Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Most of my floppy disks are badly corrupted; IIRC data on them is considered to have a lifetime of about 10 years. That being said, my 360k and 720k 3.5" floppies seem to work considerably better than my 1.44MB 3.5" floppies even though they are quite a bit older. Not sure if it's because of cheapening of production or just the higher density of data.


I got myself a kryoflux about a year ago, because I also had a bunch of corrupted disks. I was actually able to recover almost all of them..... what I found was that different drives would have different failed sectors, so I bought like 5-6 different floppy drives, and would scan the same disk with all of them. I could then patch together the working sectors from each drive to get a full image.

Kryoflux is pretty amazing

https://www.kryoflux.com/


Most of my 1.44MB 3.5" have rotted away as well. Sadly I don't have older ones left to test. I also havent seen a 5.25" drive in a long long time.

What you should really do with floppies is to quickly copy the data to another storage medium.


For whatever reason my copy of "Star Trek: The Rebel Universe" (dated 1988) which is a lower density 3.5" disk works just fine still. It's old enough to be elected as a US senator now.


Same experience here: my late-90s PC HD floppies are almost all unreadable. The early-90s DD floppies from my Amiga are almost all still readable without errors.


I'm afraid to look at the state of my Amiga 3.5s, not to mention the older 5.25s. Sadly, they are probably lost to time.


Darn that's a shame. I wonder if you could retrieve some of the data using hardware like kryoflux.


I actually just replied to the person suggesting kryoflux... I was able to recover most of my seeming corrupted disks using kryoflux and a bunch of different floppy drives... different drives were able to read different sectors, which lead me to be able to recover the entire image.


That's really cool to hear :)

Also interesting that you used different drives to achieve that too, I wonder why different drives could read different sectors. Possibly the pickup coil of the read head was slightly different size maybe? Or the stepper motor aligns the head to slightly different positions of the track?


Not exactly sure. Those were some of my theories, too.


Were they able to read those sectors correctly? (Can you tell?)


Yes. Kryoflux is pretty good at not saying a sector is good unless it actually is. I was able to mount almost all of the images I made this way, and haven't seen data corruption issues.


All of my floppies are either very old commercial software (that I could likely download from an abandonware site) or personal data that I backed up a long time ago.




Consider applying for YC's Summer 2026 batch! Applications are open till May 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: