One of the great joys of reading SQL forums is the way that even the simplest or most naive question can jog one’s thinking. Maybe it is the way that they’re approaching the problem or the very nature of the problem they’re tackling that is intriguing. Of course, it is good to see, and learn from, the expert answers but SQL Server tends to throw up questions to which the answer is 'Well, it depends….' Sometimes there are several different right answers.

SSC forums cope very well with the diffuse nature of providing answers to SQL Server problems. There is, perhaps, a slight tension with the traditional serial nature of forum software. This is why I like to use ASK SQL Server Central as well. Here is a medium that allows more freedom to float ideas, to comment on replies, to vote on solutions. Particular pleasures for me are the SQL Speed Phreak competitions. This sort of activity is a great way of exchanging ideas and techniques and one can’t fail to learn something from them.

The different types of forum seem to engender different styles of contribution. It makes me wonder if forum software doesn’t just provide a passive substrate at all but, on the contrary, they mould the way that discussions develop. The 'stackOverflow' style seems great for providing expert solutions to problems, but the SSC forums engender a more discursive and free-ranging discussion. At one point, it seemed to be somewhat quirky for SQL Server Central to have two different forums, but now I like using both, depending. I can see a possible value in extending this idea by having several different ways of supporting discussions, including maybe, a more newsgroup-styled forum, or, alternatively, a wiki. The obvious downsides are, of course, the potential for duplication and the difficulty of searching.

It may be possible someday, to square the circle. A database person can see that a single database model of a group discussion has a 'graph' structure that can be represented by a GUI in a number of different ways. If one could crack the database model that would support all these different forms of discussion, or cooperative work, then could it be that we can simply fit the 'visualization' to the nature of the particular discussion, or the preferences of the viewer.

Dogfood?

Phil Factor