Search

The games industry is growing worldwide, eclipsing the global film industry as a premier entertainment solution. The development of a commercial game is a complex, lengthy, and costly process, and increasing amounts of resources are required to build a successful game as technology improves. To this end, quality assurance (QA) is critical for producing high-quality games that are fun and reasonably defect-free. Existing work in QA for games describes game development methodologies and testing approaches, goals, and automation, but do not address the disparate contexts of independent (indie) and non-indie game development. Indie developers have limited access to funding and resources compared to their non- indie counterparts, which makes achieving a similar quality of QA difficult. Yet, indie games make up the lion’s share of newly released games each year. Therefore, we want to empower indie developers by maximizing their QA opportunities within their resource constraints. In this study, we survey 19 game developers who have experience with commercially released games to learn about their QA experiences and perspectives based on 22 of their released game projects. We compare responses about indie and non-indie projects regarding test performance, planning, goals, automation, tools, results, and resources to find key differences in practices, goals, and needs between them. We present our findings on the major differences in game testing and automation between indies and non-indies, including indies having less clear goals and plans for testing, performing tests on a conditional basis over a regular testing schedule, and having subjective test results which are open to interpretation. We also discuss issues of culture and management that respondents raise in their open-ended answers. Based on our findings, we present recommendations for each area of testing to improve quality assurance practices for resource-constrained game developers.

Twelve PGV models, MDC-2, and HIROSE, which are blockcipher-based hash functions, have been proven to be secure as hash functions when they are instantiated with ideal blockciphers. However, their security cannot be guaranteed when the base blockciphers use weak key-schedules. In this paper, we propose various related-key or chosen-key differential paths of Fantomas, Midori-128, GOST, and 12-round reduced AES-256 using key-schedules with weak diffusion effects. We then describe how these differential paths undermine the security of PGV models, MDC-2, or HIROSE. In addition, we show that the invariant subspace attacks on PRINT and Midori-64 can be transferred to collision attacks on their some hash modes.