Linux is very close to perfect as a "decision support and design" workstation. Tool performance, ubiquity, interoperability and stability are the big targets there. We have to name as a goal data and media asset durability and tool compatibility over years or decades.

Linux has virtues that end up conflicting. An occupational hazard of Linux as a tool and OS innovation platform are disruptions, disappointments, instability and dropped balls.

Other people are goofing off without enduring responsibility for a large user base. We confuse design criteria we inherit from the commercial world. Easy to use, easy to remember, easy to diagnose problems and "quick to getting going" are not the same always compatible design targets. The Window Manager wars took up tons of Linux desktop attention. Gaming is naturally an arena for hardware churn and 1-upmanship.

I have great hope for Linux desktop stability with government deployments and thus a global tug for stability against rushing breaking changes. The Linux platform and package manager contents are remarkable achievements.