>Solid state media has effectively zero seek time no matter how badly
>fragmented the data might be. There are potential issues with block reads,
>but the delays are microseconds or even nanoseconds. This is negligible
>compared to the millisecond seek times of rotating media for all but real
>time systems (which Newton is not).
Exactly, except that...
Fragmentation slows down the process because the system has to wonder
where to look next and increase the number of blocks to read.
Linear cards are divided into 64 KB blocks. If an object is in a
single block, there is no problem. If it's not, the system has to
find out in which next block it is and read/write an additionnal 64
KB block even if the object was less than 64 KB.
So in a nutshell, fragmentation slows things down a lot for linear stores.
Paul
-- Home page: http://www.kallisys.com/ Newton-powered WebServer: http://newt.dyndns.org:8080/-- This is the Newtontalk mailinglist - http://www.newtontalk.net To unsubscribe or manage: visit the above link or mailto:newtontalk-request_at_newtontalk.net?Subject=unsubscribe
This archive was generated by hypermail 2.1.2 : Sun Sep 09 2001 - 19:47:47 EDT