Reading about preservation efforts tends to put a smile on my face. But when floppies are involved and GreaseWeazle or FluxEngine aren’t mentioned, I do tend to get nervous.
There’s talk of “unreadable” disks; They shouldn’t be trying to read the disks normally in the first place! Imaging floppies at a raw level with high sampling resolution should always be the first step, and way less effort than trying to reconstruct the “unreadable” data.
Most “unreadable” disks actually read fine on the first try with the right imaging method. These tools are open source software/hardware, with the hardware part having low cost and complexity e.g. connect a few duponts between a floppy drive and a <$10 off-the-shelf stm32 development board.
They didn’t mention specific tools, but they did mention “making a clean flux rip” at one point.
Yeah, there’s such wording. But it’s such a passing mention I have zero confidence in the idea that they did use adequate tools.
With the tools being free and high quality, I very much doubt they wouldn’t desire to mention them.
Likewise, talk of reverse engineering the program or making custom builds of Dosbox, instead of reaching for PCem was weird to me.
Sure. When a program made by someone else doesn’t work, it makes sense to try the alternatives, before jumping into modifying the program. The reverse engineering part is just comical.