I am sure some people are interested in the topic on its own right, but obviously most just want to escape from systemd. In this case, please first read this paper. Even if you think the end result is crap, there are some very lucid ideas in there. You would do well to at least consider them as you adopt OpenRC or whatever in your thin Linux distribution. I swear that I have no relationship with systemd project and only occasional hobbyist relationship with Linux.

Every server is different, packages and dependencies are very much relevant. Some need a fully functional framebuffer with OpenGL support to generate web images/video.

If you are asking to configure system on build master and then mass deploy to individual servers, without any unnecessary development/configuration tools, that's a reasonable idea.

You don't debug devices like watches and lightbulbs by connecting a keyboard and booting them to single user mode. You mount them as a drive, or they replicate logs over network.

It's easy to laugh at Android for limited upgrades. But the truth is, you wouldn't like most of the upgrades if they were available. iOS8 is probably perfectly fine on iPhone 6/6+. On others devices, you should only upgrade when you need the new features. Performance is likely to go strictly downward.

Parallel is trivial.. hehe. Not when it needs to work unattended every time. Monitoring TCP/UDP sockets and starting/stopping services and their dependencies on demand in /bin/sh - interesting! Managing messages from all smart light bulbs in your house in text files? Go and try it.

Or is it complex simply because it takes complexity to do things better? Gnome is more complex than twm for example, but many people find extra functionality useful. Is traditional script-based init a good match for cell phones, watches, robotics, other new devices that Linux needs to support to remain relevant?

I fully understand that there should be server and education oriented Linux distributions where simplicity and ease of customization are more important than boot time milliseconds. Just don't think that others are doing anything wrong by catering to their own needs.

Think just for a second about how web email works, especially web e-mail that provides fast full content search. Or SMTP from outside systems. Can't read user's e-mail. Riiiight! Maybe with all open source client stack using public keys exchanged out of band.

At least in the US there are countless software/IT jobs because every business relies on software for its continuous existence in some way. Yor may have to downsize and live on lower wages for some time, but you will live.

J2ME? Palm/WebOS? iPhone? Android?

Are you saying that buying servers is the same thing as maintaining, backing up, securing, auditing servers? What is your threat profile? If NSA or Amazon hacking your data are realistic top concerns, by all means deploy your own datacenter with armed guards. If it's common crooks, big providers are more likely to discover and patch exploits, detect intrusions and withstands DDOS attacks. It's their nest egg and they focus on protecting it.

SF Bay Area, the birthplace of smartphones, badly needs dual sim phones. There are coverage gaps even with AT&T/Verizon, right alone highway 101 and major tech companies. Having a second prepaid sim would be a godsend for actually being able to call people when you need to. Especially if you want T-mobile unlimited plan rather than paying $1K phone bill because of a bug in one of your apps.

As opposed to what? Data on thousands of individual laptops? Servers in the hands of IT department in a company for which IT is not a core competency? I would think a cloud provider that specializes in this sort of stuff is less risky, all things considered.

Now, government secrets or say Coca Cola formula should be obviously stored in physically secured datacenter guarded by best security money can buy. Probably still not users laptops. But city gardening logs? I think Google Docs is fine.

Ergonomics is for 40 hours/week desk workers. If you are a gardening supervisor and spend most of the time interacting with workers, you can manage an hour/day hunched in front of a laptop filling in forms. In fact, you will prefer the flexibility to work anywhere, connected to a Windows XP cloud instance running your thousand custom applications. Obviously if you are going to spend most of the day at your computer, you should have a nice big monitor and a height adjusting desk.

Do all 8300 employees need individual desktops? This is not a software development company, and those machines still need to be managed, maintained and replaced. Keep big depos of $250 chromebooks where anyone can get one for temporary or permanent use at office or home. Then return when done, as still working or broken. No IT costs, as data is in the cloud.

For heavier use, provide computer labs with a choice of platforms, so if someone really needs to work on the latest version of Office or Photoshop, they can.

And of course, anyone who is expected to work on computer for hours every day, or handle sensitive data, should get a laptop/desktop of their choice with reasonable price constraints. Savings from all the other use cases will more than pay for the luxury.

Luckily all chromebooks come with multitouch touchpads which are perfectly capable of handling pinch/rotate gestures.