Software Evolution Visualization

1
Software Evolution Visualization
Lucian Voinea
2
Why Software Evolution Visualization ?

• Software understanding
• what is the structure of the project ?
• what is important ?
• what is difficult ?
• who work on it ?
• what parts are connected to a given part ?
• what parts refer to a given aspect ?
• Product and process assessment
• what is the process ?
• what is the core team ?
• where is the activity concentrated ?
• what is stable ?

3
Challenges

• Data acquisition
• data source ? SW Configuration Management
• long acquisition times
• large storage space, large bandwidth
• data parsing maintenance
• Data analysis
• - Enrich recorded evolution data E.g. evolution
  similarity metrics
• Data visualization
• multivariate 10s attributes (Demo 6
  attributes)
• large data amount 100s files,10s versions,
  1000s lines
• VTK 2700 files x 100 versions ArgoUML
  4452 files x 80 versions

SW Configuration Management?SW evolution mining
4
SW Evolution Visualization Toolset
data extraction

version control
management
system



CVSscan
CVSgrab
CVS

Project evolutionvisualization
File evolutionvisualization
5
CVSgrab
Color - encodes a file metric computed on a given
version E.g. File type, Version author, Size
increase, Release ID,Specific text, Activity
6
CVSgrab Example
Color encodes version author
Time
Files
7
CVSscan
if self.bShowClusterColor glBlendFunc(GL_ONE_MINU
S_... glEnable(GL_BLEND) yPos 0
bSwitch True for i in lCushionMap if
bSwitch cushions.drawColorCl.. else
cushions.drawColorClus bSwitch not bSwitch
yPos yPos i glDisable(GL_BLEND)
cushion clusters glBlendFunc(GL_ZERO,GL_...
glEnable(GL_BLEND) yPos 0 bSwitch True

Color - encodes a line metric
8
CVSscan
Line layout
9
CVSscan Example
Color encodes content
10
Conclusions

• Toolset for visualization of software evolution
  CVSgrab CVSscan
• Exploit source code history from CVS as main
  source for getting insight in the structure and
  development process
• Complements existing reverse engineering and
  code navigation tools
• Outlook Improve the data analysis step, find
  mechanisms for annotation management, improve
  selection mechanisms, enhance the visualization
  with new features, extend the data acquisition
  sources and types

www.win.tue.nl/lvoinea/VCN.html