Member Submission - Note

Date: Mar. 14, 2001
Link: http://www.w3.org/TR/newline
Source Author or Organization: IBM

This Note addresses problems with IBM mainframes processing XML that arise from the omission of the Unicode 3.0 [NEL] (Newline, or next line) character from XML 1.0's list of line ending characters.

The omission causes errors in OS/390 systems, and computers interacting with them, during Document Type Definition (DTD) and XML document processing, including: documents arriving via FTP from UNIX systems being declared invalid or not well-formed, documents created on OS/390 systems and retrieved by DOS or UNIX systems via Java Database Connectivity (JDBC) rejected by parsers because [NEL] characters are not interpreted as legal white space characters, and errors from OS/390 native system string functions recognizing [NEL], but not [LF] (line feed, or line separator), characters as white space.

Suggested text is offered to modify XML 1.0 Second Edition so that [NEL], [LF] and the two-character sequence [CR][NEL] (paragraph separator) are all recognized as legal white space characters.