Tags

, , , , , ,

I’ve mentioned before that we’ve used FPro as a type of Rapid Application Development Platform. Building a fully functioning prototype using this high-level (largely graphical) database application has allowed us, as non-coders (albeit with a decent understanding of how a properly architected database should be structured), to develop our therapy-smarter tool in an iterative way, without spending large amounts of money.

Another benefit is that the product we are using remains fully accessible to us at the architecture and data level, and is largely self-documenting. FilemakerPro produces human-readable graphical representations of its database structure as a standard report and the functional scripts are basic-like in their use of real language, and thus easy to follow.

What this means is that, for a team like ours, without coders, the scary step of commissioning native code from people whose work we will ultimately not be able to judge in detail is made much less risky. When we judge that the time is right to take the system to the next level – when we need it to go faster and handle more clients, we will have to get the thing recoded – and this will mean bringing developers in.

For startup teams without coders, this is a terrifying point. How do you find the right person? How do you choose which framework, which language, which platform, which architecture? How can you even begin to judge the recommendations you are hearing? How can you describe the features you want?

The risk is enormous – this is costly work, at the guts of your business, somewhere near the cutting edge of technology, in a fast-changing field, and where non-coders are effectively only able to contribute or engage at a very superficial level.

But with FPro, some of this risk goes away. We can point developers at a system that functions more-or-less exactly as we want it to, and say – build me something that does this, that will run fast and scale well. The terrifying inefficiency of writing a functional specification – of trying to capture a dynamic, multi-valent application in writing is largely eliminated.

Pukka coders might be a little sniffy at the idea that an FPro implementation is worth the effort. Our system was referred to as ‘just a proof of concept’ by a developer I talked to, and although I would heartily disagree, it has to be said that there are some serious issues to consider.

A major one is the provision of services on mobile devices.

FPro has had a method for producing a mobile app for a while now, and this is an integral part of our service. But you could hardly call it perfect. The client has to install a standard FPro app (FilemakerGo) onto their device, and then call up your app from within that. Although this can be streamlined a little, it remains clunky and vulnerable. And of course, Filemaker being an Apple subsidiary, it only works on iOS.

And that’s what is so interesting about a recent announcement from Filemaker., to the effect that they have released an SDK to allow native iOS apps to be developed using standard Apple xcode developer tools. There are some caveats, and no-one has yet taken an app written in this way through the app store process, so time will tell, but this can only increase the viability of our approach. It also offers the tantalising possibly of porting the code to run on android devices – again, untried as yet.

I can’t finish this post without saying that another advantage of using FPro is that because it is a proprietary product with a long evolutionary history, there is a strong community of expert freelancersĀ  who are used to dealing with systems written by others – aided by the stable platform and the self-documenting nature of an FPro solution. This means that you can get really useful work done on a ‘by-the-day’ basis, without worrying that the next person who looks at it will spend longer trying to understand what’s happening than doing useful work.

That said, all our freelancing work has been done by the highly productive, efficient and responsive James Glendinning of Glenavon Technology – I can heartily recommend his work.