This post originated from an RSS feed registered with Java Buzz
by Fred Grott.
Original Post: Synching with every IDE classpath
Feed Title: ShareMe Technologies LLC-The Mobile Future
Feed URL: http://www.jroller.com/shareme/feed/entries/rss
Feed Description: A Weblog about Java programming and digital convergence on mobile devices in such areas as P2P and collaborative technology.
In making the decision of synching with the IDE classpath of those IDEs that write their own classpath through an ant target and those IDEs not writing .classpath files by making the ant build scripts the incremental compiler/build process we get every one on the same page of using incremental compiliing/builidng through use of the Orca TDD Build Framework. The benefits of incremental builds are of course the QA and analysis reports you can get back on your incremental builds.
Why QA, unit tests, and code analysis in the incremental builds as opposed to a weekly build? Well if the errors is not caught in wrong java implementatation or wrong architecture until the end of the week how much time have you wasted?
A. 1 day, the 18 hour day after the weekend build finds the errors.
B. The several days you spent on the wrong implementation.
C. All of the Above
The obvious answer is C based on such manuals on software engineering as Frederick Brooks Jr's own tomes. It would seem that having the QA upfront within the incremental build allows you to catch wrong paths and refactor and thus even rescue that wrong implementation or wrong path into something still very usefull. But maybe I am wrong, maybe doign it the mobile game industry way of at the week end build really works, NOT!
Now, for our JAMDAT, Tirawireless, macrospace, mForma, and other mobile game publishers. Yes your build process appears to work. But once, you factor in the overtime errors after a week end build fails that illusion seems to dissolve in a rapid fashion.
Previous Topic
