This has been requested or discussed previously as well.
This is still an issue in 10 beta 9. Several times I have committed code which fails when deployed to a Linux environment because of this. I have to manually convert all files.
This should really be a simple issue if not an development language limitation.
Marco, can you implement a "no BOM" save file feature for UTF8?
BOM:
http://en.wikipedia.org/wiki/Byte_order_markRegards
Stian