Living in a world of errors
Living in a world of errors
Posted Apr 14, 2005 21:48 UTC (Thu) by swiftone (guest, #17420)In reply to: Living in a world of errors by shane
Parent article: Call to atention about using hash functions as content indexers (SCM saga)
Hard disks (well IDE) uses CRC to check for bad data.
As the article mentions, _checking_ identity (e.g. checking for bad data) is different than _assigning_ identity. All of th examples you cite are checking validity. None are assigning identity. If you look at identity algorithms in the industy (UUID, etc) you'll find they are a different breed.
Posted Aug 29, 2005 15:40 UTC (Mon)
by lunakid (guest, #32146)
[Link]
About checking and assigning (creating) identity being different:
Living in a world of errors
Quite correct. The difference between the two is context. Identity checking always assumes/implies the more specific context of e.g. change characteristics. Identity assignment, OTOH, can basically only assume some upper bound of numbers in the supported pool of identities, and little more (some life-cycle information for aging may also be available etc.).