Tim Van Wassenhove <eu**@pi.be> wrote in message news:<bv************@ID-188825.news.uni-berlin.de>...
On 2004-01-31, CountScubula <me@scantek.hotmail.com> wrote: a read from a db is also a read from a file.
file = file read
db = db + cool stuff + file read
if the db is on another server, then add + networking to other server
And why couldn't a file be considered as a database? :)
A *file* can not be considered as a *database*, because *files* are
implemented at OS level. Whereas databases are always implemented a
layer above OS.
True database adds overhead over file read operations when we read
some data. But it adds many features like synchronization,cache,search
etc over simple file read operation.
So, it all depends upon your requirement. If you need to read static
text every time it will be faster to save your data in a file. But if
you want to update some value every time you read the data, database
will be a better choice.
For faster performance it is always a better idea to cache your static
contents from database to files, and read from them.
And the basic thing is: databases always store data in file system
only.
--
Cheers,
Rahul Anand