Sure I can deploy seeds randomly for each and every question...
But then either one of these 2 options will happen:
- Different questions will not have the same values for variables and the exams are built on assumptions that one will have multiple questions using same values for variables.
- The system will pick one pre-randomized seed number for the first question in the exam and will calculate each other question with the picked pre-randomized seed number. Thus eliminating all benefits of caching.
The main goal here is that I can have multiple questions using same values for variables and to cache the pre-calculated solutions.