Abstract
We show that Naming- the existence of distinct IDs known to all- is a necessary assumption of Herlihy’s universality result for Consensus. We then show in a very precise sense that Naming is harder than Consensus and bring to the surface some important differences existing between popular shared memory models which usually remain unnoticed.
Cite
CITATION STYLE
Bhurman, H., Panconesi, A., Silvestri, R., & Vitanyi, P. (2000). On the importance of having an identity or, is consensus really universal? In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1914, pp. 134–148). Springer Verlag. https://doi.org/10.1007/3-540-40026-5_9
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
