Rationale for line-ending recommendation?

Linus Hicks linush100@gmail.com
Mon Dec 22 19:12:00 GMT 2008

On Fri, Dec 19, 2008 at 12:02 PM, Corinna Vinschen wrote:
> Yeah, right, if this were in fact the 21st century, why would we ever
> have to deal with a 30 year old, idiotic-from-the-start text file format
> using two line ending chars?

How quickly we forget...

The first devices that had any understanding of this kind of data
stream generated display characters through mechanical means - much
like an old mechanical typewriter. These devices all had a carriage
that moved left to right or right to left depending on whether you
moved the paper or the print head while it struck an inked ribbon
between a platen and character dies that supplied a limited and fixed
set of characters.

The people who designed these things obviously saw the carriage return
as distinct from the linefeed. In fact, these devices had no capacity
to buffer the input data stream and even at 300 baud, it took longer
to return the carriage than the interval between characters. If you
did not supply some non-printing characters in the data stream after
the CR/LF sequence, it would print the next character about half way
back to the beginning of the next line.

If you sent only CR as line terminator, you got all lines printed on
the same line and eventually ate a hole in the paper. If you sent only
LF, the carriage would shortly reach the far right margin and print
each line in one character position on separate lines.

Early design decisions, however well considered, sometimes remain to
encumber the future, often for reasons of compatibility. The
constraints of available technology at the time of an innovation can
play a crucial role in this process.

Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/

More information about the Cygwin mailing list