OK I have not had much time to look at this recently but here is what I have read up so far
Setting the servos up as an array is simpler and allows you to group the servos.
Define the min and max rotations on your servos will stop you burning them out. for example, rotate to a maximum of 2000 if 2000 is reached return to center.
Getting to load the servos in groups at a time prevents a power spike when the robot powers on. so for example
start thighs then start hips then start knees
Defining postures for the robot means you can make it walk in posture without coding each walking gate. For example. stand normal and walk, stand tall and walk, crouch and walk.
Inverse kinematics allows you to plot where you want the feet to go and the servos do the rest (as I understand this so far.
Its important to use the servo center as the point of reference because the robot body position varies relation to the legs depending on what posture it is in.
Based on what I have read I decided that I want to build my robot so that
it uses an IR compound eye as the main sensor,
I want to put foot switches in the feet so it wont start the next gait movement until the switch is pressed. so it wont walk off of tables and things.
I found the chopsticks robot and code and decided that this was a good foundation for me. so looking at it a lot of the code is fairly simple to read even if im not at a level to write the code at the moment.