Hacker News Reader

Comment by Angostura | Hacker News Reader

I could certainly imagine NHS England looking at this and creating something that hospitals and GPs could deploy

I'd love it if my government created a civil reserve for technology workers. Let me volunteer every weekend to help fix infrastructure so I don't have to give up my existing job.

If it was a volunteer effort without ulterior motives, it would be beneficial to society, but ultimately who is on call? Who pushes for hard, but beneficial changes that might not have immediate obvious value? Who accepts risk or responsibility.

Ultimately that’s the point of the market. Incentivize people to take risks for rewards. Allow others to improve on proven models for lower costs. Unfortunately, government does not have any risk/reward or other market pressures.

It's pretty common for traditional organised volunteer services to have "on call" aspects.

Think of like volunteer firefighters, The Samaritans, St John Ambulance, the UK lifeguards and lifeboats (RNLI). Such organisations do usually have full time paid staff too, but the bulk of the front line work is part time volunteer.

This is on one hand quite the fantastic idea, but I imagine it falls down to bureaucracy, for example for health services related stuff everyone would need pretty thorough training in the legal aspects, and insurance might be difficult, and preventing malicious actors from contributing so probably security clearance (for example if the military use it for their healthcare) or at least a thorough background check. I think open sourcing everything is far far easier than a volunteer based setup.

Sadly, very sadly, I cannot imagine it. I have seen the inside of NHS IT.

Yeh, I could imagine GDS (government digital services) having the capability, but not NHS digital whose remit this would likely fall under. They just don't have the resources or experience.

I've seen inside NHS IT too (currently work in the NHS). It's patchy, but there are some talented people. Given the push for digital - and cost-saving at the moment, I can well imagine a GDS-style core group being started that would 'productise' this for Trusts - and even offer support, the way the Accenture crew does reasonably well for Microsoft products etc

In the US, the Veteran's Administration wrote their own EHR (Vista) which was released as public domain. They've been trying (and mostly failing) to migrate to a commercial EHR for the last seven or eight years.

VistA has some great functionality and end users generally like it, but unfortunately the underlying platform and developer tooling is hopelessly outdated. It's approaching a technical dead end and there's no practical way to keep it moving forward unless someone steps forward with the funding and resources for a major refactoring / rewrite engineering effort.

https://worldvista.org/

Vista is ancient, and it's written in MUMPS, an evil twin of COBOL.

For context, many (most?) other EHRs are too, though they call it M now so it sounds less disease-ridden.

No, MUMPS (or M) is a remote descendant of JOSS, an interactive language of the 1950s. JOSS has all sorts of variants (DEC's FOCAL language of the 1960s was a dialect), but I think MUMPS is the only living one. MUMPS code is mostly unreadable, as the commands can be, and often are, abbreviated to the first letter. As a result, it looks a lot like line noise.

Regardless of its many warts, Cobol cannot be accused of being unreadable. Verbose, yes.

MUMPS was originally developed in the 1960s for use on minicomputers that had maybe 64KB RAM. At the time it was a lot more important to keep code size small, hence the single letter commands. Readability wasn't a concern then but it sure looks like a mess today.

It's hard to imagine it's an improvement over just the raw assembly.

Imagine harder. It was an enormous improvement over assembly language. Not so much for the basic coding but for the portability and built-in persistent data structures. MUMPS has an excellent "NoSQL" database built in which is a pretty good fit for a lot of healthcare use cases.

> Regardless of its many warts, Cobol cannot be accused of being unreadable. Verbose, yes.

Hence the "evil twin" comment :)

I can't. There are engineers advocating for open source in the NHS but management is mostly non technical and will go for a big corporate product every time.

Reply on news.ycombinator.com