Re: random i/o and relative efficiency

[david.kahn@xxxxxxxxxxxxxxxxxx]

Pat Barber <MBOCEANSIDE@postoffice.worldnet.att.net>

>Large "random" files can really blaze with a binary search,BUT,
>the file has to be in some order...which sort of defeats the random
>concept... stick with keyed index files....

Pat,

Are you perhaps confusing "random" with "relative"? AIUI Joel's example
_is_ with an indexed file. The choice is between:

a)   Chain to the 1st record in a group
     Chain to the 1st active record in a group
     Reade through the active records

or b)     Chain to the 1st record in a group
     Reade through the group

when each group is known to be a sequence consisting of an iteration of
inactive records followed by an iteration of inactive records. It was a
good question, and I think the only universally correct answer, as in so
many cases, is "it depends".

Dave Kahn, ABB Steward Ltd.


+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---