On Sat, Oct 03, 2026 at 07:58:07PM -0700, Per Bothner wrote: > On 10/3/26 14:38, Gavin Smith wrote: > > The claim that files should CRLF line endings should always be treated the > > same as LF line endings appears to be based on wishful thinking rather than > > reflecting common practice. > > I made no claim about "common practice". My claim is that "best practice" > is to treat LF and CRLF the same in most cases. There may be some cases where > compatibility or historical or other reasons make that difficult, > but I don't think there are any such reasons that apply for texinfo source > files. > > What is the alternative? Is there any reason for a texinfo-processing program > to handle CRLF differently from LF?
My concern is reliability and maintainability. > If you prefer you can treat this as a special case of white-space at the end > of lines: > If a program acts differently based on what kind if any invisible white-space > there is > at the end of a line, that is a design flaw. Example: if a program uses > backslash-newline > as a line-continuation indicator, it should allow and ignore arbitrary > horizontal white-space > (space, tab, or cr) after the backslash. That's my preference. > All the text editors I tried (emacs, gedit, kate, vim - on Fedora) silently > open files > with CRLF line endings, silently removing the CR. I.e. cursor motion > does not act as if there is an extra white-space character at the end of each > line. > > Output files are a different matter. Generally, I believe it is OK to write > output files using just LF even on Windows. (However, I don't have enough > Windows experience to be sure,) > -- > --Per Bothner > [email protected] http://per.bothner.com/ >
