From current@ :
“My objective is to produce the best damn UNIX and keep it viable for at least another ten years.
That means that I have to be at least three years ahead of the users and I have to make it robust, fast and innovative enough that they want to use it.
For that “good enough” will not cut it.”
I totally agree with Poul-Henning, for what my opinion is worth.
What is considered “academic perfection” or “esoteric application” or “extreme needs” today, will be “normal quality standard”, “common use”, “basic requirements” tomorrow.