Our last poll examined how software configuration management (change management / version control) is performed by organisations. Is it an informal activity or performed with formal procedures that are validated by an automated tool?
|No formal procedure and no tool||13%||15%|
|A formal procedure but no tool.||11%||10%|
|A tool but no formal procedure||17%||26%|
|A formal procedure and a tool||36%||49%|
Participants: 271 (346 in 2006)
Ending date: November 2009 (April 2006)
If we compare the current results with the 2006 numbers, we can see two major points. The first is that the number of participants that don’t use tools remains at the same level, around 25%. It is true that there has not been any strong innovation in the area of software configuration management tools in the recent years. The major open source products were also available three years ago. Things have change however for the participants that were using tools. The number participants using a formal procedure is now a clear majority, counting the people that do continuous integration in this category. The increased demand for regulation, for instance the Sarbanes-Oxley law, has surely influenced corporations to adopt formal policies in their software development practices.
We introduced in this poll a new answer to take into account the “continuous integration” way of software configuration management. This agile-linked practice is adopted by 23% of the participants to the survey. This result is similar to the survey conducted by Jurgen Apello on noop.nl. In his case, 29% of the participants had adopted continuous integration. As a funny conclusion on this topic, when I researched on this topic, I noticed that in the 2007 Agile Survey by Version One 50% of the participants declared that they were practicing continuous integration and 57% were currently using a continuous integration tool. This makes me wonder what 7% of the participants were using the continuous integration tool for. ;o) Anyhow, this always remembers us that we should be very cautious with survey number.