Ok so there's a lot of talk on this and rather than copy pasting I want to get a single short thread I can point people to.

To get the obvious out of the way, yes it's disappointing that not all tests are loggable. However given cases are on the rise I understand releasing early https://twitter.com/NHSCOVID19app/status/1309446092057202689

We also need to acknowledge the people working on the app are months behind because of the government wasting time on their own failed approach.

With that out of the way here's a quick primer on how agile software development works.

You gather the features you want and know about for your product. A "product owner" is in charge of prioritising these for the Dev team to work on.

Some features or specific functionality is only well known as you progress.

Some require other people to do work.

How you require other people to do work might also not yet be known enough for them to start.

Each of the different test providers is going to be a separate third party team that needs to do work and coordinate with, which is time expensive.

In this sort of circumstance the best way through is usually to build your minimum viable product. Prove the features work, integrate with one third party that has the best trade off of volume and complexity.

Based on the work of that initial integration you discover unknown issues and define standards, which then make it easier to work on the other providers.

Knowing the other providers will also take time you might prioritise beginning that process part way through.

The NHS itself is actually very complicated with a mix of different systems and providers of their own, while there's work towards common standards it is fragmented. I'm not sure on the state of things like this sort of testing to know the fragmentation for this.

Agile software development is about getting slices of value released rather than waiting to get everything at once.

This tends to get you to full features earlier and means you start getting value sooner. Add in the current situation and launching early is the right move.

Now we're into assumption territory. Why choose the private testing as the first integration?

I suspect they are one of the largest single systems to integrate with in terms of coverage. And as an organisation are focus on Covid so their priorities are easier to align.

Why is the NHS harder?

I don't know the specifics here so going on general knowledge. Often different trusts have different electronic patient records, which integrate with different patient management systems, all by different providers with many different customers.

That means many third parties who have to talk with their own third parties to prioritise a feature that may be easier or harder to implement on different systems, each getting a small amount of coverage.

Note I don't know specifics on test result systems so this may not apply.

Finally there's expectation setting on launch.

Comms had to balance getting installation coverage vs setting expectation for test submission. I think they could have done more on the latter but made the right priority call.

This part is never easy.

Also to note, the exposure notification system requires days of use before it starts really getting value. A positive test today only notifies people using the app since Thursday. Onboarding a new test route in a fortnight means a positive from them notifies people using it now.

Summary?

Software development is complex.
The NHS is complex.
Development on this app started months late.
They likely made the right choice to launch as is.
It is still vitally important to use the app.
Blame the government not the app or developers for the delays.