= RTEMS Code Swarm =



[[TOC(Developer/CodeSwarm, depth=2)]]


Code Swarm is a tool which visualizes project commit history.  It can produce some very cool and interesting output.

 *  http://vis.cs.ucdavis.edu/~ogawa/codeswarm/ has videos from Eclipse, Python, Apache and PostgreSQL
 *  http://code.google.com/p/codeswarm is the project's current home
= Running Code Swarm =


{{{
#download and extract
curl -L http://github.com/rictic/code_swarm/tarball/master > code_swarm.tar.gz
tar -xzf code_swarm.tar.gz
mv rictic-code_swarm-*/ code_swarm

#add the bin directory to your path
PATH=$PATH:code_swarm/bin

#convert the CVS log
convert_logs.py -c activity.log > log.xml
#get a code_swarm config file (I'd eventually like to remove this step)
cp code_swarm/bin/config.template rtems.config
#run code_swarm on the config file
code_swarm -c rtems.config
}}}

At this point, you will have a subdirectory in the code_swarm tree named code_swarm_frame with a lot of .png files.
Following the instructions for [http://code.google.com/p/codeswarm/wiki/GeneratingAVideo Generating a Video], I used this command to generate the movie.avi file.

{{{
mencoder mf://*.png -mf fps=24:type=png -ovc lavc -oac copy -o movie.avi
}}}

If that results in a file that is too large, read the Wiki and try the other alternatives.

This takes between 90 minutes and 6 hours on the RTEMS activity log depending on how many frames per day you generate.  At 2 frames per day, it is about 90 minutes and at 6, it is ~6 hours.
= Text for Video =


The video shows periods of high and low activity.  On the videos from other projects, there are subtitles indicating what actually happened in the project at that time.  

 *  beginning until non-OAR committers added - need date
 *  early 1997
 *  June 1997
 *  October 1997
= RTEMS Configuration File =


You can try enabling OpenGL support and see if it works for you. 

{{{
# This is a sample configuration file for code_swarm

# Input file
InputFile=log.xml

# Optional Method instead of MillisecondsPerFrame
FramesPerDay=2

# OpenGL has problems on some platforms.  We're closer to having it work 
# automatically out of the box, but not close enough to enable it by default
UseOpenGL=false

# Color assignment rules
# Keep in order, do not skip numbers. Numbers start
# at 1.
# 
# Pattern:  "Label", "regex", R,G,B, R,G,B
# Label is optional.  If it is omitted, the regex
# will be used.
#
ColorAssign1="Tests",".*test.*", 90,225,90, 110,200,90
ColorAssign2="Build Infrastructure",".*config.*|.*Make.*|.*cfg", 198,229,255, 178,209,235
ColorAssign3="Ports",".*score/cpu.*", 90,225,225, 110,200,200
ColorAssign4="BSPs",".*libbsp.*|.*libcpu.*|.*libchip.*", 30,144,255, 10,124,235
ColorAssign5="Source Code",".*(\.py|\.rb|\.erb|\.hs|\.sql|\.[sSchd]|\.cpp|\.cc|\.m|\.js|\.pl|\.sh|\.java|\.lhs|\.hi|\.ad[abs])", 225,90,90, 200,90,110
ColorAssign6="Documents/Images",".*(\.txt|\.html|\.tex|\.t|\.texinfo|\.texi|\.tmpl|\.css|\.xml|\.yml|\.json|\.png|\.jpg|\.gif|\.jpeg|README|COPYING|LICENSE|AUTHORS)", 90,90,225, 90,110,200

# Save each frame to an image?
TakeSnapshots=true

# Where to save each frame
SnapshotLocation=code_swarm_frames/#####.png

#Is the input xml sorted by date?  It's faster and uses much less memory if it is
IsInputSorted=true
#  - All of the given scripts for producing repository xml files produce sorted data
#  - Please file a bug if one doesn't

# Uncomment to not use avatars
#AvatarFetcher=NoAvatar
}}}