I tried playing a couple of Xbox 360 Kinect games with my four-year-old daughter, Phoenix. We had less than stellar results.

The first game was “Sesame Street TV.” Basically it’s interactive Sesame Street. We picked it up from the library to try it and I’m glad it was free.

Problem 1: She’s very small compared to me. If the Kinect sees me, it somehow stops seeing her. And vice versa - if it sees her, it stops detecting me. There seemed to be a sort of very small “magic area” in the room where it’d find both of us.

Problem 2: The interaction for that game isn’t constant. It’s more like: they sing a song, then you have a small bit of interaction, then they tell a story, then there’s a small bit of interaction. She’ll watch or she’ll interact, but she loses interest in interacting once you switch to watching.

Problem 3: Slight misrepresentation of the game on the box. The concept behind the game is like you going into the TV and being on Sesame Street. There is a picture on the box to illustrate the concept. Phoenix wants that to be the reality. It is really hard to explain that the box just shows an idea of what it’s like, that you don’t really transfer yourself into the television.

After a bit of Sesame Street, we tried “Kinect Adventures.” I did this thinking that the constant interaction would keep her engaged.

We still ran into the problem where there was basically the small area where it recognized us both, but then it was compounded with a couple of new problems.

Problem 4: Many of the games aren’t obvious to four-year-olds. In particular, the game where you have to walk from side to side and jump to control the raft - that was entirely unintuitive to Phoenix. She was far more concerned with whether or not the avatar on the raft actually looked like her, which then led to a half-hour diversion where we had to set up an avatar.

Problem 5: Auto jump-in/jump-out. The ability to jump in and out of the game quickly is great for folks that “get it” and when you have a properly sized room without the “magic area” where you’re recognized. However, every time Phoenix accidentally stepped out of the “magic area,” her avatar would disappear because it thought she was jumping out of the game, at which point I’d have to try to convince her to come back into the area - but not too close to me - so we could continue.

In the end, we decided it a better idea to just go watch some Looney Tunes cartoons we picked up at the library. Which, now that I think about it, is sort of the opposite of what Kinect is trying to get you to do - get off the couch and be active. Hmmm.

Over the years I’ve posted about my home media center developments. Back in 2008 I posted a summary with links to articles, then I did another roundup in 2014.

The problem with this sort of periodic summary is that it’s hard to get an accurate picture of how things are working right now. I might forget to blog it, or I’ll take some notes on something I found and forget to post it, or whatever.

I was keeping my media center and home networking notes in a personal wiki on PBworks but I figured it was time to make things a bit more official.

My media center and home network documentation is now live at illigmediacenter.readthedocs.io

This is the place I’ll add notes or tips on how my media center setup works. I’ve got everything from the hardware I use to my process for getting video content into the system. I’ve got my plan and analysis for how I cut cable including cost breakdowns and options. It’s all on this site.

My biggest problem in getting my media center going was that I didn’t know what I didn’t know. Information about all this stuff - hardware, software, how to get things done - is spread out all over the place. I never found a complete guide to help me on my way.

I hope this documentation can help you jump start your media center or improve the one you have. As things change in my system, I’ll be keeping the documentation here up to date so it should always have the latest info.

We finally did it: We cut the cable.

On Friday, we took all the cable boxes back to Comcast, cut off the cable TV and the phones, and we’re down to internet service and mobile phones only.

I have to say, I know I’m only a few days into it but I haven’t really noticed it. Aside from calling my various financial institutions and utilities to change my phone number with them, it’s pretty status quo. We were already watching most of our stuff on demand or through online services anyway.

If you’d like to know what I did or how I did it, I documented the whole plan. I’ll do a blog entry later for the official release of my media center documentation site, but you can read over there about my cable cutting plan: what we did and the equipment/services we use.

NDepend 6 was recently released with a ton of new features. I’ve been working with NDepend for quite some time (my earliest blog entry on it was for version 2.7) and every release gets better. It’s been a couple of years since version 5 came out. What’s new?

The first thing you notice that’s new when you start things up is the additional integrations they’ve added. It used to be just “install the extension for Visual Studio” but now there are icons for TFS, TeamCity, SonarQube, and Reflector integration.

I’m particularly interested in the TeamCity integration because that’s the build server I use. I have manually integrated it in the past using MSBuild and some manual TeamCity configuration, but with the new add-in, I can just drop NDepend on the build server and have all that work done for me. There’s even a specific NDepend build step type added and the report magically shows up in the dashboard. There are some great step-by-step walkthrough videos on the NDepend site showing how to set this up.

I decided to analyze some of the new code I’ve been working on. It was pretty easy to get my project started. I love how NDepend helps you figure out where to go next if you haven’t used it before.

The report has improved by adding “how to fix” information to rule failures. One of the challenges I’ve had in the past is that you could see what things might have failed a rule, but you didn’t really have anything clearly “actionable” you could tell folks to fix. You had to kind of “know” what a rule meant. Now there’s no guesswork.

One of my huge complaints with other tools (coverage, analysis) has been addressed - handling of async/await methods. A lot of what I’ve been working on lately has been Web API code, which is async/await from the ground up. Have you ever looked at that stuff in a decompiler like Reflector? Or a code coverage tool? I’ve found you don’t get any information on it (“Let’s just omit it!”); you get incorrect information on it (“You don’t have full coverage because you didn’t cover all the cases in the generated state machine!”); or you get confusing information (“I’ll show you all of the compiler generated methods that don’t make sense!”).

Reports are very clean and complete, but you don’t see the compiler generated state machine junk. Finally!

The metrics view just doubled in value by adding a second “dimension” to its display. You used to be able to just change the size of an item in the view based on a specific metric; now you can compare one metric to a second metric by adding a sort of “heat map” style coloration to it.

My favorite combination so far is to set the box size by “# IL Instructions” and set the color of the boxes by “IL Cyclomatic Complexity.” It gives you a pretty good indication of things that need to be refactored - just look for the huge red boxes!

My favorite new feature is the shareable rule files. We have a standard FxCop ruleset we use on all of our projects. We have a standard StyleCop ruleset we use on all of our projects. We can finally have a standard NDepend ruleset we use on all of our projects.

You can create a rule file with all of your analysis rules stored outside the project file and then tell projects to reference the central/common NDepend rules file.

Once you have a custom rules file, you can reference it from your project. You will probably want to switch the paths in your project to be relative to the project file so it works on your machine and the build server.

With every iteration, NDepend just gets more compelling. I get so much insight from it about our code and areas we need to improve - things that are hard to see when you’re neck deep in code and NuGet package references and under a deadline. You owe it to yourself to check it out.

Full disclosure: I got a free personal license from Patrick at NDepend. However, we have also purchased several licenses at work and make use of it to great benefit.

I use LastPass for a lot of things including storing my personal software license files. I use the “secure note” function to save the license information and attach the license file to the secure note.

I was working on something today and trying to save a license to my machine and kept getting a dialog saying, “Error opening attachment. Error C.” Nothing really specific and very confusing. I was able to save the attachment from the LastPass web site but not through the browser extension.

I ended up finding the solution in this forum post.

1. LastPass Icon > Tools > Advanced Tools > Clear Local Cache
2. LastPass Icon > Tools > Advanced Tools > Refresh Sites

After doing a clear and refresh, the attachment saved correctly. These are probably good steps to try whenever you get any sort of error with the LastPass browser extensions. Filed for future reference.