Fascinating read that unlocked some childhood memories.
I'm secondhand pissed at the recovery company, I have a couple of ancient SD cards laying around and this just reinforces my fear that if I send them away for recovery they'll be destroyed (the cards aren't recognized/readable by the readers built into MacBooks, at least)
My understanding is that flash memory does not do very well at all for long term unpowered data retention. flash memory is basically a capacitor(it is not really a capacitor but close to one) and will loose it's charge after a few years.
And magnetic drives will seize up. and optical disks get oxidized, and tapes stick together. long term archiving is a tricky endeavor.
It is however an interesting challenge. I think I would get acquainted with the low level protocol used by sd cards. then modify a microcontroller sdmmc driver to get me an image of the card(errors and all). that is, without all the scsi disk layers doing their best to give you normalized access to the device. Or more realistically, hope someone more talented than me does the above.
Tapes hold up really well if they're not in absolutely awful storage conditions. And the claim at least was that the early CD-ROMs were quite durable, being a straight up laser carved diffraction grating. CDRs on the other hand rely on dye which will degrade rapidly.
M discs were made, quite possibly to meet the Mormon need of post Armageddon lineage documentation, to last at a minimum of 1000 years in reasonable condition. They are just special, expensive CDRs that use a different dye system that won't break down in 4 years.
My optical disk drive was $17 and has M disc writing compatibility, and my understanding is they are meant to be read by any CD reader.
That is true about mag tape, I suspect tape to be one of the better choices for archival storage. In fact the biggest problem you will have with mag tape is making sure you have a working drive 10 - 20 years in the future when you want to look at your archives. to make things worse tape drives are getting more and more flimsy and fragile as tape tech advances.
my first job we had a vault of ibm reel to reel tapes of old business data. our attitude if we were ever asked to pull any of the data was that we would probably only get one chance at it as the ferric material on on the tape had a disturbing tendency to flake off. note that there are techniques to reduce this, but we did not have the means or motivation to apply them.
And a pedantic observation on your correct point about optical disks. you can't make backups on pressed disks, only recordable ones.
You or I can't make backups on pressed disks, but it is an interesting consideration if part of your archive consists of commercially released movies, games, music, etc. that could be stored as original Blu-Rays, DVDs or CDs.
> My understanding is that flash memory does not do very well at all for long term unpowered data retention
You need to let flash cells rest before writing again if you want to achieve long retention periods, see section 4 in [1]. The same document says 100 years is expected if you only cycle it once a day, 10k times over 20 years (Fig 8).
Last year I helped a friend recover photos from a portable WD HDD. It was formatted in FAT32 and I was forced to run R-Studio to get reliable results. There was a lot of damaged (readable, with artifacts) and corrupted (doesn't render, have wrong size) files.
I'm secondhand pissed at the recovery company, I have a couple of ancient SD cards laying around and this just reinforces my fear that if I send them away for recovery they'll be destroyed (the cards aren't recognized/readable by the readers built into MacBooks, at least)