences with the 5G version? The 5G version has good reviews as far as i can see. Feels a little bad to downgrade to a phone that cannot even record 1080p@60 and to go back to USB 2.0 and 15W charging but whatever.

So I'm open to suggestions, either on the longer lasting side or something cheap yet secure for a couple of years.

view more: next ›