[stuff deleted]
> It might solve the problems of users who just het a development
> kernel from some "distribution"-CD, but who are not reading this
> list.
> What do you think?
> First entries: "arithmetic faults in ``ps'' and ``top''" and "output
> of ``free'' and ``vmstat'' is wrong".
This is a good plan. Unfortunately many people will probably not read the
file! A good case in point is the 'make is broken' (new libc) story which
is documented but no-one seems to look at the README :-(
Perhaps if the kernel make procedure popped up a notice about the
existance of the file and then refused to continue until the user typed
yes or something it would help. Of course the 'yes' thing could be
bypassed by setting a flag in the makefile so that it would be a first
time only situation ...
--Craig