The surrounding metro areas may not have the dirt cheap rent like the city, but it is still going to be *much* cheaper than places like San Francisco. And if don't want an hour commute into the city, try living some place closer to the city than Troy.

Note that this article is about "metro Detroit", not "Detroit". Plenty of safe places to live in the Detroit metro area, especially on a tech worker salary, they're just outside the city proper. And even if the jobs were in actual Detroit, it's still possible to commute from outside the city. But whatever, it's an article mentioning Detroit, let's just bash Detroit.

Well, the key word in my original post was *eventually*. Stuff like ESPN3, MLB.tv, and March Madness on Demand work already, I don't think it's too far out there to either increase bandwidth or develop more efficient protocols to handle more customers in the not to distant future.

I'm not familiar enough with the inner workings of a CDN to say specifically, but I'm pretty sure something similar can be set up for live events. Instead of sending a single copy of an old show to a CDN to be distributed to all the downstream end users, a single live stream could be sent to a CDN equivalent and that could be forwarded to all the end users. Of course, I'm just a lowly end user spitballing here, so maybe I'm missing a huge hang up that would prevent this from happening.

Not now, no. But eventually, once we all get hooked up to Google Fiber or equivalents and CDNs are beefed up, maybe.

I'm not surprised that the cable and satellite TV companies want to their branding and interface in front of Amazon and the like. But I thought the point of these boxes was so that eventually you don't need the cable and satellite TV companies and get everything steamed over the internet to the set-top box. Cable and satellite TV companies can't control the interface if you don't use their services.

We could go back and forth forever, but if you are basing this all on smartphones being too unreliable and fidgety maybe you need a new smartphone. Or a tablet. Because we never have any issues using our Chromecast with our phones and tablets.

People are plugging the smartphones and tablets into the wall regularly due to regular use already, very rarely will you want to use it and have it be dead. And if you already have and use a smartphone or tablet, another remote is just another thing to get lost. And given my experience with the first generation Roku, I'll take a multitouch screen over arrow keys at least until the remote comes with a keyboard.

Good lord what a moronic piece of drivel. It screams "I have no idea how this product works, but I won't let that stop me from bashing it!"

• Are you the only person in the world without a charger in your house? If the battery on you phone or tablet gets that low, you can charge it while watching the show.
• It is possible to share smartphones and tablets. Somehow multiple people can control a TV despite only one remote control.
• Any device can control any Chromecast connected to the same wifi network. And can stop, pause, or rewind a program started by another device.
• And the worst, most blatant example that you have no idea how the product you are bashing works. When it comes to streaming shows from the internet like on Netflix or HBOGo or whatever, Chromecast doesn't actually stream from the device that started the program. The device gives the Chromecast the location of the stream, and the Chromecast accesses it directly. If you need to run to the grocery store, it keeps playing even if you take your phone with you.

A someone that easily uses a Chromecast with my wife and children, you need to shut up until you learn about what you are bashing.

And this is why you should buy a dumb TV and just use it as a display for smart devices. Whatever is added on to the TV is usually obsolete or dead long before the TV and can't be replaced, but a box or stick can. This has been the way to do it since they started coming out with TV/VCRs, and it's even more true more that input devices are advancing so much more rapidly.

But how many people that would be interested in a set top streaming box won't already have a tablet or smartphone to use as the source? Not exactly a large market there.

I've been running Linux on my home computers for about 8 years, starting with Ubuntu and switching to Linux Mint + Cinnamon to get away from Unity and Gnome3. I've had some issues with hardward peripheral support (less lately), software availability (less lately with everything moving to the cloud), and Office documents getting mangled formatting as they went through Open/LibreOffice and back, but as long as your Grandma doesn't do anything of these things and just uses the browser for surfing the web and email she should be fine. Linux Mint also comes with a lot of multimedia support built in - I was able to add the same stuff in Ubuntu, but it was more convenient to have it from the start in Linux Mint, so you're less likely to have to come back and figure out why some song or video won't play.

I have a French press at home and it's not that hard to use. But when I'm at work and I just need a cup of of coffee (which is often, at least until my son starts sleeping through the night), there is nothing as quick and convenient as my Keurig. Turn it on and a couple minutes later I have a cup of coffee with no mess to clean up. No, the quality of the coffee won't be as high as a press or other methods, but I'm not a coffee snob so it's fine as far as I can tell.

If they know the concepts behind a relational database, training them to write SQL queries should be easy. If they know the concepts behind the OSI model, training them to use a packet sniffer should be easy. Teaching students narrow topics like SQL queries or how to use a damn sniffer is very limiting, teaching them the theories behind them is useful in many different fields. Refusing to train your employees and expecting them to know very specific topics is very short sighted and limiting. Expecting your employees to know high-level theories and then training them for the specific needs of a particular job gives you much more adaptable employees that will be more beneficial in the long run.

High-level theories and models and UMLs and chess board Java CS projects are most definitely useful to 99.9% of tech employers. High quality employees will apply those concepts to write efficient, maintainable code regardless of the particular language being used on a project at a tech company. Once the students have learned those high-level concepts, learning how to apply them to the syntax of a particular language is the easy part, and the student will also be able to go out an learn multiple other coding languages with ease. Like the file out question you mentioned: they might not know how to do it in a 4-line script, but they will know the general algorithm behind how to do it (or at least they ought to if they've learned high-level theories). Teaching them the syntax for it in a particular scripting language is a minimal investment after they are hired.

Techology evolves, software suites come and go, but the theories will be always be applicable. Unless you want the students to become useless once a particular software suite becomes outdated, you teach the theories that will always be applicable. Tech employers that hire solely based on knowledge of a particular software suite are very short sighted and get what they deserve.