Help and info on line endings

[Les Spam]

We have been having trouble with line endings because we check out files 
on a Linux server and zip them up for download by users who do not have 
CVS clients. Win9x users end up with line oriented text files that do 
not parse because the scanner sees one long line.

However, we discovered that if we run unix2dos on these files then they 
have DOS line endings when they are checked out and unzipped. Looking 
inside the CVS files, we see they have *DOS* line endings for the 
relevant lines and Unix line endings for the rest. There is a reference 
to "canonical form" in the CVS docs for the -kb flag, from which we 
infer that the CVS files are supposed to have only Unix line endings. 
Can anyone explain how/why it is that the DOS line endings are being 
preserved in these CVS text files?

We are not that happy with this unix2dos solution because: 1) it can 
screw up parsing the files on Unix or Macs, 2) people forget to convert 
the files, and 3) it doesn't work for everyone, every time (and we 
haven't figured out why - any ideas?).  We could get a consistent result 
by converting to DOS and checking  these files in as binary, but we 
don't like that approach for reasons 1) and 2). Does anyone have other 
suggestions?

Thanks,
 Scott