Title: How about a Turner Prize for software?

Art moves forward because it is readily visible even when it is not accessible. Software like architecture usually hides the internals. Maybe, like Richard Rogers, we should try putting more of our code on the outside!

Source is openly available for FSF and OpenSource projects but only those who participate in the project look at it, and then not comparatively. Most commercial software is locked away where not even the customer can see it.

I'd like to suggest a competition for quality software focused on the code itself. Programming contests have existed for years but these are normally races against the clock. This contest would be run a bit like an art prize. Any organisation may enter provided:

Judging would be by a panel, say five people including a chair, who would assess each entry and decide on a winner. Membership would vary year to year and would be made up of respected authorities, academics and jobbing programmers.

Following the conclusion of the contest the judges' notes would be added to the web site. Entries would remain publicly available and thus serve as a reference and teaching tool. (We may choose to move the entries to a virtual gallery.)

The contest would aim to promote best practices, advance current techniques and provide large-scale examples of quality projects. This is not designed as an effort to further the open source movement; it is intended to improve software quality.

Any organisation could enter as long as the software was delivered to a real customer. Although some companies may object to their source code being publicly available they would retain copyright and in return, they would be rewarded with publicity and a free external review of their project. That said, we might need to allow some elements of a project to be left out of a submission. I don't expect to see Microsoft submitting the Word code but I would hope to see Accenture submitting projects from the Department of Social Security benefits system.

I expect consultancies such as Accenture, Cap Gemini and Logica to be potential sponsors. The prize itself need not be big, say, $1,000. Naturally there will be administration costs but the whole thing should be produced for under $5,000. In time I hope the contest would become a sought after trophy which actually meant something more than ISO 9000!

Naturally, the award would be presented at the ACCU conference thus bringing publicity for the ACCU, the winners, contestants and sponsors. I can hear it now... "The 2002 Logica-ACCU award for quality software engineering goes to..."