alt Hacker News> but I believe that success Linux has had is because of copyleft
No, the success Linux has had is because it ran on the machines people had at home, and was very easy to try out.
An instructive example would be my own path into Linux: I started with DJGPP, but got annoyed because it couldn't multi-task (if you started a compilation within an IDE like Emacs, you had to wait until it finished before you could interact with the IDE again). So I wanted a real Unix, or something close enough to it.
The best option I found was Slackware. Back then, it could install directly into the MS-DOS partition (within the C:\LINUX directory, through the magic of the UMSDOS filesystem), and boot directly from MS-DOS (through the LOADLIN bootloader). That is: like DJGPP, it could be treated like a normal MS-DOS program (with the only caveat being that you had to reboot to get back to MS-DOS). No need to dedicate a partition to it. No need to take over the MBR or bootloader. It even worked when the disk used Ontrack Disk Manager (for those too young to have heard of it, older BIOS didn't understand large disks, so newer HDDs came bundled with software like that to workaround the BIOS limitations; Linux transparently understood the special partition scheme used by Ontrack).
It worked with all the hardware I had, and worked better than MS-DOS; after a while, I noticed I was spending all my time booted into Linux, and only then I dedicated a whole partition to it (and later, the whole disk). Of course, since by then I had already gotten used to Linux, I stayed in the Linux world.
What I've read later (somewhere in a couple of HN comments) was that, beyond not having all these cool tricks (UMSDOS, LOADLIN, support for Ontrack partitions), FreeBSD was also picky with its hardware choices. I'm not sure that the hardware I had would have been fully supported, and even if it were, I'd have to dedicate a whole disk (or, at least, a whole partition) to it, and it would also take over the boot process (in a way which probably would be incompatible with Ontrack).
Replies
I'd say with modern hardware, like the xe Intel iGPUs on 11th gen Intel and up got driver attention quickly. Some things like realtek 2.5gb NICs took a little while to integrate but I think realtek offered kernel modules. I remember NIC compatibility was sparse when I started playing with it around 1999-2000. What trips me up is command flags on gnu vs freebsd utils, ask me about the time I DOSed the Colo from the jump machine using the wrong packet argument interval.
I don’t disagree with what you say. But why did Linux work on all that hardware? I assert that if you trace that line of thinking to its conclusion, the answer is the GPL.
Many people and organizations adapted BSD to run on their hardware, but they had no obligation to upstream those drivers. Linux mandated upstreaming (if you wanted to distribute drivers to users).