The Epic Fail of the FDA

This week, I received my first insulin pump. After nearly 27 years treating my Type 1 diabetes with shots — sometimes six or more of them a day — I’m moving to the high tech solution to give me better control. It’s a neat system, complete with a Continuous Glucose Monitoring device that interacts with the pump, and a new glucose monitor that can also transmit its data to the pump. We’re getting baby steps closer to something like the artificial pancreas that people have dreamed of for decades.

It’s not without its shortcomings, of course. All the usual proprietary crap creeps up in this corner of medical commercialism, too. Namely, competing companies marketing competing products have proprietary ways of getting, transmitting, and storing that data. There’s lock-in. The best CGM isn’t necessarily compatible with the best pump. And there’s no third party application that could monitor all of this for you, because those medical companies are protecting their research and development expenditures by keeping their secrets and protecting their data.

Damned frustrating it can be, but it is what it is.

Wait for it — the computer angle of this is about to start.

My pump came to me from Medtronic, which also supplied a compatible blood glucose monitor and CGM. I have all the parts of the system I need with full interactivity. Starting at this point in the evolutionary timeline of diabetes control means I get to have some of the benefits of being a late comer.

The center of the system is a website named Carelink, which can retrieve all your readings — blood sugar readings from both the CGM and the glucometer, and insulin levels from the pump — and put them in one easily accessible place on the web. Charts, graphs, histories, etc. It’s all there for you and your doctor to see. You also have the option of letting Medtronic see it so they can continue to gather data on how well their devices are working. I’m fine with that, so I opted in.

So now, let’s get to the amazingly frustrating thing that puts everything else so far to shame.

I went to sign up for Carelink the other day, only to get this error screen:

I’ll type out the relevant portions for you:

Required Settings: Windows 8 users: Close any browser launched from the Start screen and open your browser from the desktop.
Windows XP/Vista/7, Windows 8 Desktop, MacOSX 10.5 (Intel), MacOSX 10.6/10.7/10.8”


Required Settings: IE 7/8/9, IE 10 Desktop (Windows), Safari 4/5/6 (Mac OS), Firefox 5”

All the power of standardized web browsers and web languages brings us to a system that lacks a clear equality between Mac and PC users. Suddenly, I’m stuck in in 1995 all over again. (I’d put up a fight for Desktop Linux users, but I doubt either of them are reading this.)

If you’re a PC person, you can be using Windows 8 and Internet Explorer 10. If you updated to 8.1 — and I haven’t read a review yet that says not to — you’re out of luck. If you use IE 11, you’re done, too.

And let’s face it, if you’re running a Microsoft box, you better damn well run the most up to date versions of all their software if you want to have any chance of having a secure computer.

If you upgraded to Mac OS X 10.9 six months ago when it came out, you can’t use Carelink. If you use Safari 7, you’re toast. You need to be running Safari 6 — a web browser released in the summer of 2012 — to be compatible.

Your other option is to use Firefox 5. That’s only about 23 versions out of date.

Chrome is not an option.

(Neither is Opera, but I think there are more Desktop Linux users than Opera fans at this point…)

I expressed my displeasure with this ridiculous state of affairs. Medtronic tweeted back:

I should have guessed it would be a government agency protecting me by not letting me use the best, most secure, and most compatible browsers and operating systems on the market. I guess I’m spoiled. I pay attentiong to the Mac software community, where people trip over themselves to get their products in the App Store on Day One of a new operating system release. They test on the betas, do a last minute quick check on the release, and they’re good.

With the FDA stepping in to “protect” me, they’ve not only slowed down the process, but they’ve managed to make everything worse.

Needless to say, this greatly shakes my faith in this software. What’s the FDA’s role in approving this software? Security? That’s a joke, because every browser they’ve approved to be used is less secure than current ones; every operating system they don’t approve is more secure than current ones. By nearly three years.

This is a joke.

I can download an old copy of Firefox and run it off in a corner on my computer at home, but I can’t change the operating system. I guess I could go out and buy a copy of VM Ware or Parallels and dig up an old disc from Snow Leopard, maybe? But that seems like an awful lot of work to go through to dumb down and downgrade my system for the sake of a website.

Unfortunately, there is a solution. My wife has a Windows 8 machine that she hasn’t upgraded to 8.1 yet. As of today, she won’t be. I need it now.

This is ridiculous. By now, this should be approved for the most recent operating system, at the very least. It’s not just an inconvenience, but an actual security threat. Your federal tax dollars at work…

Elixir Talks Go Live

The recent Erlang Factory San Francisco conference featured an Elixir track on one of its two days. The videos for these talks are starting to appear on YouTube now.

Judging by the videos, this track of the con was held in the janitor’s closet in the back of the conference center, with someone’s iPhone pointing at the screen. It’s fun being the red-headed stepchild at a conference. But, hey, after telling the Erlang folks all the stuff they’re doing wrong, you didn’t expect bright lighting and full post-production on your talks, did you? Nah.

The first two talks up should likely be watched in reverse order if you aren’t familiar with macros yet. I’m not.

Here’s the “Write Less, Do More (and Have Fun)” talk that deals with Macros:

Then you can go back to the “Pipe Dreams” talk which is mostly about how to reimplement pipes with macros for better error handling:

More to come!

The Amiga 3000

A quick trip down memory lane…

I grew up with a Commodore 64, but always wished for an Amiga. Those things were powerful, right? They even had models with hard drives in them so you didn’t need to switch floppy disks out of the disk drive so often to load a word processor or whatnot.

Here’s a walkthrough, 20 years later, of the Amiga 3000 (released June 1990), from the paint programs to the web browsers. Pretty impressive for a machine that old.

Read. Butt In Seat. Code.

Book learning is easy. Book learning is great.

It’s also useless on its own.

You don’t know the questions you need answered until you sit in front of an open editor with a blinking cursor and attempt to get something done. That’s when the book learning guides you to go to the right places, ask the right questions, and discover new things.

The book is there to get you excited to sit down and start hacking away at some toy code. It’s a learning process, not necessarily a production one. Don’t start by trying to produce your application. Start by following the examples — TYPING THEM IN, not DOWNLOADING them — and then messing around with them. Repeat things. Try completely crazy new things. Does that crazy idea work? Give it a shot. The worst that can happen is that you’ll get a syntax error.

Experiment. Play. HACK.

Learning vim

I’m a vi editor user kind of guy. It’s just the way I started in college, and is the way I’ve always felt comfortable. I’m happy the Emacs users are happy, but it’s not something I’m interested in converting to right now, thanks.

I’m getting serious about how I use vim lately, though. I actually booted up the vimtutor for the first time in my life and am going through it. I’ve learned some neat things. Most of it I already knew, but there are some little nuggets that I plan on using in the future. Some examples:

  • rx

Replace the current letter with an ‘x’ (or use your own character here)

  • cw

Delete the rest of this word and go to insert mode

  • :set ruler

Gives you that neat line number/character position stat at the bottom of the screen always, as well as what percentage of the way into the file you are.

And here’s where things get really embarrassing:

  • :!

Runs an external command. I’ve always just saved out and run the command, or kept a second window open and switched over to it to run that script. But I can do it from inside the vim session, which is huge when doing test driven development. Heck, it’s so huge that I plan to start setting up my tests in my daily Perl programming in a better way to take advantage of this.

Then I’ll simplify it with a keyboard shortcut either through Text Expander or the .vimrc, which is something else I’m still fiddling with.

But you can also grab a quick directory listing or cat a file or something else. It auto-completes the file names, too. Lovely.

  • v

Visual mode. Command mode and Insert mode I had the hang of. Visual mode was completely foreign to me. Being able to highlight a bunch of lines and then save them to a separate file? That’s awesome. To do that:

  • :w filename

…after you’ve highlighted what you want to save out.

Next up, I’m going to peruse some other folks’ .vimrc files on Github, watch that Ben Orenstein vim talk, and dive into Derek Wyatt’s ridiculously large number of vim editor tutorial videos.