Investigating technical and non-technical factors influencing modern code review

When submitting patches for code review, individual developers are primarily interested in maximizing the chances of their patch being accepted in the least time possible. In principle, code review is a transparent process in which reviewers aim to assess the qualities of the patch on its technical merits in a timely manner; however, in practice the execution of this process can be affected by a variety of factors, some of which are external to the technical content of the patch itself. In this paper, we describe empirical studies of the code review processes for large, open source projects such as WebKit and Google Blink. We first consider factors that have been examined in previous studies — patch size, priority, and component — and then extend our enquiries to explore the effects of organization (which company is involved) and developer profile (review load and activity, patch writer experience) on code review response time and eventual outcome. Our approach uses a reverse engineered model of the patch submission process, and extracts key information from the issue-tracking and code review systems. Our findings suggest that these non-technical factors can significantly impact code review outcomes.     

Some observations on "paralytic myoglobinuria" of cattle in Britain

Characteristics of 44 outbreaks of "paralytic myoglobinuria" in yearling cattle reported to veterinary investigation centres during 1974 and 1975 are described. The clinical syndrome and associated nutrition and management of cattle are discussed in the light of present knowledge of this disease with particular reference to the diagnostic and epidemiological value of plasma creatine phosphokinase and blood glutathione peroxidase activities.     

Influence of Stress Ratio on the Elevated-Temperature Fatigue of a Silicon Carbide Fiber-Reinforced Silicon Nitride Composite

The influence of stress ratio on the tensile fatigue behavior of a unidirectional SiC-fiber/Si₃N₄-matrix composite was investigated at 1200°C. Tensile stress ratios of 0.1, 0.3, and 0.5 were examined. Fatigue testing was conducted in air, at a sinusoidal loading frequency of 10 Hz. For peak fatigue stresses below the proportional limit of the composite (approximately 195 MPa at 1200°C) specimens survived 5 × 10⁶ cycles, independent of stress ratio. At peak stresses above the proportional limit, fatigue failures were observed; fatigue life decreased significantly as the stress ratio was lowered from 0.5 to 0.1. Creep appears to be the predominant damage mechanism which occurs during fatigue below the proportional limit. Both mechanical cycle-by-cycle fatigue damage and creep contribute to specimen failure at peak stresses above the proportional limit.     

Plasma-Sprayed Aluminum and Titanium Adherends: III

The durability of aluminum and titanium adherends, plasma-sprayed with polymeric coatings, and bonded with an epoxy and a polyimide adhesive has been investigated. Organic-polymeric coatings were plasma-sprayed using epoxy, polyester, polyimide, and cyanate ester components. Durability was investigated using a wedge-type specimen by exposing the specimens to an environmental cycle that included low temperature, high relative humidity at elevated temperature, high temperature at atmospheric pressure in air, high temperature in a vacuum, and room temperature. The systems exhibiting durability comparable with that for adherends treated using standard solution methods, included aluminum or titanium coated with a bis-maleimide/cyanate ester (B-CE) or a bis-maleimide-LaRC TPI-1500^® (B-TPI) mixture and bonded with an epoxy or a polyimide adhesive. For these B-CE- and B-TPI-coated specimens, failure during exposure to the environmental cycle occurred in the adhesive, indicating a favorable adherend/plasma-sprayed coating interaction.     

Diamond vacuum field emission devices

This article reports the development of (a) vertical and (b) lateral diamond vacuum field emission devices with excellent field emission characteristics. These diamond field emission devices, diode and triode, were fabricated using a self-aligning gate formation technique from silicon-on-insulator wafers using conventional silicon micropatterning and etching techniques. High emission current >0.1 A was achieved from the vertical diamond field emission diode with an indented anode design. The gated diamond triode in vertical configuration displayed excellent transistor characteristics with high DC gain of 800 and large AC output voltage of 100 V p–p. Lateral diamond field emission diodes with cathode–anode spacing less than 2 μm were fabricated. The lateral diamond emitter exhibited a low turn-on voltage of 5 V and a high emission current of 6 μA. The low turn-on voltage (field 3 V/μm) and high emission characteristics are the best of reported lateral field emitter structures.     

Supercritical water oxidation of quinoline in a continuous plug flow reactor—part 2: kinetics

The results of a detailed investigation into the kinetics of quinoline oxidation in supercritical water are presented. The novel kinetic data presented were obtained in a continuously operated, plug flow reactor where parameters such as temperature, pressure, residence time and stoichiometric ratio of oxidant to quinoline were investigated and detailed in the companion paper (Pinto LDS, Freitas dos Santos LMF, Al‐Duri B and Santos RCD, Supercritical water oxidation of quinoline in a continuous plug flow reactor—part 1: effect of key operating parameters. J Chem Technol Biotechnol). An induction time was experimentally observed, ranging from 1.5 to 3.5 s, with longer times observed in experiments carried out at lower temperatures. A pseudo‐first‐order rate expression with respect to quinoline concentration (with oxygen excess) was first adopted and the activation energy of 234 kJ mol^?1 and a pre‐exponential factor of 2.1 × 10¹⁴ s^?1 were estimated. Furthermore, an integral power rate model expression was established, attributing a reaction order for quinoline as 1 and for oxygen as 0.36. An activation energy and pre‐exponential factor for this model were determined as 224 kJ mol^?1 and 3.68 × 10¹⁴ M^?0.36 s^?1, respectively. A global rate expression was then regressed for the quinoline reaction rate from the complete set of data. The resulting activation energy was 226 ± 19 kJ mol^?1 and the pre‐exponential factor was 2.7 × 10^{13 ± 2} M^?0.1 s^?1. The reaction orders for quinoline and oxygen were 0.8 ± 0.1 and 0.3 ± 0.1, respectively. It was shown that the least‐squares regression method provided the best‐fit model for experimental results investigated in this study. Copyright © 2006 Society of Chemical Industry     

Studies on the refractive index and dispersion of American tung oil

1. Tung oil has a refractive index and a dispersion so far above those of any other common oil that both are valuable criteria for identification purposes. With proper equipment the dispersion, in addition to the refractive index, can be determined with little extra effort and would confirm the conclusions drawn from the refractive index.
2. Mixtures of tung oil with another vegetable oil (except oiticica and other rare conjugated oils) can be analyzed to within 0.5% from the refractive index for either the sodium or the mercury line if the refractive indices of the separate oils are known. The mixtures can be analyzed from the dispersion to within about 1% of the correct composition if the dispersions of the separate oils are known. If the adulterating oil is not known the adulteration can be more closely estimated from the depression of the dispersion than from the depression of the refractive index.
3. When tung oil is bodied by heat the refractive indices for the sodium and mercury lines and the dispersion fall rapidly and continuously to the point of gelation, but the changes are so similar that no worth-while additional information is obtained by determining more than one refractive index. The fact that refractive index decreases as viscosity increases suggests the use of the refractive index in controlling the bodying of tung oil.
4. Other things being equal, the refractive index for the mercury line should give more accurate information on tung oil than that for the sodium line because of the greater changes in the refractive index for the mercury line upon adulteration or heating.
5. A correlation coefficient of 0.83 was found for refractive index with the diene number of tung oil. A lower correlation coefficient was found for refractive index with the iodine number, but the latter would probably be higher if a more accurate method for the determination of the iodine number of tung oil were available.

     

Tensile Creep and Creep-Recovery Behavior of a SiC-Fiber─Si3N4-Matrix Composite

The tensile creep and creep-recovery behavior of a unidirectional SiC-fiber/Si₃N₄-matrix composite was investigated at 1200°C in air. A primary objective of the study was to determine how various sustained and cyclic creep loading histories would influence the creep rate, accumulated creep strain, and the amount of strain recovered upon specimen unloading. The key results obtained from the investigation can be summarized as follows: (1) A threshold stress of 60 MPa was identified, below which the creep rate of the composite was exceedingly low (∼10⁻¹² s⁻¹). (2) Periodic fiber fracture was identified as a fundamental damage mode for sustained tensile creep at stresses of 200 and 250 MPa. (3) Because of transient stress redistribution between the fibers and matrix, the creep life and failure mode at 250 MPa. were strongly influenced by the rate at which the initial creep stress was applied. (4) Very dramatic creep-strain recovery occurred during cyclic creep; for cyclic loading between stress limits of 200 and 2 MPa, 80% of the prior creep strain was recovered during 50-h-creep/ 50-h-unloading cycles and over 90% during 300-s-creep/ 300-s-unloading cycles. (5) Cyclic loading significantly lowered the duration of primary creep and overall creep-strain accumulation. The implications of the results for microstructural and component design are discussed.