"... In PAGE 12: ... One important criterion to help selecting an appropriate technique to mine a given system for aspects is what implicit or explicit assumptions that technique makes about how the crosscutting concerns are imple- mented. Table5 summarises these assumptions in terms of preconditions that a system has to satisfy in order to find suitable aspect candidates with a given technique. Technique Preconditions on crosscutting concerns in the analysed program Execution patterns Order of calls in context of crosscutting concern is always the same.... In PAGE 15: ... Preconditions. All of the techniques make different assumptions about how cross-cutting concerns are implemented in the system, in order for the technique to find viable aspect candidates (see Table5 ). Since the as- sumptions of the techniques we studied seem quite complementary, and... ..."

"... In PAGE 22: ... As an example, consider the message list given in Table 13. The HYbrid STAtic Dynamic scheduling algorithm, or HYSTAD, determines the schedule (1, 2, 5, 7, 8, 3, 4, 6) shown in Table14 . Also shown are the results from the dynamic FCFS technique as well as the SAG static scheduling technique from the previous section.... ..."

"... In PAGE 16: ...1 Performance Fluctuations Where the real advantage of the hybrid approach occurs is when the actual information about the network deviates from the expected. Consider the message list of Table5 but with some slight deviations in the arrival times. Table 6 shows this list and the results for the three scheduling techniques being considered.... ..."

"... In PAGE 3: ... First, LU factorization, matrix multiplication, and bitonic sorting were analyzed to determine the message passing that oc- curs when they are mapped to a two-dimensional mesh architec- ture. ACT values are given in Table3 for the results of the SCORE static scheduling algorithm utilizing re-routing [13], a FCFS ap- proach both with and without re-routing, and the HYCORE tech- niques. In this table the variance is 0 so the static approach again performed the best.... In PAGE 3: ... Further note that the HYCORE technique out- performs the FCFS approach by approximately 20%. Table3 also shows the results when the variance is 4. Static scheduling no longer works best as it must compensate for worst case times, and HYCORE still works better than FCFS although the percentage is not as great.... ..."

"... In PAGE 20: ....3.1 Application-based message tra c Three common numeric operations, LU factorization, matrix multiplication, and bitonic sorting, were analyzed to determine the message passing that occurs when they are mapped to a two-dimensional mesh architecture. Three scheduling algorithms were used on these tra c patterns and the ACT values are given in Table8 . Note that when the variance is 0, a static scheduling approach (the SCORE algorithm) performed the best.... ..."

"... In PAGE 13: ...1 Application-based message tra c Three common numeric operations, LU factorization, matrix multiplication, and bitonic sorting, were analyzed to determine the message passing that occurs when they are mapped to a two- dimensional mesh architecture. Four scheduling algorithms were used on these tra c patterns and the ACT values are given in Table5 . Note that for each application the SCORE algorithm provides the best performance.... ..."

"... In PAGE 16: ... Consider the message list of Table 5 but with some slight deviations in the arrival times. Table6 shows this list and the results for the three scheduling techniques being considered.... In PAGE 16: ...38. The statically determined schedule and its results are also shown in Table6 . Here the ACT is 42.... In PAGE 16: ... This makes sense because the statically determined schedule must allot for worst case times and therefore cannot incorporate arrival times which may be sooner than the estimates. Table6 also shows the results of applying the hybrid scheduling technique. The ACT for this schedule is 34.... ..."

"... In PAGE 23: ... Consider the message list of Table 13 but with some slight deviations in the arrival times. Table15 shows this list and the results for the three scheduling techniques being considered.... In PAGE 24: ...38. The statically determined schedule and its results are also shown in Table15 . Here the ACT is 42.... In PAGE 24: ... Again, this makes sense because the statically determined schedule must allot for worst case times and therefore cannot adapt to actual arrival times which may be sooner than the estimate. Table15 also shows the results of applying the hybrid scheduling technique. The ACT for this schedule is 34.... ..."

"... In PAGE 17: ... The largest deviation, or variance, is 3 time units. Table7 shows this list and the results for the three scheduling techniques being considered. The abbrevi- ations `Act.... In PAGE 17: ... The e ects of larger variances will be studied in the next section. Applying the HYCORE technique yields the best schedule with the results shown in Table7 and an ACT 28.63.... ..."