ZeroTurnaround Release Results from Survey of Indian Developers at JavaOne
ZeroTurnaround have compared and contrasted the state of the technologies used for software development, and analyzed the tool-usage aspect of Indian productivity. The survey covers application servers (containers), IDEs, frameworks and build tools used by Java teams across India and globally.
The report includes several comparisons between the data received from the India survey and the Java EE Productivity Report, which the firm published earlier in 2011. Combined, the two surveys contain results from more than 2,400 developer-respondents. The firm made use of a rented email list to contact developers in India, which the report acknowledges introduces some forms of bias. The sample size is also very small, and there may be some cultural differences and problems of interpretation in play, so any conclusions need to be treated with some scepticism. The data is however freely available for review.
In terms of demographic, 31% of the Indian developers in the sample work in the "services" sector, which encompasses system integration plus managed and professional services. Independent Software Vendors (ISV) took the number two position at 16%, with consulting firms, outsourcing firms, telecommunications providers and academic institutions each making up roughly a further 10% of the market.
Amongst the notable points:
For application servers, whilst the major players are all represented in India, the market shares are significantly different when compared with the RoW. The greatest difference can be seen with IBM WebSphere, which is the number 2 player in India with 26% of respondents using it, as against just 7% in the RoW sample. WebLogic is the number three application server for both India and globally, but has much higher usage in India than in the rest of the world sample.
IntelliJ IDEA is popular outside of India but not within it: 22% of global respondents use IntelliJ vs. just 2% in India. IDEs like MyEclipse, NetBeans, JDeveloper and RAD showed the opposite findings, although the differences are less dramatic.
For frameworks, the survey found significant differences in the use of JSP, Struts (1 and 2), Hibernate, JSF and JPA. Spring has a significantly smaller representation in India. There is some suggestion of a more conservative approach to newer Java EE technologies, with a greater use of EJB 2.0 in Indian versus the RoW sample, and a lower use of EJB 3.0.
Maven also seems to have lower use in India than in the rest of the world. Globally, developers are using Apache Ant and Maven almost equally, but in India only about 17 percent of developers reported using Maven (compared to 53 percent globally).
As you might expect, given that ZeroTurnaround builds and sells JRebel, the survey goes on to look at turnaround time as a way of measuring productivity. The survey implies average redeploy times between India and RoW are quite different. In the last report, the average redeploy time for RoW came out at 3.1 minutes with a standard deviation of 2.8. For India, the average is 4.8 minutes with a standard deviation of 4.2. Nearly a quarter (23%) of the Indian survey population require at least 10 minutes for each redeploy, as opposed to 7% for the rest of the world.
In related news, ZeroTurnaround's JRebel product was awarded a Duke's Choice award at JavaOne.