Consumers of video and other content in today's networks have very diverse 
display and computing equipment ranging from mobile phones and handheld devices 
to desktops, and HDTVs. Downloaded items range from simple stock quotes to news 
stories to full movies. Current schemes for content download such as standardized 
rateless codes, are shown to be inefficient at addressing such high heterogeneity. 
Because of that, we in particular address the problem of optimizing rateless codes 
for single-server streaming to diverse users, assuming that diversity in users is 
present not only because they have different channel conditions, but also because 
they demand different amounts of information and have different decoding capabilities 
(e.g, some even cannot code). We propose several performance measures for such 
heterogeneous networks, and optimize the performance of the rateless code accordingly. 
Optimized performance shows great improvement in total bandwidth consumption over using 
the conventional approach of sending separately-encoded streams to different types of 
user nodes over independent channels.