I would be reallllly careful using a heat gun - that can very easily destroy the big BGA chip on that board.

If I had to guess, the pads aren’t properly cleaned and also your iron’s tip may not be cleaned and tinned properly. Ribbon cable pads aren’t very small, so solder should melt onto them and stick quite quickly - you don’t have to heat a board long for that to happen.

Also, the metal shield you’re referring to is exactly that - an EMI shield, usually. A big piece of metal that is grounded and meant to help prevent interference. They’re typically made of steel, so solder usually does not stick directly to them! You’d have trouble soldering wires right onto an EMI shield unless it happened to be copper, which again, it likely is not.

So for padding, it sometimes depends on how your compiler works, but usually, it doesn’t pack bytes by default - that needs to manually be done. Otherwise, a uint32 followed by 2 uint16s, for example, will take up the space for 3 uint32s (in a 32-bit native compiler). If you manually specify packing (implemented differently depending on your compiler and such), then it will pack those all properly into just 2 uint32s.

I do imagine 24 bits followed by 16 more in a bit field for a 32-bit number would potentially cause problems. But it’s late here and I could certainly be wrong so take that with a grain of salt.

That also said, I typically don’t use bitfields directly in structures - it’s not usually good practice, at least where I work. I’d either do a uint8[3] or use a whole uint32 that is a union, and in the union would be your :24 followed by a reserved : 8, if that makes any sense. It’s sometimes worth it to leave a few extra bytes in there just from an organization standpoint.

Worked for me; I can see it fine

I do like that idea. I think Apollo (or one of the Reddit apps) had something that expanded the hit box on links to prevent that as well - that could potentially be useful, too.

I wouldn’t say to disable it - but if there was an option to toggle on/off somewhere, that would be awesome.

I’d personally be super surprised if they were outsourcing their firmware engineering - but I do suppose it’s technically possible.

Ahh, yeah. Neither would I. I would expect my USB sticks to last longer than that, lol.

That aside - here’s a fun fact. We sell the NAND from scrapped SSDs that we no longer need for development to a third-party vendor that actually desolders it and uses it for flash drives. So… you never really know what kinda flash storage you get on your flash drives! (Or… we did do this, until the program recently got shuttered because NAND is so damn cheap now)

I do agree with the plastic brick part - but there is actually reasoning behind that second part - the read-only mode. That happens when the flash is down to a very low amount of life left (usually predetermined by the manufacturer). It is by design because the flash will degrade further if you continue to write to it, so by forcing it to read-only mode, users can still recover their data in a failing/aging SSD. Not to say it isn’t a huge pain in the ass when that happens though, lol

These failures don’t have to do with where they’re manufactured - it seems like this is some sort of firmware bug. NAND doesn’t really just choose to wipe itself at random. Actual NAND chip failures are few and far-between, so this is very likely much more than a hardware issue.

That said, I personally have done a lot of testing with WD-manufactured NAND, compared other companies’ NAND - and the WD NAND is pretty crap. I can’t really go into further details than that, though.

Source - I’m an SSD firmware engineer.

You may be able to find scrapping places around you (e.g. metal scrappers) - my local scrapping place also takes circuit boards. Which gets me ~$1/lb or so. But they also just take misc. electronics for like $0.10/lb, phones are like $5/lb, etc. there’s prices for it all. Best part is, this way you can actually get paid for your trash, instead of doing it the other way around.

Among what everyone else is saying. Customizable color schemes! If I could browse Lemmy with a solarized dark-esque color scheme, I may actually cry…

And then you’d have the following processors be called the Intel Core 360, followed by the Intel Core 1