On Thu, 2012-11-15 at 11:37 +0800, Thomas Goirand wrote: > On 11/15/2012 10:07 AM, Ben Hutchings wrote: > > On Thu, Nov 15, 2012 at 01:11:32AM +0000, Roger Leigh wrote: > > [...] > >> So systems which put additional logic in PID 1 are going to increase > >> the probability of bugs being present, and those bugs could kill > >> your system. > > [...] > > > > This is also true for the kernel, which is why we generally prefer > > to use Hurd... or not. > > > > Ben. > > The fact we aren't using Hurd has *nothing* to do with > the fact it is a micro kernel, and you know it. Aside from the fact that micro-kernels are grossly impractical. > If Hurd has the same level of hardware support as > Linux, as many contributors, and as many features, > then probably, it would also have as many users. It turns out that stupid architectures make it hard to attract and keep contributors. > Mixing the discussion around feature and engineer > design does *not* work, even in the case of kernels. Whether the kernel or init survives a crash is completely unimportant if the applications the computer is supposed to run become unavailable. What good is a smaller, less buggy init if it can't keep (say) apache and sshd running? Ben. -- Ben Hutchings friends: People who know you well, but like you anyway.
Attachment:
signature.asc
Description: This is a digitally signed message part