Over the past week or so, there’s been noticeable concern amongst the CGM in the Cloud and Looping communities, all part of WeAreNotWaiting movement, at what appeared to be a crackdown by Heroku targeting them. Many woke late last week to see the following:
However, this isn’t a focused attack on those users from that community. It appears to be a glitch with Heroku. It served as a timely reminder, amid the outcry, that the use of free or sandbox services comes with a little recognised cost. No resilience and in reality, no contractual obligation to provide service. There is no tort involved, even if you are using it for what you consider to be a lifesaving service. This is something that is never at the forefront of people’s minds when they join in with something like this.
Are we taking a sensible approach to this?
Let’s consider that last point carefully. You are using Nightscout (open source software) to monitor the glucose levels of someone you care for. You are concerned that without it, then they may be at risk of hospitalisation. You pay for Dexcom, or you’ve paid for things to convert a Libre to a CGM. You’ve paid for an Apple developer’s license, or a RileyLink, or a phone. But you don’t want to pay for the monitoring service, even though you think that’s a critical part of the infrastructure.
Let that sink in for a moment.
While I recognise that it’s an unpopular opinion, running a service that you consider life-critical on a free platform with no resilience and no obligation to maintain that service is perhaps not the best thing to do. And that’s without the other costs associated with WeAreNotWaiting software and services, which may not be obvious.
The real costs of WeAreNotWaiting tools
Let’s take a step back and look at where NightScout started. It was a user interface, hosted in the cloud to allow a user to remotely follow their child’s Dexcom G4 CGM.
Since then, NightScout alone has grown into a very capable monitoring system able to monitor and generate reporting around glucose levels, food consumption, monitoring of closed loop systems (both DIY and Commercial) and as a common source for a lot of data for a lot of apps. The number of apps and solutions that have developed with NightScout as a key component has continued to grow.
Maintaining the source code
And of course with that expansion comes an increase in requirement for support, maintenance and updating of the software, as well as troubleshooting and a requirement to align different groups. All of which is done by individuals, in their spare time, unpaid.
Many rely on these systems to provide data based around commercial APIs, such as the Medtronic Cloud to Cloud Bridge or Dexcom’s share service. All require not only maintenance in line with the APIs, but potentially lobbying and communication with the manufacturers to ensure access remains available. If we bear in mind that each of these manufacturers operates within their own silo, and can and will make changes based on internal plans, they are unlikely to take data interoperability with DIY solutions into account, breaking things for many.
There’s quite a long held view in the DIY user community that these changes don’t matter too much. Someone will step up and reverse engineer a solution. And yet, some changes may not be possible to reverse engineer… This requires some of those participating to get involved in working with these providers to advocate for access and continuation of access. Something that isn’t an easy sell.
But what’s the point of all this? It’s that many people have given up their time and effort in order to make these tools available. But not only available for those of us who use them to do so, but also in making sure that manufacturers, regulators and healthcare professionals are aware and bought in to what we’re doing and understand why things need to be a certain way.
All of this takes time. As does giving your time to support users on Facebook and the other forums that are used to disseminate information and help users. Whether that be updating system documentation (an area of much complaint and precious little participation) or sitting in Telegram, Zulip or Gitter and helping answer user questions. You’ll generally find a small group who do this regularly while a much larger group are able to benefit.
The real costs? Time and manpower. However you look at it.
So what’s your point?
We have users who aren’t willing to pay for the support of and access to services that many consider to be life changing, if not life saving. We see concerns when people try and provide commercial versions of WeAreNotWaiting services. And then we see complaints when free services don’t maintain an expected level of support. But if we’re willing to pay for CGM, why wouldn’t we pay for other life-saving services?
Is it any surprise that no large scale producer of Diabetes care systems is interested in developing a real time, shareable, product agnostic tool? As a community we’ve done an excellent job of demonstrating that we don’t think we should pay for it and therefore there’s no money in it. While some might argue that well, no, there shouldn’t be any money in it, we’ve also demonstrated that we don’t care enough about the service level we get from the various providers we use to provide our own free services to even hand over a credit card that never gets charged, let alone pay for a service.
We’re happy to let our healthcare plan or service pay for it for us, but if we do that, we shouldn’t be surprised that our choices diminish.
Other paid for services?
And if we want that choice and level of resilience, then we either pay for a patchwork solution that we ourselves build, or we outsource that to others, such as Sensotrend
but we shouldn’t complain about them offering paid for service. If we want a reliable, supported environment that is far less likely to suffer the vagaries of a free service that has no service level agreement with us, and no requirement to maintain an uptime level, then maybe we should consider that payment for a service is worthwhile.
What are our real reasons?
Ultimately, if we pay for a CGM system because we think it’s life-saving, but we don’t want to pay for a remote monitoring system that we also consider to be life-saving for our cared ones, then we can’t complain when that remote monitoring system disappears. Whether that’s due to the service provider or dwindling numbers of people involved in maintaining it. WeareNotWaiting shouldn’t mean that WeAreNotPaying, and if we want service levels and hardware support then maybe, just maybe, we should consider paying for it in some way.
The Heroku “outage” that prompted this should serve as a timely reminder that WeAreNotWaiting exists because of the people involved in it. For it to continue, people need to get involved and use their skills to help. Whether that’s reverse engineering technology, lobbying various groups, investigating and sharing new algorithms or providing support to users.
Without all of this, there is no WeAreNotWaiting community. There is only a group of increasingly frustrated WeAreNotWaiting users. Who will soon once again be waiting…