V-Base Development Blog

As a response on the (excellent) post of James Shore on the decline and fall of agile, where he discusses why agile often fails when implemented, I responded with what we did before we really started implementing agile; here is my response on his blog:

Very good post, although I would suggest also to focus on how to measure good engineering practices. In the project I currently work on (my job is to implement/streamline development processes) I’ve installed several tools which auto generates lots of reports for our scrum master so she has a lot of tools the measure to the quality of the development process. 
 
And before I even try to implement ‘Agile’ in a software project a want these questions asked and answered: 
 
* Does it compile? 
Does it compile on a non developer machine? Do the latest changes compile?  
A: Continuous Integration 
 
* Does it run? 
Can we execute a recent version of the application without having to wait for a developer to build it, also valid for non gui back end apps! 
A: Daily builds  
A: Release candidates 
 
* Is it tested? 
A: Unit tests  
A: Integration tests  
A: QA team  
A: Hallway usability testing 
A: Code coverage reports 
 
* Who changed what and when? 
A: Version control  
A: Issue tracking 
 
* Did we produce source code daily? 
My main goal when I have to start a project: produce code on a daily basis, get the momentum/routine right 
(instead of having meetings on a daily basis ) 
A: Version control reports 
 
* Second opinion? 
A: Code reviews 
Not of everything, just the areas where unit tests fail regulary 
 
* Can we track bugs? 
 
* Is it documented? 
Not in a word processor document, but one liners in the code do wonders 
 
I call this my ‘extended sprint zero’, first good and disciplined engineering before we start implementing fancy development process theories  
 
Olivier