A Higher Dimension - Spreadsheets, Pt2

Image - Shutterstock

 

 

Yesterday I outlined the basic problems with spreadsheets and the very real impact they have had over the last forty years, good and bad. The issue of the Covid tests is a real-world example of over-reliance on tools that have long been suborned for tasks they were never designed for, nor capable of doing, reliably.

There was at one time however, a spreadsheet-like application that not only overcame these limitations of the standard spreadsheet, but gave it powerful and most importantly reliable predictive modelling facilities customisable to companies and institutions from the small to the very large. Its key feature was that it dynamically scanned the model as it was constructed, any errors by the operator being flagged in real-time. It also had the singular ability to have a complete sheet embedded in a single cell of another sheet; the possibilities for use practically infinite.

It was produced here in North Wales by the company run by the former colleague I mentioned in yesterday's post. I first saw it demonstrated back in the late nineties, having seen a glowing technical review of the software in a business journal we used to take. It took me all of ten minutes to realise that they had a truly class-leading product that could revolutionise the way that accounting and arithmetical modelling was done. The profit potential for this, properly marketed, was enormous and should have made them very wealthy and perhaps turned the history of the spreadsheet and its pernicious effects on its head.

However, their approach was from a very different angle. They ran the business as a service rather than as a vendor of product, producing large, complex and completely bespoke accounting models for institutions; principally large general hospitals across Europe - that could model the finances of those institutions down to the last paperclip. The institution would then hold a licence to use their model and the company we were kept on a retainer for any updating or modification necessary over time.

The software and the company no longer exist, nor sadly does the chief software engineer who died some time ago. Our former colleague now works with a veteran support organisation; he himself being ex-services.

However, the idea is still there, waiting in the wings for a developer to make real again. This current limbo state applies to all manner of significant ideas that never see the commercial light of day, some being unused features of existing software, like Apple's QuickTime. The power locked away in the original QuickTime API  [Application Programming Interface] and the SDK's [Software Development Kits] available to us as developers was truly staggering, and yet only a tiny fraction of its capabilities was used by Apple. For those who don't remember QuickTime, it was
Apple's attempt at a standard multimedia platform for the Mac and later Windows, although support for the latter disappeared fairly quickly, as has the original incarnation of the framework.

It was essentially a multichannel, frame-based 'movie' format and most applications of it stuck to the basics of movies and audio. But if you looked hard at the API it was obvious that so much more could be done with it. The layered structure of QuickTime made it so that it could import, manipulate, play and export a variety of data forms, including video, audio, still images, MIDI, animation, textual and numerical data and QuickTime VR, Apple's format for immersive photography.

The singular aspect of the format that stood out for me, though was the potential to embed a complete movie in a single frame of another movie file. I could see that in theory you could build a multi-dimensional structure for the storage and retrieval of large and complex collections of multimedia assets, using the inbuilt database abilities of the format for search and cataloguing the assets. I used a technique based on this idea for the prototype and early versions of a project we were involved in at the time, only changing away from the technique for the final turnkey version of the software which was recoded in a well-known general-purpose multimedia package, making future maintenance easier and not dependent on my one-off solution.

This revelation about QuickTime's untapped potential came at more or less the same time that I first saw the modelling software demonstrated and to this day I still feel the synergy between the two was remarkable and that there is still plenty of mileage in it all somewhere. It would be good to find someone out there with the talent and enthusiasm to carry on where we left off. It makes me wonder just how much more little-known stuff is lurking in the dusty corners of the world of code.