On Thu, Nov 10, 2022 at 1:14 AM Theodore Ts'o <tytso@mit.edu> wrote:

I'd also argue that (a) the differences between the Linuxes aren't as big some people would make it out to be

It's all a matter of degree and what things bite you. This is why we had to develop the cluster checker because Linux is all over the place >>still<< -- one famous ISV had 144 possible differences between 'standard' Linux cluster configurations they needed to test. BTW: that had been a developer during the UNIX wars and pure Fortran on VMS/IBM etc days. I suspect they would claim it was not much better for them.

Is it better in some areas? Sure, in another no. It depends if it's something that bites you and you care about. Things like the "systemd wars" did not help which I would contend was not much less glorious that the Unix wars -- just different players. In the end, it's all about who is making the call of what is different/what is changing. I'm too old to get too religious about this.

The 'sameness' is because of UNIX not because of Linux. Ken and Dennis's core ideas created an industry and set of systems that are all 'close enough that we all can get work done. In the end, at any given time, one group or any other will have the incentive to have their way lead the market - and the hard part for many of us to accept is that the incentive is most often economic not technical.

Larry's old paper nailed the basic issue. It was use the core ideas of Ken and Dennis - (call it anything you want) - and make it freely available to remove the economic barriers. This is what Linux exploited, and that is what made it successful. I'm pleased to say, that Larry's 'Sourceware' came about --> today its in the form of Linux. But the ideas (the core IP) is called UNIX. And yes there are specific differences. But in general; they are close enough.