The root causes of Linux's problems are very hard to fix. It needs to aim to attract commercial developers by adding cohesive APIs and stable ABIs. That's goi

The root causes of Linux's problems are very hard to fix. It needs to aim to attract commercial developers by adding cohesive APIs and stable ABIs. That's going to be a real struggle. Linux is too fragmented, there's too much opposition to proprietary software, and the kernel developers don't believe in stable ABIs. Perhaps some sort of VM, à la Android's Dalvik or .NET, is the best solution for working around most of these problems in one fell swoop. Being OS independent also opens up approaches to solving driver problems in the long-term. I thought Haiku had a decent shot of avoiding these problems, but my optimism has waned. Its ageing rapidly, development is slow, and rather than trying to attract commercial developers they seem happy to repeat Linux's packaging mistakes.