Sounds absolutely exhausting. Playing around with problems like that was entertaining in a way 20-30 years ago. I just want my computer to work nowadays.
I wonder... perhaps linux driver issues were more entertaining 20-30 years ago because there were less peripherals to worry about. Eg no wifi, no power management, etc etc. Nowadays computers are more complex, it's more involved, more exhausting?
20 years ago Linux still had issues with wifi drivers and power management. You had to be very careful with which network card you had (wifi and ethernet), as only a few had good in-kernel support.
Actually, it was very analogous to Asahi too. At the time, there was Yellow Dog Linux which was a project to make Linux run on (at the time current) PPC macs. I ran it on both Xserves and Powerbooks (although I only remember really being successful with the Xserves). Even then, the restricted amount of hardware present in Apple machines made it much easier to support that small subset in Linux. Funny enough -- Yellow Dog was also where the "yum" program started.
None of these things are that new... all of these have happened before and will happen again.
There were more peripherals to worry about, and fewer generic drivers.
It's not like people are trying to individually purchase PCI sound/ethernet/modem/graphics; they just get these components from the south bridge and they don't think about it at all. This makes things much easier, and gives a clear target for community notes/contributions.
I suspect that the reason he is describing this systems admin work as exhausting is because we've moved on. Personally, I only care about the thing I'm doing with the computer I happen to be sitting at, as opposed to computer itself.
Separately, power management is not a new detail: https://www.lpthe.jussieu.fr/~talon/pentiumII.pdf
The Pentium II processor supplies a STPCLK# pin to enable the processor to enter a low power state. When STPCLK# is asserted, the processor puts itself into the Stop-Grant state. The processor continues to snoop bus transactions while in Stop-Grant state. When STPCLK# is deasserted, the processor restarts its internal clock to all units and resumes execution. The assertion of STPCLK# has no effect on the bus clock.
But, if you had a good non-winmodem (external or more expensive internals), it was very easy to get working. Same as now... some peripherals are better supported than others. The more standard, the better. And this is true regardless of the OS.
30 years ago, there was no “just works” option. Windows was at WfW 3.11 / NT 3 , so you had to choose between being “entertained” by Windows driver issues and crashes (which you could hardly do anything about), or Linux (which you could!)
When Win95 came with Plug and Play, it still took a few years to go from “Plug and Pray” to “Plug and Play”. It’s only with 2000/XP that windows got to the “just works” stage.
Nowadays, everything just works, so the choice to deal with driver issues is a deliberate choice, not a “lesser evil” one.
This was my rationale for switching over to Macs (from a mix of Windows and Linux) many years ago. I liked the flexibility of working within Linux to get a workstation to be setup exactly as I wanted it. If I had problems, I was able to (usually) figure out a fix or script a workaround. Eventually I got tired of the thousand cuts of desktop Linux and decided to just use a Mac for the "daily driver" stuff and keep my Linux work on servers.
That said, I do feel that these projects are helpful. Aside from the joys of tinkering, I think that adding more OS support for older Apple hardware is useful. Not only are we adding more architectures to the list that Linux supports, but projects like Asahi will help to keep hardware useful well after Apple stops supporting it.
That said, we still have the same problems... new unsupported drivers, poorly supported power management... re-re-redesigned desktop environments that aren't yet uniform.
What I was disappointed with the article though was a lack of attempting to compile their own software. Their two major complaints were missing Google Chrome and an outdated version of Postgres. I know there are some differences with Chrome and Chromium, but they should have been able to compile a new version of Chromium and Postgres.
