Or by my own calculation, 41.32 lightning bolts! Great Scott!

Here we have an identity crisis within the linux community, and I find it distressing how few people see the underlying opportunity. The decision between X and wayland/mir depends on what you think linux is. Is it an industrial-strength swiss army OS used by the technically inclined, or is it the platform upon which the tablet renaissance is being built? Of course it's both, so quit with the civil war and pay attention to what's happening in computing.

If general purpose computing is going to survive Apple, Microsoft and Google, we need a rich, high-performance compositor that can run on embedded devices AND a next generation framework for network transparency in applications, preferably in separate packages. Since I'm being dragged into cloud computing, I want to become my own cloud: I want to blur the line between my laptop, server, desktop, and tablet, but I want to do it in an open-source, platform agnostic way. I want to leave my CAD software running on my desktop and connect to it from my tablet to get dimensions for some part. I want automatic syncing ala dropbox for my LAN. I want to stream audio and video to my stereo without using airplay. I want generic compute jobs to be distributed to idle computers on my personal network. I want to lease an EC2 instance just for the week that I have to do some high-quality rendering and have my desktop parcel the job up and send it out to be executed with a minimum of manual plumbing.

In other words, I want network abstraction for input and display, a toolkit to aid with responsive UI design, local openGL compositing, a framework for exporting big, blind compute jobs, and some network utilities to help me get my services configured correctly, and I want them to be designed to work well together. Some of this is Hard but all of these technologies already exist in some form, they just haven't been integrated into a single open-source platform. Usable by consumers. Yet.

The open source community has the opportunity to stake a claim while the world of computing has been turned on its head. Fretting about X11-style network transparency at this point is like sweating over the future of IRC. (Hint: all my chatroom correspondence is now owned by some shitty company overvalued at $27 a share). When all new software is designed to run on top of webkit, will your remote GIMP even matter?

I'll say this as a cynical adjunct: the instructors who are the most integrated with CMS are the instructors who are likeliest to be replaced by a MOOC. Not to discount online learning, but since I prefer it the old-fashioned way I've changed my approach to emphasize the strengths of conventional classroom instruction. My IT needs are a lab, projector, audio system, LAN file share for course materials and submissions, and a whiteboard - anything more is likely to be more trouble than it's worth.

I love this question! Let me help you pick an oscilloscope.

If you're not designing a motherboard, but instead working on medium-speed (Agilent MSOX3054A. If $12k is too much, you can get the "lower" grade MSOX2014A, 100MHz, 8 digital inputs for an almost reasonable $3100. Agilent is the heir to the classic Hewlett Packard geeks all know and love, and the infiniivision x-series compares favorably to tektronix, dollar for dollar.

You need a multimeter. Just get the Fluke 87 and forget the rest. It is also useful to have a function generator and a frequency counter, even if you are doing low-speed digital/analog work. If you're doing RF work or designing analog amplifiers, you might also want a spectrum analyzer. These can get freakishly expensive depending on the type of work you want to do. If you didn't get a mixed-signal oscilloscope, look into getting a logic analyzer. For professional applications it's probably cheaper to just get the MSO.

Not as high tech, but equally important: a soldering station and a fume extractor. For working on mains-powered equipment, you are going to want an isolation transformer and potentially a variac. You will need some way to mitigate ESD in your lab, so look into grounded tables, heel/wrist straps, and ionized air blowers. Not the consumer grade stuff, you want something that senses static charge.

If you are the DIY type who doesn't mind getting your hands dirty, it is nice to be able to fabricate your own PCBs. You can get PCB mills but they are expensive and low-precision. Better to do it the old-fashioned way, in a printmaking studio. You need a UV exposure unit, a laser printer or inkjet that can lay down a high density of ink on a transparency (no recommendations, sorry, I'm still trying to find a good one myself), an etching tank with aquarium pump, and a sink. Just buy your PCBs pre-sensitized. For cutting and drilling you want a small bandsaw or shear and a drill press. You can Harbor Freight the bandsaw, but try to get a high-quality drill press with no spindle wobble or else you'll blow through drill bits like crazy. If you're OCD about this you might even consider getting a small mill, just make sure it has enough of a throat to handle the boards you want to work with.

If you want to do SMD in house you'll need a reflow oven, solder paste, some tiny tools, and possibly a low-power stereo microscope depending on your visual acuity.Throw in a hot-air reworking system too, they're essentially heat guns with chips that match common SMD packages.

You are going to be accumulating many, many tiny parts, so storage is essential. It will need to be versatile: bulk resistors are one thing, but over time you will accumulate reels of SMD parts, coils of wire, long plastic tubes full of DIP and other through-hole parts and all kinds of other junk. It has to go somewhere and stay organized. In my lab I use the tiny trays available at model shops for stuff that's not ESD-sensitive, but in pro labs I've seen large metal cabinets reminiscent of library card catalogs and flat files lined with anti static foam organized with mazes of dividers.

Finally you'll need good EDA tools, so be prepared to pay good money for professional grade software. Most of it runs on Windows so budget for a new PC as well. Cost is no object? Look into Altium Designer. It's the kind of software that doesn't have a listed price. Eagle is a realistic option although its user interface is like taking a time warp into the nineties. No free software I've tried can be cajoled into doing what I can routinely do with Eagle, so my advice is to accept that you'll be using a proprietary toolchain and budget accordingly. Good luck.

The photos of the Z machine have to be seen to be believed, and even then, it is grade A sci-fi: http://www.sandia.gov/z-machine/ The "Z pinch" is an alternative method of containing the hot plasma. Tokomak reactors use magnetic confinement of a continuous plasma, while the Z machine uses inertial confinement for shorter lived plasmas. IIRC the web of lightning shown in Sandia's publicity photos is produced when thousands of tungsten filaments are vaporized in order to generate x-rays. The fuel pellet sits in the center and the X-rays compress it into criticality -- if it sounds like an H-bomb, that's because it probably is.