[Mailman-Users] Long subject lines get extra padding from mailman

[Mark Sapiro]

Miles Waller wrote:
>
>I have verified this problem in mailman versions 2.1.2 and 2.1.5.  The 
>problem seems to be that the second and subsequent lines of a long 
>header are indented when they are folded onto new lines.  From my basic 
>testing, the isseu seems to be related to the presence of the indents, 
>rather than the actual line breaks.
>
>The impact is that several mail clients (tested thunderbird 0.9, outlook 
>express 5.5, outlook 2000) try faithfully to display all the extra 
>spaces between words, causing a gap to appear in the subject line which 
>can be quite noticeable in certain circumstances.

I wouldn't rely on lookout - er outlook or outlook express to provide
an example of correct behavior in am MUA.

>Here is a sample of the headers:
>
<some snips> 
>Subject: This is a really long subject line. The quick brown fox jumped
>	over the lazy dog, but the dog wasn't so lazy after all! The fox had roused
>	a sleeping giant, so to speak, and that dog jumped right up and bit him on
>	the behind! Not quick enough, Mr Fox!
>X-BeenThere: previewlist-text at localhost.localdomain
<more snips>
>
>Is this a bug in mailman, or is it something in the list settings that I can change?
>Has anyone else had this problem, and how have they worked round it?

Actually, the indents are a single horizontal-tab character. The
process of folding and unfolding long headers is described in RFC
2822, sec 2.2.3
(http://www.cse.ohio-state.edu/cgi-bin/rfc/rfc2822.html#sec-2.2.3).
The standard suggests that folding SHOULD be done at higher level
syntactic breaks, but there are no such things in Subject: headers.

I think it is actually wrong of Mailman to effectively change a "space"
to a "tab" when it folds a Subject: header, so in that sense this is a
bug, but that said, different MUAs will react differently to this.
Many will still unfold in the RFC 822 way which drops the CRLF and the
following white space character which results in words being run
together at the folds. Others will just drop the CRLF which is the RFC
2822 way and effectively results here in a "space" being replaced by
"tab".

Everyone has the problem. I think most people just ignore it. The bug
has been reported -
http://sourceforge.net/tracker/index.php?func=detail&aid=736559&group_id=103&atid=100103

--
Mark Sapiro <msapiro at value.net>       The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan