Olaf Noehring wrote:
Ähm, du hast schon gemerkt, daß diese Newsgroup englisch ist?
For those that aren't able to read the german text above:
So what about your problem. Why don't you guys test it? I guess the
database will be faster in the average case, but this also depends on
the database structure for example which indices have to be updated
after an update and of course of the structure you use in the files,
because a database also stores its data into a file ;)
So lets think about an avarage case:
10000 of datarows in the most naive file solution a CSV file. Imaging
you want to alter the 5000th row, you need to load the entire file (at
once or in parts), alter the row and store the whole file. A database
doesn't do so, so it would be of course faster. But if you have say a
B*-Tree structure behind your files you may get the same results a
database get. Well just test it.