"Paul J. Schinder" <schinder@pjstoaster.pg.md.us> writes >But it *does* accept a reasonable definition of line-ends, the only one >that makes any sense: accept the ones on the platform you're on, but no >others. That's what the base Perl code must do, that's what MacPerl does. I'd question this - in what way would supporting CR|LF|CRLF affect how perl interprets the script? There would be the issue of whether to keep the line-end for here-text or translate it into the platform one, but anything else? PostScript for instance has platform-independent line-ends in the interest of script portability. The ANSI C standard is silent on this issue, but Java explicitly accepts CR|LF|CRLF as a line-terminator. Note I'm talking about scripts here, not input files where the programmer can adjust Perl's idea of line-ends to suit. >And it's far easier to solve this "problem" by a few lines of Perl rather >than making major changes to the Perl interpreter itself. sure it's easy enough to translate line-ends, but the question was multi-part 1) is it easy to add line-agnosticism to perl? Without looking at the code-base, it's not unlikely that there's only one place that reads in script lines, in which case it would be easy making the change. 2) if it's not easy, then it would be nice to have a warning if the interpreter found only one line. Because if your script starts with a shebang, the whole script is treated as a comment line and the naive user wonders why nothing happens - at present it seems to *silently fail* which is rarely a user-friendly approach. cheers, Danny Thomas PS I doubt it's related to line-ends, but the current MacPerl often reports the wrong line number in warnings/errors. Has anyone else noticed this? Next time it happens I'll try an check that the discrepancy isn't coming from a block of text with non-platform ends-of-line. ***** Want to unsubscribe from this list? ***** Send mail with body "unsubscribe" to mac-perl-request@iis.ee.ethz.ch