There was an interesting article published in 2015 about the retirement aged 80 of Larry Zottarelli who was the last surviving original engineer on the Voyager project at JPL/ NASA.
http://money.cnn.com/2015/10/27/technology/voyager-nasa/His imminent retirement led to something of a panic within NASA at the time, and a more general call-out for younger programmers who could understand ancient computer programming languages like Fortran, Algol and assembler that were needed to keep an almost 40 year old space probe functional.
The thrusters recently re-activated on Voyager 1 for the first time in 37 years are called 'Trajectory Correction Maneuver' (TCM) devices.
http://edition.cnn.com/2017/12/01/us/voyager-1-thrusters-fired-first-time-since-1980/index.htmlThe control signals took 19 hours and 35 minutes to reach Voyager 1. But before they could get as far as transmitting the control commands, the team at JPL had to audit reams of archaic computer code to make sure they understood what they were attempting to do.
There is a certain lack of public information about exactly how the Voyager probes were originally coded. FOIA requests on the subject are still being blocked by NASA, apparently because the answers could compromise IT and comms security.
The Voyager computer hardware is said to be controlled by custom-built GEC interrupt driven processors, with a system memory capacity of just 64 Kilobytes. The main programming was written in an early version of Fortran, but some system sub-routines such as thruster control are hard-coded in assembler to achieve optimum use of memory space.