Talk:Deadlock prevention algorithms
Add topic | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||||||
| ||||||||||||||||||
As of 12 May 2013, this article is largely nonsense, written by persons who have no understanding of the issues. It could be safely deleted without loss to the wikipedia. It is so bad, reading it is currently a waste of time.
The title of the article is meaningful, however, so with some relevant content, it could be useful.
The "Distributed Deadlock" section is a verbatim lift from the wikipedia article entitled "Deadlock". The use of jargon is excessive, but the section is the only one in the article that has any well defined substance.
What the article should describe:
Multiprocessing algorithms for preventing irresolvable requests for shared resources that lead to the failure of a computation to proceed. This really has nothing to do with the halting problem. The banker's algorithm is one such algorithm, although it is neither especially good nor especially practical.
Recursive locks are perfectly valid tools, and so do not need to be "prevented", and in any case have nothing to do with the deadlock issue: like any lock, they merely ensure that competing processes do not corrupt one another's data. — Preceding unsigned comment added by 216.251.132.13 (talk) 22:41, 12 May 2013 (UTC)