On Tue, Nov 18, 2008 at 9:51 PM, Andreas J. Koenig <[EMAIL PROTECTED]> wrote: >>>>>> On Tue, 18 Nov 2008 21:13:35 -0500, "David Golden" <[EMAIL PROTECTED]> >>>>>> said: > > > I've been meaning to hack at Test::Reporter to wrap long lines, but I > > really have no idea if that will work or not. > > But isn't wrapping long lines an evil approach to an unverified > assumption? Fom the nature of what we are doing, long lines are to be > expected, like long diagnostics lines of long lines in perl -V, etc.
Wrapping is evil, except that the SMTP protocol does specify a maximum line length of 1000 characters that I think some reports could be exceeding. Net::SMTP doesn't actually check so it just takes whatever it's given and puts it on the wire. As I said, it's a hypothesis and I don't know that it will work, but at the same time, if we're violating SMTP, that could explain why this is a heisenbug for some testers but not others. -- David
