Someone recently contacted me because my Thunderbird "undigestify" add-on <https://addons.mozilla.org/en-US/firefox/addon/undigestify/> is failing to unpack digests produced by Mailman 2.1.9.
He sent me a sample digest, which ends like this:
------------------------------
--
test mailing list
test@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/test
End of test Digest, Vol 93, Issue 20
************************************RFC 1153 has this to say about the digest trailer (emphasis added by me):
The Trailer immediately follows the blank line of message separator
following the last enclosed message. The Trailer consists of two
lines. The first line /*must */begin with the words, End of,
followed by the listname a blank and the word Digest which is
usually followed by volume and issue number on the same line. The
second and last line of the Trailer and the entire message is a line
of asterisks serving to underline the line immediately above it.I looked at the mailman source code and it doesn't appear that the Fedora Project is running a hacked version; it appears that the standard Mailman source code inserts the list footer as shown above as part of the digest trailer, in violation of the RFC.
I can and will work around this violation of the standard in my code, but it makes the code significantly more complicated, and it's kind of annoying to see the RFC being violated like this.
jik
Hello list,
I'm new to mailman and have recently installed the software on my CenOS server.
All is working quite well, accept for the specified join & leave dedicated addresses for members to subscribe/unsubscribe.
When sending to ios-join@a11y.us, the message is rejected, as a "mailbox doesn't exist" type of error. Same goes for ios-subscribe@a11y.us.
Any ideas on what I should explain to my I.T rep to review, so these requests can work?
Thank you.
Carlos
Hello list,
I'm new to mailman and have recently installed the software on my CenOS server.
All is working quite well, accept for the specified join & leave dedicated addresses for members to subscribe/unsubscribe.
When sending to ios-join@a11y.us, the message is rejected, as a "mailbox doesn't exist" type of error. Same goes for ios-subscribe@a11y.us.
Any ideas on what I should explain to my I.T rep to review, so these requests can work?
Thank you.
Carlos
As an update to this issue:
We've discovered that replying to the verification emails also results in mail rejection.
Any ideas or exact details on what is wrong would be highly appreciated.
Carlos
On Nov 7, 2011, at 12:39 PM, Carlos Palomino wrote:
Hello list,
I'm new to mailman and have recently installed the software on my CenOS server.
All is working quite well, accept for the specified join & leave dedicated addresses for members to subscribe/unsubscribe.
When sending to ios-join@a11y.us, the message is rejected, as a "mailbox doesn't exist" type of error. Same goes for ios-subscribe@a11y.us.
Any ideas on what I should explain to my I.T rep to review, so these requests can work?
Thank you.
Carlos
Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://wiki.list.org/x/AgA3 Security Policy: http://wiki.list.org/x/QIA9 Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/ Unsubscribe: http://mail.python.org/mailman/options/mailman-users/carlos.palomino%40me.co...
Carlos Palomino wrote:
As an update to this issue:
We've discovered that replying to the verification emails also results in mail rejection.
Any ideas or exact details on what is wrong would be highly appreciated.
Your incoming MTA needs aliases or other programmatic instruction for delivering to all 9 (-admin, -bounces, -confirm, -request, -owner, -join, -leave, -subscribe and -unsubscribe) of the LISTNAME-*@... addresses in addition to LISTNAME@....
As Adam McGregor suggested, see section 6 of the installation manual at <http://www.list.org/mailman-install/mail-server.html>.
-- Mark Sapiro <mark@msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
As an update to this issue:
We've discovered that replying to the verification emails also results in mail rejection.
Any ideas or exact details on what is wrong would be highly appreciated.
Carlos
Original message:
Hello list,
I'm new to mailman and have recently installed the software on my CenOS server.
All is working quite well, accept for the specified join & leave dedicated addresses for members to subscribe/unsubscribe.
When sending to ios-join@a11y.us, the message is rejected, as a "mailbox doesn't exist" type of error. Same goes for ios-subscribe@a11y.us.
Any ideas on what I should explain to my I.T rep to review, so these requests can work?
Thank you.
Carlos
Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://wiki.list.org/x/AgA3 Security Policy: http://wiki.list.org/x/QIA9 Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/ Unsubscribe: http://mail.python.org/mailman/options/mailman-users/carlos.palomino%40me.co...
On Mon, Nov 07, 2011 at 12:39:41PM -0500, Carlos Palomino wrote:
I'm new to mailman and have recently installed the software on my CenOS server.
All is working quite well, accept for the specified join & leave dedicated addresses for members to subscribe/unsubscribe.
When sending to ios-join@a11y.us, the message is rejected, as a "mailbox doesn't exist" type of error. Same goes for ios-subscribe@a11y.us.
Configuring your MTA?
(add a second (and others, should you so wish) nameserver, too)
Any ideas on what I should explain to my I.T rep to review, so these requests can work?
Did you get to step 6 of the installation instructions? http://bazaar.launchpad.net/~mailman-coders/mailman/2.1/view/head:/doc/mailm...
-- "Sitting in a church doesn't make you a Christian, any more than sitting in a garage makes you a car."
Jonathan Kamens wrote:
He sent me a sample digest, which ends like this:
-- test mailing list test@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/test
End of test Digest, Vol 93, Issue 20
RFC 1153 has this to say about the digest trailer (emphasis added by me):
[...]
I looked at the mailman source code and it doesn't appear that the Fedora Project is running a hacked version; it appears that the standard Mailman source code inserts the list footer as shown above as part of the digest trailer, in violation of the RFC.
And if you looked at the source, you should have seen the following comment immediately preceding the adding of the footer.
# BAW: This is not strictly conformant RFC 1153. The trailer is only # supposed to contain two lines, i.e. the "End of ... Digest" line and # the row of asterisks. If this screws up MUAs, the solution is to # add the footer as the last message in the RFC 1153 digest. I just # hate the way that VM does that and I think it's confusing to users, # so don't do it unless there's a clamor.
I can and will work around this violation of the standard in my code, but it makes the code significantly more complicated, and it's kind of annoying to see the RFC being violated like this.
If you are intending to raise a clamor, you can file a bug report at <https://bugs.launchpad.net/mailman/+filebug>.
Suppose the above were changed to
------------------------------
Digest-Footer:
--
test mailing list
test@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/test
------------------------------
End of test Digest, Vol 93, Issue 20
************************************Would your code handle that or would it complain the Digest-Footer: is not an allowable header and/or none of Date:, From:, To:, Cc:, Subject:, Message-ID: and Keywords: were included?
What if it were just
------------------------------
--
test mailing list
test@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/test
------------------------------
End of test Digest, Vol 93, Issue 20
************************************-- Mark Sapiro <mark@msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
On 11/07/2011 10:24 PM, Mark Sapiro wrote:
Jonathan Kamens wrote:
He sent me a sample digest, which ends like this:
------------------------------ -- test mailing list test@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/test End of test Digest, Vol 93, Issue 20 ************************************RFC 1153 has this to say about the digest trailer (emphasis added by me): [...] I looked at the mailman source code and it doesn't appear that the Fedora Project is running a hacked version; it appears that the standard Mailman source code inserts the list footer as shown above as part of the digest trailer, in violation of the RFC.
And if you looked at the source, you should have seen the following comment immediately preceding the adding of the footer.
# BAW: This is not strictly conformant RFC 1153. The trailer is only # supposed to contain two lines, i.e. the "End of ... Digest" line and # the row of asterisks. If this screws up MUAs, the solution is to # add the footer as the last message in the RFC 1153 digest. I just # hate the way that VM does that and I think it's confusing to users, # so don't do it unless there's a clamor. LOL. Yes, I was looking right at that section of code, and my eyes just glossed right over the comment because my brain knew exactly what the code in question was doing and therefore saw no need to read the comments documenting it. D'oh.
I can and will work around this violation of the standard in my code, but it makes the code significantly more complicated, and it's kind of annoying to see the RFC being violated like this.
If you are intending to raise a clamor, you can file a bug report at <https://bugs.launchpad.net/mailman/+filebug>. I have raised as much of a clamor as I intended to. "I just hate the way that VM does that" is not in my mind a legitimate reason to violate the RFC. If you agree, then you can create your own bug report and fix it. If you disagree, then there's no point in my creating one. Suppose the above were changed to
------------------------------ Digest-Footer: -- test mailing list test@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/test ------------------------------ End of test Digest, Vol 93, Issue 20 ************************************Would your code handle that or would it complain the Digest-Footer: is not an allowable header and/or none of Date:, From:, To:, Cc:, Subject:, Message-ID: and Keywords: were included? My code would have handled that without complaint. What if it were just
------------------------------ -- test mailing list test@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/test ------------------------------ End of test Digest, Vol 93, Issue 20 ************************************I think it would have handled that without complaint as well... The second blank line after the line of hyphens would have been interpreted as the terminating line after an empty header.
Note that I say "would have handled" in both of these cases because I've already modified my code to accommodate the non-complaint Mailman functionality. I would have to do that regardless of whether you fixed Mailman to be compliant now, because such a fix would not make it out to a substantial majority of Mailman sites for several years, even if you released it today.
Thanks for the reply.
jik
On Nov 07, 2011, at 10:39 PM, Jonathan Kamens wrote:
I have raised as much of a clamor as I intended to. "I just hate the way that VM does that" is not in my mind a legitimate reason to violate the RFC. If you agree, then you can create your own bug report and fix it. If you disagree, then there's no point in my creating one.
https://bugs.launchpad.net/mailman/+bug/887610
Right now it's milestoned to MM3.0beta1, and I'll leave it up to Mark to add a bug task for MM2.1.
Cheers, -Barry
On Nov 07, 2011, at 07:24 PM, Mark Sapiro wrote:
And if you looked at the source, you should have seen the following comment immediately preceding the adding of the footer.
# BAW: This is not strictly conformant RFC 1153. The trailer is only # supposed to contain two lines, i.e. the "End of ... Digest" line and # the row of asterisks. If this screws up MUAs, the solution is to # add the footer as the last message in the RFC 1153 digest. I just # hate the way that VM does that and I think it's confusing to users, # so don't do it unless there's a clamor.
Just to mention, I haven't used VM in a bajillion years, and in fact I very rarely use digests at all these days. So while this was very definitely a case of "I get to make my playground work for me", I have no problem with this code getting fixed to actually support the RFC, and not allowing a workaround for an obscure MUA break things for everyone else.
Cheers, -Barry
participants (5)
-
Adam McGreggor -
Barry Warsaw -
Carlos Palomino -
Jonathan Kamens -
Mark Sapiro