The primary reason for purchasing and using a rip is to allow the absolute control of the colour output of the printer (this can not be done the same by using print drivers through the operating system). A rip will communicate directly (usually by network, USB or Firewire) with the printer in it’s native language, bypassing the operating systems drivers. OS drivers do not allow for the selection of both CMYK and RGB input profiles, this means that results may not be very accurate depending on your requirements. They also don’t have the feature that most rips offer of calibration the printer.
A printer usually has very little knowledge and would best be described inside as “dumb”. By this I mean that it knows only drop of ink, or no drop of ink. The rest is left up to where you process the file – without a rip this must be done by the operating system. As an example, Microsoft Windows makes some assumptions that cannot be controlled. Windows assumes that all digital files are intended only for viewing onscreen or over the internet (all RGB output, unlike print). This means that if you submit a CMYK file through a windows print driver, it will always firstly convert that file to RGB so that it can be processed, and then be converted to CMYK data for the printer to produce. This adds 2 conversion that will change the colour of your file and affect the output results.
Windows also makes the assumption that any files being created are likely to be office presentations or internet images, and uses “Saturation” rendering intent. This will shift almost all colour leaving little chance of achieving accurate spot colour reproduction.