What is the difference between the absolute thermodynamic temperature scale and the practical temperature scale: ITS-90?
The absolute thermodynamic temperature scale is a scale fixed by absolute zero and (at least for the kelvin, I'm not sure about the Rankine scale) the triple point of water, where absolute zero is defined as 0 K and the triple point of water is defined as 273.16 K. The rest of the scale is just a straight line from zero through 273.16 K off to infinity. However, this scale relies on only two defined points (only one of which can actually be measured, absolute zero being established from thermodynamic relations).
In practical use, for the purposes of calibrating thermometres and the like, this is not a great scale as the two calibration points may be far away from the temperature region one is interested in actually measuring. The usual way to deal with this is to use other reference points outside the actual definition. So that different people do this as consistently as possible, ITS-90 was created as a practical implementation of the absolute scale, setting additional reference points (all phase transitions) to defined values, interpolating between them to get other values. It's technically not a scale in itself, just a standard for calibrating temperatures that are as close to the absolute scale as possible. By making sure everyone uses the same reference points, measurements can be more consistent than if reference points are chosen arbitrarily and possibly assigned different values.
For example, if one wanted to make a thermometre for general lab use, one might pick the triple point of mercury, the triple point of water, and the melting point of gallium from the ITS-90 standard for reference points, knowing that other thermometres in this range conforming to the standard would also be calibrated this way, rather than choosing say the melting and boiling points of water, increasing the likelihood of consistent measurements with other thermometres.