summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--Mailman/Queue/NewsRunner.py10
1 files changed, 5 insertions, 5 deletions
diff --git a/Mailman/Queue/NewsRunner.py b/Mailman/Queue/NewsRunner.py
index bc247bdf0..3626f838b 100644
--- a/Mailman/Queue/NewsRunner.py
+++ b/Mailman/Queue/NewsRunner.py
@@ -100,7 +100,6 @@ def prepare_message(mlist, msg, msgdata):
else:
# Newsgroups: isn't in the message
msg['Newsgroups'] = mlist.linked_newsgroup
- #
# Note: We need to be sure two messages aren't ever sent to the same list
# in the same process, since message ids need to be unique. Further, if
# messages are crossposted to two Usenet-gated mailing lists, they each
@@ -121,16 +120,17 @@ def prepare_message(mlist, msg, msgdata):
if hackmsgid:
del msg['message-id']
msg['Message-ID'] = Utils.unique_message_id(mlist)
- #
# Lines: is useful
if msg['Lines'] is None:
# BAW: is there a better way?
count = len(email.Iterators.body_line_iterator(msg))
msg['Lines'] = str(count)
- #
- # Get rid of these lines
+ # BAW: It appears that some news servers have problems if the message
+ # contains X-Trace:, NNTP-Posting-Host: or duplicate Cc: headers. Should
+ # we retain X-Trace or NNTP-Posting-Host as backup X-* headers?
del msg['received']
- #
+ del msg['x-trace']
+ del msg['nntp-posting-host']
# BAW: Gross hack to ensure that we have only one
# content-transfer-encoding header. More than one barfs NNTP. I don't
# know why we sometimes have more than one such header, and it probably