No worries! I'm a government employee involved with electronic medical records, and I tend to get a bit thin-skinned about such things. Sorry if I sounded grouchy! (Off in search of coffee.)

Neither the story nor the summary mentioned anything about government employees. The private sector is just as capable of screwing up as the government is.

You do raise some good points about VistA. There are some folks in the old guard that cling to M, but I think the vast majority of the the VA's OI&T group have realized that M needs to go away at some point.

The problem is that the "D" in DHCP really became a joke. Due to some of the quirks of the M language, and the way they were abused back in the late '80s to early '90s, the whole system is really a big hairball at this point. So, the first thing that the VA has to do is to tease all of the separate subsystems out from each other. These can then be modernized onto new systems (the strategic direction going forward is Java on the middle-tier and client, and either M or Oracle on the backend, the last I heard). And, since the VA these days uses Cache almost exclusively, those old M globals can be mapped to SQL tables when the time comes to grab the data and move it.

So, the Feds can't just help by throwing resources at PatientOS (or anything else, for that matter). Without those slow backend upgrades, you'd be left with a whole load of data in a legacy system that would be virtually impossible to migrate. If the migration is going to succeed, we have to get the data onto something like a standard system first, and THEN we can start looking at options to go into the future.

So, they switched to VistA about 10 years or so ago, and look what Microsoft did.