Text tools, line endings, and O_TEXT

[John Cowan]

Eric Blake scripsit:

> If you disagree with the cygwin philosophy, or with the particular
> implementation of that philosophy as embodied by ptx, you are better
> off raising the issue on the cygwin mailing lists rather than here.

I don't in fact disagree with it; I was simply clarifying the OP's
situation (which appeared only in the Subject: line).

> This argument, on the other hand, is appropriate to bring up here, as you
> are claiming that ptx's behavior should be able to (optionally) ignore \r
> on all platforms.

Indeed, I'd like to see all text tools accept all six known line ending
conventions:  CR+LF, CR alone, LF alone, NEL alone, CR+NEL, LS.
(NEL is U+0085, LS is U+2028.)

> In which case, Jim's proposed patch of adding O_TEXT is
> inappropriate on two counts - it will not affect Linux (while you argued
> that Linux should also have a way to ignore \r), and it will make the
> default behavior of a cygwin binary mount diverge from behavior of Linux
> (which goes counter to cygwin's goal of behaving like Linux when used on
> binary mounts).

Are you saying that an O_TEXT open under Cygwin, given a binary mount,
will cause newline translation?  I had assumed that O_TEXT would be
ignored under Cygwin just as it is under Linux.

That said, I believe it is good for portability to use O_TEXT
and O_BINARY (and their stdio equivalents) in all software intended
to be portable.  Unix-tradition systems won't benefit, but it doesn't
cost anything either; other systems will be enabled to DTRT as a result.

-- 
Even the best of friends cannot                 John Cowan
attend each others' funeral.                    address@hidden
        --Kehlog Albran, The Profit             http://www.ccil.org/~cowan