Shape performance question - one larger file vs. several smaller

[Gambin Dejan]

Hi,

I have several line shape files showing the borders of different types.
I am wondering if it is better to merge them using ogr2ogr and tehn
classifying them in map file, or use all the files as layers? 

Does anybody know the difference in speed/performance? 

By the way, I have one more question related to performance. Is it
better to place all symbol definitions in symbol file or define them
directly in map file?

thanks very much

regards, dejan