The characters at the beginning of the file are the standard signature
for UTF8 and Unicode files, if you are not using this formats you can
modify the default encoding (in the default document class and in the
rest of document classes you use) to Ansi (this will use as encoding
your default codepage). For C programs I always use Ansi encoding to
avoid problems with the compiler, for HTML and TeX files I use UF8.
I'm not entirely sure if many editors in desktop PCs take notice of
this "signature" characters, mine (TextPad) does it properly, but if
that is the not the case for other editors It could be a good idea to
allow unsigned utf-8 and unicode encodings, but this makes the opening
slower and error prune since it needs to detect the kind file it is
reading.
Regards.
Esteban