Starting MyChart
I was recently asked to talk about a departure at Epic and I’m going to turn that general idea into a series of posts about the early days and months of MyChart. If you have topics you’d like to hear about … please send them!
I don’t remember what month Carl met with me to “ask” if I would like to lead a new patient portal product team. It was early 1999. Yeah, 1999. I can’t believe that MyChart is 27 years old as a team, although it wasn’t released in 1999.
We talked about the general idea for a while and I know I was still leading the Foundations teams and was working on this other new platform that was called “EpicDesktop.” (You likely have heard of Hyperspace … yeah, that’s just EpicDesktop with a new name.)
I was enjoying working on EpicDesktop and leading the Foundations team, so I did ask if I could do all of it, but he wanted me to focus the majority of my time on MyChart.
Why was it named MyChart? Three reasons generally that I remember:
- It was completely following the then “trend” established by Microsoft where the super hip operating system Windows 98/ME had My Documents, My Pictures, My Recipes, My … etc. etc.
- The domain was available and Carl had already purchased it. Seriously, if MyChart.com hadn’t been available, I sincerely doubt that we would have stuck with that name. Amusingly though, even back in 1999, there was an idea that there would be a central destination that patients could go to and then be taken to their healthcare provider’s MyChart servers. As you may know … that didn’t happen (and still is a rough user experience today). That was a HUGE complex technical and security beast on its own and deprioritized for decades.
- It was a reasonable description of the product, albeit boring. Epic wasn’t great at product names … look to “Legacy” for example.
On a positive note, there are worse names that Epic could have selected.
I agreed to make the switch after some deliberation to leading this new team and product. It did cause an itch though which will be saved for a later post.
A few core requirements were established that were very unusual for Epic released software at the time. The most noteworthy and architecturally complex was that MyChart would be released out of band with other software releases and application.
Epic’s applications and systems are a heavily interconnected and integrated and the connections between applications are TIGHT. Dependencies are made and not broken. This isn’t just a micro-service type level of connection, it’s down to specific functions, function signatures and behaviors (and as I think I’ve mentioned, an unhealthy dose of variables and data that is assumed to be available for a running process). This is in the era of COM being the Microsoft Way on the Windows desktop, HL7 v2 (pre-XML, and JSON) for data interchange, and CORBA thankfully failing to take hold of any reasonable market share. Epic R&D had no architectural or technology patterns to support this essential requirement.
I can’t understate how this requirement was very difficult to navigate as it meant that MyChart code and application needed to adapt to many versions of Epic software and be constantly updating and improving, yet have minimal dependencies on existing code. That code was a dependency, and duplicating logic was very undesirable if not impossible. There was a LOT of head scratching and white-boarding.
This wasn’t Kobayashi Maru levels of impossible requiring cheating to solve, but yikes it felt big back then (especially in M/MUMPS). (Yeah, it’s Star Trek’s 60th, so I worked it in.)
This was a BIG FREAKING EPIC DEAL.
I’ll talk more about the beginnings in my next post.
Squeegee 2.0
My wife and I squeegee the shower tiles in our bathroom after every use. I know some of you don’t do this — and that’s fine (but probably causes more cleanup work later if you have a tiled shower). If you do however and you don’t have a decent Squeegee … let me tell you … it’s shower cleanup changing. We had a small 8 inch (20cm) squeegee that was included in a kit …, and while functional … it wasn’t great.
I recently bought a new 14” squeegee (35cm) from Amazon and while it’s not life changing, it works so well and the size makes the process so much more efficient. I don’t even mind doing it anymore! Highly recommended (or one like it). It does not come with any way to motivate you or members of your household to actually using it. That’s on you. The sticky holder for it has worked well and feels very secure on the tile surface.
WLED and ESP32 and mmWave …
Earlier today, I received a box of many hundred dollars worth of 10m RGBW LED strips for our basement hallway ceilings (ugh, tariffs with shipping costs SUCK). Prior to the drywall being finished, I added 2 channels in each hallway and pulled cables for power and control for each (it looks vaguely like a runway on the ceiling).
This week I plan to install the LED strips and several QuinLED · Dig-Next-2 controllers and Mean Well transformers I’ve already purchased (I normally buy transformers at TRC Electronics when I want something outside the norm). The goal … an amazing ceiling display.
In the next post, I ideally will have some pictures of the completed project. Eventually, I plan on having some PoE motion sensors (using mmWave) that will trigger the lights to “follow” ahead of someone walking down the hall. (I’ve got the pieces, just need to assemble and program it all).
Humble Bundles
Time limited of course, but there are 2 Humble Bundles right now that I’ve picked up that I’d recommend (especially for the price).
Even if you only are interested in a few of the books, both bundles are good deals and the books are mostly very recent (and a small portion goes to a charity). Call me old-school if you want, but I still enjoy reading text written by a human for humans to learn something new.