A year ago you could buy 10k+ units of most common parts -- in stock -- from each of the major distributors.
Some people hoard toilet paper, some people hoard baby formula, some people hoard Shanghai apartments, some people hoard microchips...
Got burned with same STM32s recently. A purchaser been shopping parts, and ordered a given model of MCU by muscle memory. The price has moved one zero overnight, and we ran for quite a sum.
That is not sympathetic, that's very normal.
What you're describing is sympathetic and hard to deal with. Their example was lucky about it and still ended up somehow quoted in the story.
However, their software sucks (okay, so do all of the other options...), their peripherals are not as good as some of the other options (in particular I'm thinking of peripheral clock trees when I say this, among other things), and their availability is the _worst_.
Seriously, unless you're big enough that ST management actually knows your name, expect to have sudden availability issues with the STM32. This has been true for the last decade. It will remain true. (Many of these availability problems start at ST's wafer fab, so they're not shared by the other vendors.) I always advise clients who care about such things to consider other vendors.
Now find a single person who can program bare hardware on a short notice on above Arduino level.
The current chip shortage has claimed 6 of our firmware devs, all hired by companies ready to spend just any money for anybody "who can replace that ____ing STM|NXP|Renesas thing"
People downstream in the industry greatly overestimated their knowledge of the industry.
Lots of tech companies around who had zero prior knowledge of embedded development, now jumping on it, and breaking their teeth.
I haven't heard more fabulous questions like "Are there other microcontrollers than Intel, and AMD in the world?" this year than any time before.
Wait, there's real money in that?
Maybe I shouldn't waste my time just doing yard work.
Sidenote: I just asked my wife about parts shortages at her workplace. Evidently you guys are right. I should get out more.
Firmware engineers are the out-of-sight-out-of-mind component in the product development spectrum.
Although the money has been decent, programming complicated devices has spoiled me for one thing..I have just about essentially zero interest in web technologies, products, languages. I've always gotten to work on the sexier version of embedded work and so feel really blessed, it's not all about the lucre.
Which is the saddest part of it all.
But a thing to remember is that people currently hiring themselves been likely neglecting their RnD before the crisis hit them.
Condolences on the loss of coworkers.
But this seems interesting to software people who haven't been doing embedded, but who can wrangle C well enough to adapt (quickly?) to embedded constraints and interfaces.
It's knowing how low level hardware operates and knowing how to dig yourself out of the various gotchas like for example figuring out that a last minute replacement of the crystal oscillator by the HW team with a cheaper one is now screwing up your boot process or that the reason your products are starting to go haywire after 6 months is an interrupt triggering too many writes to the flash, wearing it out prematurely, etc.
Sometimes being an embedded dev is dead easy. Other times you are dealing with a GCC chain from 2003. 2 layers of some 3rd party of libs that someone bought in 2005 because the VP was good friends with a buddy in some other company that made something similar (which you may may not have the code for). With 64k of flash and RAM. And the vendor ground off the chip numbers and put their own sticker in place. Playing detective, historian, and making good guesses as you may not even have access to the firmware anymore. Then on top of that sometimes you are really lucky and have a debug port and can play the printf logging game to see what is going wrong.
Are you in the US? Because EU is full of embedded devs.
It looks easy to find STM32L0x2 chips in stock. Those are not supported by the QMK firmware, but they are very similar in terms of peripherals and features to the supported (and hard-to-find) STM32F0x2 lines. The main differences are a lower minimum operating voltage and more power management options, IIRC.
One vendor is having trouble finding any chip to design for. The software stack supports a couple of dozen STM32 chip series, but none of them are even remotely available. I've seen a lead time of over 11 months. Before COVID, pretty much every single chip was in stock almost all of the time.
Respinning a board isn't too difficult, but good luck doing that if there's no chip to respin it for.