Sloppy Systems (or why did spreadsheets and HTML take over the world?)

In his ISOCO4 talk, Adam Bosworth sought to explain the effectiveness of HTML and spreadsheets on the malleability of the medium, and how forgiving the medium is.

In Scribbling in the Margins, Jon Udell’s thesis is that extensibility was the mark of enduring design. Jon likened the extension of DNS records for use in SPF to how people scribble in the margins of a document.

Extensibility is one hallmark of malleability. Margins are all over the place in spreadsheets. HTML, in the real world is not full of semantic markup, but of presentational markup.

A key characteristic of malleability is in the responsiveness of a system to experimentation. For instance, adding 3 and “string” results in #VALUE! almost instantly. Refreshing the browser allows the user to see immediately the results of one’s work.

Another aspect of malleability is in disconnectedness. A html page and a spreadsheet are pretty self contained, and can function well enough without outside references. This makes it easy and less risky to mould.

One aspect not discussed by either Jon or Adam are the openness of the medium. Anyone can view the source behind a spreadsheet or a piece of HTML designed by an exponent, and learn from it. ² It makes these technologies the ideal medium for domain experts, not IT experts. I’ve seen organisations where a piece of long loved Access database gets turned into an Oracle application and it immediately starts to “rust”, because the users can no longer modify the system at will. This is particularly so for Government departments charged with implementing legislative requirements. Sometimes the rules and regulations change so fast, IT departments will have difficulty staying relevant.

Another aspect is the ease with which one could start with a working example and then modify it turns into something else altogether. Subtext is an interesting experiment that brings concepts of spreadsheets into general programming. I’m not sure if I’m quoting Jon again here,¹ but most programmers start learning an API by looking at working code or sample code in the documentation. (Note to platform designers: no matter how hard your customers ask for obfuscation technologies, it’s against the interest of the platform because nothing bootstraps developers like being able to “View Source”, and learn).³

Where does this leave the IT department of the future? The IT department of the future will make it easy for people to organize themselves. The provisioning of applications should not be the exclusive domain of the IT department. Instead, an IT should be more like a hosting service, where employees can throw together php applications, hooked to to an instance of their database, where IT provides the necessary hardware, backup and uptime. IT could provide services so that little groups can secure their little Access application for the few people in the group. ⁴

¹I can’t find the document on the web anymore, it was about launching programs on an NT remotely.

²If html were invented by the commercial rather than the scientific community, would it be as open? I’m doubtful.

³Could it be the fact that Access applications can’t be easily compiled actually made it easier for people to learn how to develop applications?

⁴Interestingly this Gartner analyst called for more agility in software.