And I'm being downvoted for that? Seriously? In 13 years of networking I have never once had to reload machine to help with OSPF or BGP convergence. Good networking architecture and planning should mitigate anything other than a couple of minute outage. No routing change should ever require a reload of a server or end node.
Those who are still posting on HN are orders of magnitude more sensitive than those who post on Imgur. The communities are similar-size, yet Imguraffes are much, much more accepting of my comments. What merits a handful of upvotes there brings a downvote or two on this site.
I believe it. Networking is seen as a commodity now. It's transparent until it fails. There's a whole lot of technical debt lurking out there. I personally have seen the dark shadow of spanning tree suck the light out of DevOp engineers eyes.