Point/Counterpoint

POINT: AOP Will See Widespread Adoption Adrian Colyer, Rob Harrop, Rod Johnson, and Alexandre Vasseur AOP offers both the short--and long-term return on investment needed to ensure its widespread adoption. COUNTERPOINT: AOP Has Yet to Prove Its Value Danio Beuche and Cédric Beust Where is the proof that nonhierarchical modularization into aspects makes systems development cheaper or better?     

Spray cooling effects on milk production, milk, and rectal temperatures of cows during a moderate temperate summer season

During summer 1982, responses of lactating Holstein and Guernsey cows were measured by milk temperature recorded by a Digital Dataloger with thermocouples attached to Boumatic flow meters. Maximum air temperature and temperature-humidity index averaged 30.8 degrees C and 75.6 for July. Breed did not affect milk temperature, but within-breed milk temperature increased with production. In a second study, benefits of spray cooling were evaluated with 24 Holsteins in midlactation assigned randomly to two groups of 12 and maintained under loose-housing conditions. Spray nozzles were installed in the walkways and under the manger shade for the spray treatment group. Maximum temperature and temperature-humidity index during the spray study were 27 degrees C and 73.9. Rectal temperature taken following milking averaged less for treatment than control (38.8 versus 39.1 degrees C). Milk temperature was similar (37.8 versus 38.1 degrees C). Daily milk yield was .70 kg higher than controls. Milk temperature may provide reliable indication of climate stress similar to rectal temperature, and spray cooling improves cow comfort and lessens summer decline of milk production.     

Constructions of particular random processes

This paper reviews how to construct sets of random numbers with particular amplitude distributions and correlations among values. These constructions support both high-fidelity Monte Carlo simulation and analytic design studies. A variety of constructions is presented to free engineering models from “white or normal” limitations embodied in many current simulations. The methods support constructions of conventional stationary and normally distributed processes, nonstationary, nonnormal signal and interference waveforms, nonhomogeneous random scenes, nonhomogeneous volumetric clutter realizations, and snapshots of randomly evolving, nonhomogeneous scenes. Each case will have specified amplitude statistics, e.g., normal, log-normal, uniform, Weibull, or discrete amplitude statistics; and selected correlation, e.g., white, pink, or patchy statistics, clouds. or speckles. Sets of random numbers with correlation, nonstationarities, various tails for the amplitude distributions, and multimodal distributions can be constructed. The paper emphasizes aspects of probability theory necessary to engineering modeling