Alan Manuel Gloria suggested: > > 1. NLF (new line function) is LF | CR | CR LF | NEL > > 2. Treat any kind of NLF the same. > > 3. "A readline function should stop at NLF, LS, FF, or PS"
John Cowan: > Take it from the guy at the sharp point of the XML 1.1 mess: the > only newlines anyone cares about are CR, LF, and CR+LF, and even CR is > obsolescent. NEL is used only on EBCDIC systems, and conversion to ASCII > usually changes it to LF rather than U+0085. LS was Unicode's attempt > to kill CR/LF/CR+LF, which failed completely. Nobody uses FF or PS. > (CR NEL, which is supported in XML 1.1, is the result of half-assed > round-tripping of CR+LF via EBCDIC.) > > Should we adhere to the Unicode specs more closely, even at the cost > > of a more annoying Guile behavior? > > No. I agree with John Cowan on this one. Treating NLF (aka the EOL sequence) as LF | CR | CR LF handles the only cases anyone cares about, and even lone CR is rare. Perhaps most importantly, these are the only sequences defined in R7RS draft 9 as a "line ending", so it's consistent with where Scheme is going. I think we do NOT want to claim that a line ends at FF or VT. Typical text editors treat these as *characters* that are *on* a line, not as line-ending characters by themselves, There's not much call for supporting VT, but there's still some call for supporting FF, and supporting both VT and FF the same way costs practically nothing and is by some views the right thing to do (so we may as well do it). --- David A. Wheeler ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_feb _______________________________________________ Readable-discuss mailing list Readable-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/readable-discuss
