After porting a satellite guidance and control program to a newly acquired computer with a RISC architecture, one of my colleagues observed that the very large program started to trigger numerous memory unaligned exceptions. It was discovered that the exceptions were due to the fact that the new faster processor enforces memory alignment restrictions but the previous processor did not. The colleague dreaded having to track down and modify the numerous instructions that were causing the memory unaligned exceptions and argued that programmers should be free to write their software without concern for memory alignment restrictions. One proposed solution was to include an exception handler that intercepted each unaligned access and substituted byte accesses in place of the word and double word accesses that were causing the problem. I was assigned the task of writing such an exception handler. What would be your opinion on this issue and what would you propose to resolve it?