Re: [ECOS] Cygwin & i386-target

[Bart Veer <bartv at redhat dot com>]

>>>>> "Rosimildo" == Rosimildo daSilva <rosimildo@hotmail.com> writes:

    <snip>

    Rosimildo> These strange errors are most likely caused by the
    Rosimildo> wrong handling of CRLF <--> LF. You most likely needs
    Rosimildo> to cleanup the source files with some utility such as
    Rosimildo> dos2unix.

    Rosimildo> This is *by far*, the most serious problem that CygWin
    Rosimildo> has. Nothing else comes close, and I am amazed that in
    Rosimildo> the last 4 years they have been unable to fix it.

The current situation is by no means perfect, but there is no solution
that is going to satisfy everybody. Sure, many people would have
preferred it if cygwin just worked entirely in binary mode, but that
would have made it impossible to share files between a cygwin
environment and typical Windows applications. That is important for at
least some of the cygwin user base.

If you have suggestions on how to improve cygwin, please use the
appropriate cygwin mailing list. Also, please feel free to send bug
reports to any operating system vendors who believe that two
characters are needed to mark the end of a line in a text file.

For the record, I believe that the current gcc CVS tree works much
better in this regard, i.e. it will just ignore spurious carriage
returns rather than throwing up errors like the above. This is not
quite as simple as you might think, there are subtleties in the ISO C
standard which must be observed by a conformant compiler. In due
course, probably after the gcc 3.0 release, I would expect that to
become available for cygwin users.

Bart