[fpc-pascal]using big files with high performance

[Christian Müllerke]

Hi!

Sorry for bad english, and, maybe this is a beginner question:

I'am trying to write my own small NNTP(Usenet)-daemon (experimental - just 
want to increase my knowledge).

What whould be the best method for storing articles in files divided by group 
(one group, one file)? I want a good performance, ram usage should be small. 

Is BerkeleyDB a solution? Or writing file-access with streams and a 
index-file? Next problem would be the locking of files (can BDB handle this 
for me?).

I'am using Linux and FreeBSD. Never programmed software for big files with 
locking before.

thanks in advance, Christian
-- 
Christian Muellerke, cm at cmuellerke.de, http://www.cmuellerke.de
Seite zum Webauthoring: http://www.mywebresource.de