Mailman on Mac OS X Server 10.3: Outgoing messages stuck in qrunner/in folder
![](https://secure.gravatar.com/avatar/30ddf4e6b418c22b78fbe3ecdc0ec256.jpg?s=120&d=mm&r=g)
Hi,
I am running Mailman on an Xserve running Mac OS X Server 10.3.9. The
version of Mailman is the one included with Mac OS X Server 10.3.9.
(It's 2.1.2.)
I've just started experimenting with mailing lists on this server
using Mailman. Things worked fine until I actually got to the stage
of mass-subscribing a series of nearly 500 email addresses.
I used the web-based Mailman interface to submit the list of 500
email addresses for new subscribers, through the "Mass Subscription"
web form. I just copied the list of email addresses (separated by
return chars) from a text editor and pasted it in the field and
submitted. My mistake (I think) was that I submitted the entire list
of 500 email addresses at once. Initially it SEEMED to work, i.e. I
got a confirmation page listing all the subscribers that had been
added to the list.
But then I became unable to access the admin page for the list
altogether. The server would simply not respond. I could still access
other parts of the web-based Mailman interface, but not the pages for
this particular list.
After a bit of on-line research, I discovered that the problem was
probably due to "locks" in the
/var/mailman/locks/
folder. Indeed, when I looked inside that folder, I found a whole
series of files named:
enseignants.lock enseignants.lock.www.cprp.ca.421.1 enseignants.lock.www.cprp.ca.7150.0 enseignants.lock.www.cprp.ca.7166.0 enseignants.lock.www.cprp.ca.7216.0 enseignants.lock.www.cprp.ca.7228.0 enseignants.lock.www.cprp.ca.7286.0 enseignants.lock.www.cprp.ca.7470.0 enseignants.lock.www.cprp.ca.7696.0 enseignants.lock.www.cprp.ca.7819.0 enseignants.lock.www.cprp.ca.7918.0
("www.cprp.ca" is my server and "enseignants" is the name of the list)
The instructions for dealing with this were not exactly clear, so I
just removed all these files so that the "locks" folder was empty.
After that, I was able to access the list's admin pages again. So I
thought everything was good.
Then I tried to send a message to my subscribers. (It's a receive-
only list, so I'll be the only one sending messages.) However, the
message that I sent never got sent to the mailing list's subscribers.
After more online research, I found out that my message was actually
stuck inside the
/var/mailman/qfiles/in/
folder. I also found that I could force Mailman to send the message
by using the command:
/usr/share/mailman/bin/qrunner -r All
That worked and the message got sent out to the list (although I then
had to control-C the process to return to the prompt).
But I find that this now happens every time I want to send a message.
The message never gets sent automatically as expected. Instead, it
gets stuck in that /qfiles/in/ folder and I have to run the qrunner -
r All command to get it sent.
This never happened with this list before I added the 500 subscribers
and had to fiddle with the /locks/ folder. Prior to that, when I was
still in the stage of testing the mailing list with only a handful of
test subscribers, everything worked fine and the messages would get
sent automatically.
My suspicion is that the problem is due to what I did in the /locks/
folder, that maybe I wasn't supposed to remove all the files. (I have
kept them on a local disk in case I need to move them back to that
folder.)
But unfortunately further on-line research and a search through the
mailman-users archives didn't return anything that would help me. I
found people describing various similar problems, but the
troubleshooting steps seem suggested totally foreign to my actual
problem.
Hence this message. I would REALLY appreciate it if someone could
help me with this. I simply do not have the required Mailman
expertise. I just want to use Mailman for a single announcement list.
I suppose I could erase everything and start from scratch again, but
maybe someone here has an easier solution.
Thanks in advance for any help.
Pierre Igot
Pierre Igot, administrateur des systèmes / Systems Administrator www.cprp.ca 902-837-7391
![](https://secure.gravatar.com/avatar/746f7519ba02fb0d815e59f305c53fa2.jpg?s=120&d=mm&r=g)
Pierre Igot wrote:
See <http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.021.htp>
Yes, it worked. The hangup was probably because of trying to send notifications.
Yes, the problem was these locks.
So the mailmanctl daemon and the qrunners aren't running unless you start the qrunners manually as above, but this isn't the way to do it.
If you had done
/usr/share/mailman/bin/mailmanctl start
you wouldn't have had to kill the qrunners.
I think in Mac OSX server, Mailman is installed as a service and the service runs mailmanctl. Somehow you have stopped the service and you need to restart it.
I doubt that removing the locks had anything to do with it, but you might check ALL of Mailman's log files to see what kinds of errors and other conditions occurred.
-- Mark Sapiro <msapiro@value.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
![](https://secure.gravatar.com/avatar/30ddf4e6b418c22b78fbe3ecdc0ec256.jpg?s=120&d=mm&r=g)
On 06-12-13, at 19:31, Mark Sapiro wrote:
I am aware of this page and checked it before sending my request for
help. The page didn't provide any information that addressed my
issue, as far as I could tell.
Right--but I didn't see any error messages about anything. (I checked
all the logs.) So it was a failure that was not exactly graceful :).
OK, so based on what you are saying, deleting these locks doesn't
cause any further problems, right?
Well, there was nothing anywhere that indicated that I had to run
this command :).
Like I said, before the mass subscriptions, everything ran fine and I
never had to run this command. If "mailmanctl" somehow stopped, it
certainly didn't notify me. And there was nothing in the user
interface for Mailman that indicated that this was the problem.
Running this command did indeed fix the problem. The messages are now
sent out automatically. Thanks!
Well, I did not stop the service intentionally. The Mailman service
is part of the "Mail" service (which also controls POP and SMTP) and
there is nothing in the server admin software that indicates that
mailmanctl is not running. Probably a shortcoming of Apple's server
admin software.
In any case, now I know that if things are not going out, I need to
check if this "mailmanctl" is running. Thanks.
It's the kind of thing that is probably obvious to most readers of
this list, but I had to subscribe to the list and ask the question to
find out myself (or spend many hours reading the entire Mailman
documentation, I suppose).
OK. I did check all the log files, and didn't see any significant
errors.
Thanks again.
Pierre
Pierre Igot, administrateur des systèmes / Systems Administrator www.cprp.ca 902-837-7391
![](https://secure.gravatar.com/avatar/746f7519ba02fb0d815e59f305c53fa2.jpg?s=120&d=mm&r=g)
Pierre Igot wrote:
The relevant part of that FAQ is the part that says
All other support issues relating to using the Apple-provided versions of Mailman under MacOS X should be first directed to Apple, since they are the ones that modified the Mailman installation to suit their environment and they have not shared their changes with us.
There are other FAQ articles
Mark Sapiro <msapiro@value.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
![](https://secure.gravatar.com/avatar/746f7519ba02fb0d815e59f305c53fa2.jpg?s=120&d=mm&r=g)
Mark Sapiro wrote:
Sorry, I sent that before I finished it. I meant to say
There are other FAQ articles such as <http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq03.014.htp> that may have been helpful, but we don't know if they are applicable to the Mac OS-X Server Mailman or not because it is a modified version of our package, and we don't know what the mods are.
WRT to mailmanctl, apparently the Mailman qrunners were running at some time and then stopped. I don't know how they were started in the first place or why they stopped. Perhaps a reboot of the server would have started them again, perhaps not, but IMO, the fact that all this is not documented for you is Apple's failure, not ours.
I think I can understand your frustration, and I am not trying to be unsympathetic. If you had installed our package, and found our installation manual lacking, we would do our best to supplement it or improve it. However, we really can't create documentation for packages whose details we have even less information about than you do. </rant>
-- Mark Sapiro <msapiro@value.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
![](https://secure.gravatar.com/avatar/d5a34df919f6a79644d998b34dcee064.jpg?s=120&d=mm&r=g)
http://www.list.org/mailman-install/node50.html
I'm not sure just how much this may help the original poster or not.
--
Bye for now, Terry Allen
___________________________________________________________________
hEARd
Postal Address: hEARd, 26B Glenning Rd, Glenning Valley, NSW 2261, Australia Internet - WWW: http://heard.com.au http://itavservices.com EMAIL: hmag@ozemail.com.au Phone: Australia - 02 4388 1400 / International - + 61 2 43881400 Mobile: Australia - 04 28881400 / International - 61 4 28881400
Non profit promotion for new music - since 1994
![](https://secure.gravatar.com/avatar/01aa7d6d4db83982a2f6dd363d0ee0f3.jpg?s=120&d=mm&r=g)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Dec 14, 2006, at 3:18 AM, Terry Allen wrote:
I sent some docs through to Barry Warsaw a long time back, but some oif the info contained in it are here:
Hi Terry, would you consider fleshing out OS X support on the Mailman
wiki? That's a much better place to maintain this type of
information than in the docs.
- -Barry
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin)
iQCVAwUBRYFqOnEjvBPtnXfVAQKr6AQAs8GAEs/LkA4+iPA1wb73ZkynjlLosX9Y dg6qSSNynLgSLffy0xB/e5VXCwx/B+q9iA8ZKfzpkRbjTywpDb0t04h8V0KjvYQz De0In3pS4MMj0OAHpINwPBUS48MQ4lMvAj4kyuwKW4Gb8EsmbKSGil8TTlN9kb// MZITo2oHEic= =FwrF -----END PGP SIGNATURE-----
![](https://secure.gravatar.com/avatar/30ddf4e6b418c22b78fbe3ecdc0ec256.jpg?s=120&d=mm&r=g)
On 06-12-14, at 03:51, Mark Sapiro wrote:
Right, I understand your point of view perfectly.
On the other hand, the reality is that I, as a user, have to deal
with "Apple's failure," as you call it, and that the best avenue for
dealing with it, as far as I could tell, was to ask for help on this
list. The fact that you were able to help me so quickly and with a
single instruction proves my point :).
I *could* have submitted a bug report to Apple and waited for weeks
or months until I got a reply.
I *could* have spent hours searching through Apple's "Discussions"
forum on Mac OS X Server's Mail service, trying to find posts with
the exact same issue, knowing full well that I barely had enough
expertise to figure out the best keywords to use to describe the
problem. ("messages," "stuck," "in folder" are not exactly very
effective keywords, in that they are very generic)
I *could* have posted a request for help on the forum and waited for
days to see if, by chance, anyone might be able to figure out what
was going on on my machine.
I *could* have spent hours reading the Mailman documentation myself
until I had figured out the fundamentals of Mailman myself.
But the reality was that, in all likelihood, this was a pretty basic
problem that had to do with Mailman itself, that it probably had
nothing to do with Apple's customizations, and that the most
efficient way to get help was probably to submit a request to this list.
And, as I said, the quick (if reluctant) response you gave me proved
my point.
The question is whether it really costs Mailman experts such as
yourself so much effort to just provide such basic help from time to
time to "newbies" like me using a possibly somewhat non-standard
version of Mailman. (Based on what I have read, I don't think there
is much difference between the Mailman 2.1.2 installed with Mac OS X
Server 10.3 and your packages. It's just that everything is already
included and installed, and can be turned on through the Server Admin
GUI. And some file/folder locations are somewhat different from the
standard Unix locations. But all the rest is probably the exact same.)
For what it's worth, if you ever need any basic help with some Mac-
specific Mac OS X stuff, which is my expertise, I will be more than
glad to help, even if you are not using the exact version of Mac OS X
that I am currently working with :).
Thanks again.
Pierre
-- Pierre Igot, administrateur des systèmes / Systems Administrator www.cprp.ca 902-837-7391
![](https://secure.gravatar.com/avatar/b273ab068bc220d17a3e4c710c401c4b.jpg?s=120&d=mm&r=g)
[not speaking for Mark, Barry, or really anyone but me]
This is really the classic IT problem of unsupported configurations.
In a flurry of recycled electrons, Pierre Igot wrote:
The problem that comes up far too often is that someone, such as Mark, says "what's in the vette log?" or "delete /usr/local/mailman/lists/fred.lock" and the questioner says "where is it?" or "But I don't -have- a /usr/local/mailman" directory". We then start playing 20 questions- do you have root access? which version is installed? which pakcaged version is installed? etc. (iirc, all if which is in faq 3.14)
Given that mailman and the support are free, I think it's a bit much to expect that the maintainers know exactly how each distributor builds their package (directories, options, etc). They can only say "on a stock system, it works like this", and sometimes "there's a vette log somewhere, please find it".
z! former IT manager
![](https://secure.gravatar.com/avatar/30ddf4e6b418c22b78fbe3ecdc0ec256.jpg?s=120&d=mm&r=g)
On 06-12-14, at 13:18, Carl Zwanzig wrote:
Believe it or not, I did check the FAQ before submitting my request.
Try searching for all keywords "message stuck qfiles/in" or simply
for "qfiles" in the FAQ, for example. It doesn't return anything that
would have helped in my case.
OK, but that didn't happen in my case as far as I know. My problem
did not require any references to Mac OS X-specific things. I only
provided the Mac-specific information as a courtesy, because it was
part of list etiquette.
Mark provided me with a one-line Unix command that happened to work
with my version of Mailman in Mac OS X, but if it hadn't worked, I
would have simply tried to find where "mailmanctl" was on my system,
without bothering the list with such a detail. As it happens, I
didn't have to do this, but I would have done it if required. So I
don't feel that there was anything in my case that forced anyone to
deal with non-standard issues.
And that's all that I was asking for. I didn't ask for anything
beyond this.
Pierre
-- Pierre Igot, administrateur des systèmes / Systems Administrator www.cprp.ca 902-837-7391
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 2:03 PM -0400 12/14/06, Pierre Igot wrote:
Believe it or not, I did check the FAQ before submitting my request.
I do believe it.
The problem is that we cannot know beforehand all the possible search terms that people might use to describe any given particular type of problem. We can guess, and we can try to capture what we think are the most likely types of descriptions, but that's always a guess and there will always be someone who doesn't make the same guess and instead searches for something else. We can get better at covering more potential different descriptions, but we cannot ever be perfect.
Actually, there is an Apple-specific issue here. Apple has their own way of starting and stopping the Mailman queue runners and associated processes. If your queue runners aren't running, then you're going to have the kinds of problems that you saw -- regardless of what method is used to stop or start them.
If you're using the Apple-provided code to manage Mailman, you're not going to know anything about the Mailman queue runners. All you're going to know is that something is broken.
So, we have to work backwards from the "it's broke" state to figure out what is broken and who is responsible for that code. Now that we know your problem was a result of the queue runners not actually running, we can tell you that you need to start the queue runners, but since you're using MacOS X Server we can't tell you precisely what the "correct" way is to start them. We can tell you what the Mailman-standard was is to start them, but Apple has created their own code to manage this aspect of Mailman operations and they haven't shared that with us.
Which leads us to the general issue of people expecting us to provide support for a system that someone else has taken and modified, but without sharing any of those changes with us.
Even if the core Mailman system itself is unchanged from what we shipped, they have changed the interface that is used to manage Mailman, but none of the people on this list know anything about it.
Hell, the guy who runs lists.apple.com doesn't know anything about it, and he's running Mailman on machines at Apple that are running MacOS X Server, for Apple users on Apple-specific products -- but even he's not using the Apple-provided version of Mailman.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/f10bf4ea18a58c83f99e6d5a4b12e322.jpg?s=120&d=mm&r=g)
Brad Knowles wrote:
Apple does provide the source code for their mailman packages. You can browse it here:
http://www.opensource.apple.com/darwinsource/Current/mailman-117/
And download a tarball of that stuff:
http://www.opensource.apple.com/darwinsource/tarballs/other/mailman-117.tar....
AFAICT, the -117 is an Apple internal number. Browsing the full list shows similar, non-upstream numbers for many other packages.
The full list of open source code for Mac OS X 10.4.8 for PPC is at:
http://www.opensource.apple.com/darwinsource/Current/
So, while it seems fair to say that many of the very knowledgeable folks here don't know the specifics of the Apple packaged mailman, it doesn't seem quite as fair to say that Apple hasn't shared their changes. They are available for anyone to download and check out.
Compare that to cPanel where the best you find if you're not a customer is a half-backed and outdated patch (which I was provided only after persistent requests).
Disclaimer: I don't know anything about the Apple packages and the only Apple product I own is an iPod. :)
-- Todd OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp
Don't take life seriously, you'll never get out alive.
![](https://secure.gravatar.com/avatar/334b870d5b26878a79b2dc4cfcc500bc.jpg?s=120&d=mm&r=g)
Todd Zullinger writes:
Brad Knowles wrote:
We're not talking about "shared" as you will find it defined on webster.com, this is "shared" as used by Richard Stallman.[1] "Sharing" is active contribution to the mainline, it's not merely posting your code for others to pick up and analyze.
This usage makes some sense, because the creators of a forked product have already done most of that analysis. That doesn't mean they're not liable to substantial additional work, of course, and open source doesn't oblige them to do it. But by the same token the Mailman developers are forced (by lack of resources to do such analysis themselves) to take the position that this code hasn't been contributed to the project, so they can't know much about it.
It's not "fair", but that's the customary usage of the verb "to share" in the parts of this community that I frequent.
Sure, Apple is doing the open source thing by the letter of the law, and that is no small thing. But the problem presented to the maintainers of the mainline by Apple's fork is of the same variety as the problem presented by cPanel.
Footnotes: [1] No, I don't approve of this usage, that's why I attribute it to a famous extremist. "Extremism in the defense of freedom is no vice."<wink>
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 5:32 PM -0500 12/14/06, Todd Zullinger wrote:
Apple does provide the source code for their mailman packages. You can browse it here:
Even if they have made the source code available for everything they've done with regards to Mailman (which includes all their proprietary management tools), this is not the same thing as contributing that code back to the Mailman project.
Likewise, webmin also includes a component for managing Mailman, and all their code is publicly available, but we don't say that webmin has contributed all their code back to the Mailman project.
Same any other group that takes our code and makes modifications to it -- even if they make all their source code publicly available, that's not the same thing as contributing it back to the Mailman project.
That is worse, yes. But that's also not the point.
Disclaimer: I don't know anything about the Apple packages and the only Apple product I own is an iPod. :)
Whereas I was an Apple fan from 1982, back in the original Apple ][ days, and before the introduction of the 16K language card, the Apple II+, the Apple IIe, the Apple //c, or any other more recent Apple product. I've been a MacFanatic since December of 1983 when I saw an early prototype behind closed doors. I've been a Mac owner since the day when I bought a Mac SE -- before the internal hard drive model was available.
I've got a long history with the company, but I'm not blind to things that they do poorly. At least, not totally blind.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/f10bf4ea18a58c83f99e6d5a4b12e322.jpg?s=120&d=mm&r=g)
Brad Knowles wrote:
But what value would MacOSX specific integrations be to the Mailman project? I think that they'd likely be of limited value. Just as Debian or Red Hat specific changes are. I know that John Dennis of Red Hat contributed the changes Red Hat made to make mailman fit in better with the FHS. But even those changes are not accepted and they are far more likely to be of general use than a Mac OSX service manage script. (I'm not arguing that they should have been, just using it as an example.)
If a vendor make changes to the functionality that improve upon things in Mailman, they yes, those changes would be great to see merged back in and it's good to chide those that don't do so. I am not aware that those are the sort of changes anyone has said Apple's made.
ISTM that often when a vendor makes such changes, they make them without the flexibility that would be needed to get them into the main code base. For example, cPanel has patched mailman to allow for lists of the same name and a different domain name within a single mailman install. But they have done it in a rather hacky and inelegant way so it's doubtful that even if they sent a patch to the developers list that it'd get accepted. It's only partially baked. :)
To me it makes a difference what sort of modifications you're talking about. Most of the modifications in this thread and similar threads deal with changes a vendor or distro make to integrate mailman into their specific way of doing things. And I think those changes aren't likely to be useful to the project as a whole. So that's the part where I think it's unfair to criticize vendors and distros over.
Obviously, those changes do impact the ability of folks here to offer help to those who ask for it. It is a good thing to remind those posters that they should check in with their vendor/distro to ensure that the problems they're having aren't vendor/distro specific. This is somewhat akin to the need to ask a poster where they installed mailman before anyone here can say "run this" or "edit that."
I certainly won't argue that you've got me handily beat in terms of experience and breadth of exposure to various systems. :-)
My point mainly is that it sometimes comes across that vendors/distros who install mailman and add to or change it to integrate it into their system gets painted as doing something wrong. I don't think that's the intent, but I felt it worth mentioning and pointing to the Apple code. For anyone interested in checking specifically what Apple has changed about the Mailman code, it's there.
Ideally a few Apple users would hang out on this list and could help to guide those who are asking for assistance and using the Apple installed Mailman. In that way, we'd all learn a little more about how to solve people's problems with Mailman.
-- Todd OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp
The limits of tyrants are prescribed by the endurance of those whom they oppress. -- Frederick Douglass
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 10:20 PM -0500 12/14/06, Todd Zullinger wrote:
But what value would MacOSX specific integrations be to the Mailman project?
Well, if they fed their changes back to us, that would allow us to incorporate that into future versions of the software, which would then be trivially easy for the vendor to upgrade to. Apple would no longer have to maintain their own proprietary modified version.
So, yes -- I think that there is a very strong benefit to the vendor, if they contribute their code back to the project.
I think there is also a very strong benefit to the user community (and the vendor), because that would allow us to better support the software on their boxes.
If there are things that need to be done to get Mailman to fit better into the FHS, and they aren't just the minimum changes hacked in but are appropriately and fully integrated into the software, I see no reason who those contributions would not be warmly welcomed.
But yes, if you just hack crap in, that's not likely to be accepted. I'm not saying that this is what John did -- I have no knowledge of his changes, one way or the other. I'm just saying that, in general, hackish type changes are much less likely to be accepted than ones where the added functionality is fully and properly integrated into the code.
Unfortunately, a lot of the changes most people submit are of the hackish variety.
Right, the hackish changes I was talking about.
Actually, I think it's just as valuable for a vendor to contribute those changes back to us as ones that actually have to do with core functionality.
This way, when someone comes to us with a problem on platform Y, we can look into the code and see where platform Y puts things, and we can give them a straight answer as to where the log files are, where the commands are, etc.... Otherwise, our only option is to say that the Mailman-standard location for a specific file is mumblefrotz, and that it's up to the questioner to figure out what this means for their platform.
When they do that without contributing back to us whatever changes they've made to the Mailman code or scripts, the locations of files, etc... then I think that they have done something wrong. Some could claim that they are in violation of the GNU copyleft, but at the very least I think that they are guilty of having failed to give adequate support back to the community from whom they are taking a given system.
Or maybe Apple could actually provide some actual support to their customers who've paid real money to get MacOS X Server.
I think your suggestion is more likely, because there are a couple of people in this situation who have done at least some of what you have suggested. I just don't think that it's fair to put on their shoulders the whole sum total of providing support for all MacOS X Server users who need help with the Apple-provided version of Mailman and who cannot get Apple to give them the support that they deserve.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/f10bf4ea18a58c83f99e6d5a4b12e322.jpg?s=120&d=mm&r=g)
Brad Knowles wrote:
It seems like that'd get to be a lot of work maintaining vendor specific integrations.
So, yes -- I think that there is a very strong benefit to the vendor, if they contribute their code back to the project.
Code is one thing, vendor specific integration is another.
They're part of the design goals for MM3. I'm not sure what the status is for 2.2 and I wasn't able to find any hits for FHS on wiki.list.org, but I may very well be doing something completely wrong there. :)
I don't think John's changes were of the hacked in variety, but the changes are enough that they haven't been integrated into the current 2.1 branch, though the patch was submitted in 10/2004. I think it's simply a case where those changes were very important to Red Hat (in that FHS compliance helped with other distro issues like integrating SELinux access controls) but they weren't important enough to the broader community of Mailman contributors to warrant making the changes in the 2.1 branch.
Nothing wrong with that. It's why open source is so nice. I have the choice to find vendors and projects that share more of my values than others may. If FHS compliance is really important to you, you'll like the Red Hat mailman packages. If you prefer simplicity and less scattering of your files, the stock mailman install will suit you better.
John's message and patch are in the mailman-developers archives here: http://www.mail-archive.com/mailman-developers@python.org/msg08110.html
Unfortunately, a lot of the changes most people submit are of the hackish variety.
That sounds like a good reason not to ache for too many more to me, especially if they're just vendor specific tweaks. ;)
Well, many of the changes that are made to the way mailman is installed aren't really in a format that a vendor or a distro can send to the Mailman project as a patch. These changes may be made in an rpm specfile or a debian control file or countless other distro and vendor specific methods. Being inundated with all of that doesn't help to support users better, IMO.
What helps is having people on the lists that know something about the various distro and vendor packages that can point out where something is to a new poster. Having used Red Hat systems for a number of years, I'll sometimes chime in when someone posts and says they have mailman-2.1.5-35 on Fedora or something like that, recognizing that they've got mailman installed via rpm. To help someone find out where mailman's parts are installed requires knowing a little about the package management system - unless mailman started shipping with a script in a standard path that you could point to and say, run "mailman_where_are_my_parts" or something. :)
That's a much broader view of the requirements of the GPL than I subscribe to. Changing install paths should be easy to do via configure. And if it's not and someone has to change the code to alter the paths, that's something that ought to be fixed in the mailman configure scripts (something akin to the FHS patch).
Changing the code in other ways is also allowed as long as those whom they ship the code to have the right to the source. That's the letter of the license as I understand it. And I agree with you that I don't much care for that as it does skirt the spirit of the license. I don't think that's the category that most vendor tweaks fall into though. They are likely in the former.
Well, yeah. That'd be even better. But since I don't give any money to Apple, I'm in no position to demand that they do anything for their customers. There's nothing wrong with making customers aware that they should expect support if they've paid for it, of course.
But if someone posts here and others can help them out, all the better as the knowledge goes into the list archives and FAQ.
I can't speak for Apple, but my experience using Red Hat and Fedora has been very good. I've found the community forums quite helpful at solving many problems. The users that get help in that way often never end up here, so we don't see how helpful the vendor/distro channel may be.
-- Todd OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp
The worth of a state, in the long run, is the worth of the individuals composing it. -- John Stuart Mill
![](https://secure.gravatar.com/avatar/334b870d5b26878a79b2dc4cfcc500bc.jpg?s=120&d=mm&r=g)
Todd Zullinger writes:
The problem is that most of the people who get referred to the "not our job, mon" FAQs chose Red Hat or Apple not for FHS compliance, but for turnkey installation convenience and system integration. If that fails, it's the vendor's responsibility, and the vendor is in the best position to answer questions.
I think it's important to insist on that point.
![](https://secure.gravatar.com/avatar/f10bf4ea18a58c83f99e6d5a4b12e322.jpg?s=120&d=mm&r=g)
Stephen J. Turnbull wrote:
True. Many times those folks are in over their head trying to configure an MTA, web server, and tuning mailman. They'd probably be better off reading up on their vendors package management system so that when someone here asked where the init script was, they could find it. :)
But to what extent? That's the point I was trying to get at in my original reply. The extreme is that there should be a FAQ and maybe even a section in the list welcome that says, "Use anything but mailman built from source? Don't ask questions here, ask your vendor." I don't believe that anyone here is in favor of that.
I agree that if someone comes here with questions that are obviously very dependent on some customization that their vendor has made that they should be directed to check with the vendor. (Same goes for users who need more basic help learning to use their OS of choice.) But many, if not most, questions from users who are using a Red Hat or even Apple package don't fall into that category, IMO.
-- Todd OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp
I got stopped by a cop the other day. He said, "Why'd you run that stop sign?" I said, "Because I don't believe everything I read." -- Stephen Wright
![](https://secure.gravatar.com/avatar/334b870d5b26878a79b2dc4cfcc500bc.jpg?s=120&d=mm&r=g)
Todd Zullinger writes:
I agree. But nobody's suggesting a "don't post here if you're using Red Hat" policy. The policy is that for installation problems, problems soon after installation, and missing functionality, ask the vendor *first*, unless you're sufficiently expert that you already know what a standard installation from source looks like and can explain the differences.
Note that without that expertise, you can't figure out what's "obviously very dependent." I think that if the user doesn't know, the default should be to go to the vendor, who knows more about installation issues and who has an interest in bullet-proofing their package.
![](https://secure.gravatar.com/avatar/746f7519ba02fb0d815e59f305c53fa2.jpg?s=120&d=mm&r=g)
Pierre Igot wrote:
I believe you did check the FAQ, and further, I don't believe you did anything wrong. If you feel we are picking on you, I apologize for my part in that. We're really only picking on Apple.
I actually didn't mean to send the partial reply at <http://mail.python.org/pipermail/mailman-users/2006-December/055006.html>. There was something in your reply at <http://mail.python.org/pipermail/mailman-users/2006-December/055011.html> that annoyed me. I still can't put my finger on exactly what, but I don't think my annoyance was warranted. Maybe I'm annoyed at Apple. I started to reply to that post three times, and each time, I thought better of it and canceled the message except the third time I accidently hit send instead of cancel.
Then having sent that obviously incomplete reply, I followed up with <http://mail.python.org/pipermail/mailman-users/2006-December/055007.html>
So that's my story, and again, I don't think you did anything wrong.
-- Mark Sapiro <msapiro@value.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 10:59 PM -0400 12/13/06, Pierre Igot wrote:
Okay, I split the MacOS X and MacOS X Server stuff into two separate FAQ entries. 1.21 now talks about using the Mailman-standard version of the software, while 1.29 now talks about the Apple-provided version of the software.
Please see <http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.021.htp> and <http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.029.htp>, respectively.
If there are any further clarifications or other things you'd like to add to these subjects, please feel free to edit the FAQ Wizard entry itself, and/or discuss that subject on this list.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/30ddf4e6b418c22b78fbe3ecdc0ec256.jpg?s=120&d=mm&r=g)
On 06-12-14, at 22:40, Brad Knowles wrote:
IMPORTANT: Please note that the correct spelling for Mac OS X and Mac
OS X Server is with a space between the "Mac" and the "OS".
I think part of the problem here stems from the very name of this
list. It's called "mailman-users." I am a "mailman user." So I feel
that this is the right list for me to get help about using Mailman.
There is nothing in the list description that warns users of non-
standard Mailman distributions away from the list. On the contrary,
this description puts the emphasis on users vs. developers. So
naturally "mailman users" such as myself feel that it's the natural
place to go and ask questions:
This mailing list is for users and other parties interested in the
Mailman mailing list management system. There is a separate mailing
list for people interested in discussion about development of the
system - mailman-developers.
(http://mail.python.org/mailman/listinfo/mailman-users]
That said, I am in full agreement that the FAQ needs to be read
first, and your updates are important--as long as you spell "Mac OS
X" properly. Otherwise, Mac OS X users might not even find the
relevant section :).
Pierre
-- Pierre Igot, administrateur des systèmes / Systems Administrator www.cprp.ca 902-837-7391
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 8:53 AM -0400 12/15/06, Pierre Igot wrote:
IMPORTANT: Please note that the correct spelling for Mac OS X and Mac OS X Server is with a space between the "Mac" and the "OS".
Those of us who have a long history with Mac OS (like, back to the days of the single floppy disk Mac Plus) will tend to have problems with this particular spelling issue.
That said, I will make an effort to find and correct all entries where this is misspelled. I probably wrote most of them, so I have the responsibility to make sure that they are correct.
There's two places where the mailing lists are described. One place is on the web pages at list.org (and mirrored at gnu.org), and one is within the Mailman listinfo description itself. Only Barry can change the description at list.org/gnu.org, but I will go in and update the listinfo description.
Not a problem.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 8:40 PM -0600 12/15/06, Brad Knowles wrote:
Found and corrected, in a surprising number of places.
I corrected the listinfo description at <http://mail.python.org/mailman/listinfo/mailman-users>. Please take a look and see if you have any further suggestions.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/01aa7d6d4db83982a2f6dd363d0ee0f3.jpg?s=120&d=mm&r=g)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Dec 15, 2006, at 9:40 PM, Brad Knowles wrote:
What kind of changes are you looking for? Do you have specific text
you'd like to change the online pages to use?
- -Barry
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin)
iQCVAwUBRYN6D3EjvBPtnXfVAQKsawP/by1uXjjSlijUTa3kAtHpzOeTL24WPfGJ mA4sSsmZVGsudSZu+0ElYVi9EhG7XjHVy97IZzdYww0rA8fTcY+MPhOM+XUraSof 0zLcRXQLY3CGLmGJMe0m+ruV/YOSxJf2WHWCl68Eh8u8GOynwxiypsCiGe2ft/C3 KF1/Q5ENzXU= =W4LZ -----END PGP SIGNATURE-----
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 11:46 PM -0500 12/15/06, Barry Warsaw wrote:
Well, I updated the listinfo page for mailman-users in three places. The first paragraph now reads:
This mailing list is for users and other parties interested
in the Mailman mailing list management system, as provided
for download via the resources shown at
http://www.list.org/download.html. There is a separate
mailing list for people interested in discussion about
development of the system - mailman-developers.
In particular, I added the reference to which version of the Mailman code we're talking about. I then added another paragraph, which reads:
Unofficially, we will try to provide what support we can
for versions of Mailman that come from other sources
(e.g., vendor-provided binary packages, vendor-provided
pre-installed software, etc...), but there will be a
limit to the level of support that we can provide for
versions of Mailman other than those which were built
directly from code downloaded directly from the
resources shown at http://www.list.org/download.html.
The idea being to make it clear to people that we will do as much as we can to support Mailman regardless of where they got the software and what modifications may have been made to it by the vendor, but to also recognize that there will be limitations to the level of support that we can provide for these versions. Oh, and to reinforce which version of Mailman we're talking about being the one that we officially support here.
Finally, I added another paragraph, which reads:
PLEASE MAKE SURE THAT YOU HAVE READ ALL THE RELEVANT
FAQ ENTRIES, ARCHIVE MESSAGES, ETC... BEFORE POSTING
ANY QUESTIONS TO THE LIST -- ESPECIALLY IF YOU HAVE
ANY QUESTIONS REGARDING A VERSION OF MAILMAN THAT
COMES FROM AN ALTERNATIVE SOURCE.
And yes, I went out of my way to make sure that this paragraph was in ALL CAPS. For the benefit of the whole Mailman community, I think it's really important to remind people that they are responsible for reading all the FAQ entries, archive messages, etc... that might be relevant to their question. This is doubly true for versions of Mailman that come from other sources.
Now, how much you might want to take from this and incorporate into the pages at list.org/gnu.org, I would say that would be up to you. If there are any other changes you'd like made to the listinfo pages, you're welcome to let me know and I'll be glad to make those for you -- if you like.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/b273ab068bc220d17a3e4c710c401c4b.jpg?s=120&d=mm&r=g)
In a flurry of recycled electrons, Brad Knowles wrote:
Brad- I like those descriptions, although I might further clarify that "uesrs" doesn't mean people sending mail through mailman, it means people installing/managing/troubleshooting mailman.
Also, could we use a short faq page that lists some common search terms and their local format/style? This would cover things like the space in "MAC OS".
z!
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 9:38 AM -0800 12/18/06, Carl Zwanzig wrote:
Good point. I'll make that change.
Also a good idea. Do you want to get something started, and then let others fill in/correct as necessary?
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/01aa7d6d4db83982a2f6dd363d0ee0f3.jpg?s=120&d=mm&r=g)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Dec 16, 2006, at 1:05 PM, Brad Knowles wrote:
Brad, sorry for the delay in responding. Thanks for all the updates
here. As far as what I'll add to the static web pages, well,
probably not much. I really want to move more content into the wiki
and less in the static pages, so as that happens I'll definitely keep
this info in mind.
Thanks!
- -Barry
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin)
iQCVAwUBRZ7Ee3EjvBPtnXfVAQJjjwP/VI7BIUbFwqbf8AyRUyXc20vnEJJUc1kB 1oHBPIrJhjr2AuyNpxQpD4RwwIWdd7ENNnzIqv1/3cP04wXZOSufevfKFZhPDMOd 5GNkGA5VZ58xAcyMkVlvd7n7jThKSHsSXwPdBSa4M3kj/z2o2F37tdEqDXzCRE5g eAmJC44xszo= =O0k8 -----END PGP SIGNATURE-----
![](https://secure.gravatar.com/avatar/746f7519ba02fb0d815e59f305c53fa2.jpg?s=120&d=mm&r=g)
Pierre Igot wrote:
See <http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.021.htp>
Yes, it worked. The hangup was probably because of trying to send notifications.
Yes, the problem was these locks.
So the mailmanctl daemon and the qrunners aren't running unless you start the qrunners manually as above, but this isn't the way to do it.
If you had done
/usr/share/mailman/bin/mailmanctl start
you wouldn't have had to kill the qrunners.
I think in Mac OSX server, Mailman is installed as a service and the service runs mailmanctl. Somehow you have stopped the service and you need to restart it.
I doubt that removing the locks had anything to do with it, but you might check ALL of Mailman's log files to see what kinds of errors and other conditions occurred.
-- Mark Sapiro <msapiro@value.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
![](https://secure.gravatar.com/avatar/30ddf4e6b418c22b78fbe3ecdc0ec256.jpg?s=120&d=mm&r=g)
On 06-12-13, at 19:31, Mark Sapiro wrote:
I am aware of this page and checked it before sending my request for
help. The page didn't provide any information that addressed my
issue, as far as I could tell.
Right--but I didn't see any error messages about anything. (I checked
all the logs.) So it was a failure that was not exactly graceful :).
OK, so based on what you are saying, deleting these locks doesn't
cause any further problems, right?
Well, there was nothing anywhere that indicated that I had to run
this command :).
Like I said, before the mass subscriptions, everything ran fine and I
never had to run this command. If "mailmanctl" somehow stopped, it
certainly didn't notify me. And there was nothing in the user
interface for Mailman that indicated that this was the problem.
Running this command did indeed fix the problem. The messages are now
sent out automatically. Thanks!
Well, I did not stop the service intentionally. The Mailman service
is part of the "Mail" service (which also controls POP and SMTP) and
there is nothing in the server admin software that indicates that
mailmanctl is not running. Probably a shortcoming of Apple's server
admin software.
In any case, now I know that if things are not going out, I need to
check if this "mailmanctl" is running. Thanks.
It's the kind of thing that is probably obvious to most readers of
this list, but I had to subscribe to the list and ask the question to
find out myself (or spend many hours reading the entire Mailman
documentation, I suppose).
OK. I did check all the log files, and didn't see any significant
errors.
Thanks again.
Pierre
Pierre Igot, administrateur des systèmes / Systems Administrator www.cprp.ca 902-837-7391
![](https://secure.gravatar.com/avatar/746f7519ba02fb0d815e59f305c53fa2.jpg?s=120&d=mm&r=g)
Pierre Igot wrote:
The relevant part of that FAQ is the part that says
All other support issues relating to using the Apple-provided versions of Mailman under MacOS X should be first directed to Apple, since they are the ones that modified the Mailman installation to suit their environment and they have not shared their changes with us.
There are other FAQ articles
Mark Sapiro <msapiro@value.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
![](https://secure.gravatar.com/avatar/746f7519ba02fb0d815e59f305c53fa2.jpg?s=120&d=mm&r=g)
Mark Sapiro wrote:
Sorry, I sent that before I finished it. I meant to say
There are other FAQ articles such as <http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq03.014.htp> that may have been helpful, but we don't know if they are applicable to the Mac OS-X Server Mailman or not because it is a modified version of our package, and we don't know what the mods are.
WRT to mailmanctl, apparently the Mailman qrunners were running at some time and then stopped. I don't know how they were started in the first place or why they stopped. Perhaps a reboot of the server would have started them again, perhaps not, but IMO, the fact that all this is not documented for you is Apple's failure, not ours.
I think I can understand your frustration, and I am not trying to be unsympathetic. If you had installed our package, and found our installation manual lacking, we would do our best to supplement it or improve it. However, we really can't create documentation for packages whose details we have even less information about than you do. </rant>
-- Mark Sapiro <msapiro@value.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
![](https://secure.gravatar.com/avatar/d5a34df919f6a79644d998b34dcee064.jpg?s=120&d=mm&r=g)
http://www.list.org/mailman-install/node50.html
I'm not sure just how much this may help the original poster or not.
--
Bye for now, Terry Allen
___________________________________________________________________
hEARd
Postal Address: hEARd, 26B Glenning Rd, Glenning Valley, NSW 2261, Australia Internet - WWW: http://heard.com.au http://itavservices.com EMAIL: hmag@ozemail.com.au Phone: Australia - 02 4388 1400 / International - + 61 2 43881400 Mobile: Australia - 04 28881400 / International - 61 4 28881400
Non profit promotion for new music - since 1994
![](https://secure.gravatar.com/avatar/01aa7d6d4db83982a2f6dd363d0ee0f3.jpg?s=120&d=mm&r=g)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Dec 14, 2006, at 3:18 AM, Terry Allen wrote:
I sent some docs through to Barry Warsaw a long time back, but some oif the info contained in it are here:
Hi Terry, would you consider fleshing out OS X support on the Mailman
wiki? That's a much better place to maintain this type of
information than in the docs.
- -Barry
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin)
iQCVAwUBRYFqOnEjvBPtnXfVAQKr6AQAs8GAEs/LkA4+iPA1wb73ZkynjlLosX9Y dg6qSSNynLgSLffy0xB/e5VXCwx/B+q9iA8ZKfzpkRbjTywpDb0t04h8V0KjvYQz De0In3pS4MMj0OAHpINwPBUS48MQ4lMvAj4kyuwKW4Gb8EsmbKSGil8TTlN9kb// MZITo2oHEic= =FwrF -----END PGP SIGNATURE-----
![](https://secure.gravatar.com/avatar/30ddf4e6b418c22b78fbe3ecdc0ec256.jpg?s=120&d=mm&r=g)
On 06-12-14, at 03:51, Mark Sapiro wrote:
Right, I understand your point of view perfectly.
On the other hand, the reality is that I, as a user, have to deal
with "Apple's failure," as you call it, and that the best avenue for
dealing with it, as far as I could tell, was to ask for help on this
list. The fact that you were able to help me so quickly and with a
single instruction proves my point :).
I *could* have submitted a bug report to Apple and waited for weeks
or months until I got a reply.
I *could* have spent hours searching through Apple's "Discussions"
forum on Mac OS X Server's Mail service, trying to find posts with
the exact same issue, knowing full well that I barely had enough
expertise to figure out the best keywords to use to describe the
problem. ("messages," "stuck," "in folder" are not exactly very
effective keywords, in that they are very generic)
I *could* have posted a request for help on the forum and waited for
days to see if, by chance, anyone might be able to figure out what
was going on on my machine.
I *could* have spent hours reading the Mailman documentation myself
until I had figured out the fundamentals of Mailman myself.
But the reality was that, in all likelihood, this was a pretty basic
problem that had to do with Mailman itself, that it probably had
nothing to do with Apple's customizations, and that the most
efficient way to get help was probably to submit a request to this list.
And, as I said, the quick (if reluctant) response you gave me proved
my point.
The question is whether it really costs Mailman experts such as
yourself so much effort to just provide such basic help from time to
time to "newbies" like me using a possibly somewhat non-standard
version of Mailman. (Based on what I have read, I don't think there
is much difference between the Mailman 2.1.2 installed with Mac OS X
Server 10.3 and your packages. It's just that everything is already
included and installed, and can be turned on through the Server Admin
GUI. And some file/folder locations are somewhat different from the
standard Unix locations. But all the rest is probably the exact same.)
For what it's worth, if you ever need any basic help with some Mac-
specific Mac OS X stuff, which is my expertise, I will be more than
glad to help, even if you are not using the exact version of Mac OS X
that I am currently working with :).
Thanks again.
Pierre
-- Pierre Igot, administrateur des systèmes / Systems Administrator www.cprp.ca 902-837-7391
![](https://secure.gravatar.com/avatar/b273ab068bc220d17a3e4c710c401c4b.jpg?s=120&d=mm&r=g)
[not speaking for Mark, Barry, or really anyone but me]
This is really the classic IT problem of unsupported configurations.
In a flurry of recycled electrons, Pierre Igot wrote:
The problem that comes up far too often is that someone, such as Mark, says "what's in the vette log?" or "delete /usr/local/mailman/lists/fred.lock" and the questioner says "where is it?" or "But I don't -have- a /usr/local/mailman" directory". We then start playing 20 questions- do you have root access? which version is installed? which pakcaged version is installed? etc. (iirc, all if which is in faq 3.14)
Given that mailman and the support are free, I think it's a bit much to expect that the maintainers know exactly how each distributor builds their package (directories, options, etc). They can only say "on a stock system, it works like this", and sometimes "there's a vette log somewhere, please find it".
z! former IT manager
![](https://secure.gravatar.com/avatar/30ddf4e6b418c22b78fbe3ecdc0ec256.jpg?s=120&d=mm&r=g)
On 06-12-14, at 13:18, Carl Zwanzig wrote:
Believe it or not, I did check the FAQ before submitting my request.
Try searching for all keywords "message stuck qfiles/in" or simply
for "qfiles" in the FAQ, for example. It doesn't return anything that
would have helped in my case.
OK, but that didn't happen in my case as far as I know. My problem
did not require any references to Mac OS X-specific things. I only
provided the Mac-specific information as a courtesy, because it was
part of list etiquette.
Mark provided me with a one-line Unix command that happened to work
with my version of Mailman in Mac OS X, but if it hadn't worked, I
would have simply tried to find where "mailmanctl" was on my system,
without bothering the list with such a detail. As it happens, I
didn't have to do this, but I would have done it if required. So I
don't feel that there was anything in my case that forced anyone to
deal with non-standard issues.
And that's all that I was asking for. I didn't ask for anything
beyond this.
Pierre
-- Pierre Igot, administrateur des systèmes / Systems Administrator www.cprp.ca 902-837-7391
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 2:03 PM -0400 12/14/06, Pierre Igot wrote:
Believe it or not, I did check the FAQ before submitting my request.
I do believe it.
The problem is that we cannot know beforehand all the possible search terms that people might use to describe any given particular type of problem. We can guess, and we can try to capture what we think are the most likely types of descriptions, but that's always a guess and there will always be someone who doesn't make the same guess and instead searches for something else. We can get better at covering more potential different descriptions, but we cannot ever be perfect.
Actually, there is an Apple-specific issue here. Apple has their own way of starting and stopping the Mailman queue runners and associated processes. If your queue runners aren't running, then you're going to have the kinds of problems that you saw -- regardless of what method is used to stop or start them.
If you're using the Apple-provided code to manage Mailman, you're not going to know anything about the Mailman queue runners. All you're going to know is that something is broken.
So, we have to work backwards from the "it's broke" state to figure out what is broken and who is responsible for that code. Now that we know your problem was a result of the queue runners not actually running, we can tell you that you need to start the queue runners, but since you're using MacOS X Server we can't tell you precisely what the "correct" way is to start them. We can tell you what the Mailman-standard was is to start them, but Apple has created their own code to manage this aspect of Mailman operations and they haven't shared that with us.
Which leads us to the general issue of people expecting us to provide support for a system that someone else has taken and modified, but without sharing any of those changes with us.
Even if the core Mailman system itself is unchanged from what we shipped, they have changed the interface that is used to manage Mailman, but none of the people on this list know anything about it.
Hell, the guy who runs lists.apple.com doesn't know anything about it, and he's running Mailman on machines at Apple that are running MacOS X Server, for Apple users on Apple-specific products -- but even he's not using the Apple-provided version of Mailman.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/f10bf4ea18a58c83f99e6d5a4b12e322.jpg?s=120&d=mm&r=g)
Brad Knowles wrote:
Apple does provide the source code for their mailman packages. You can browse it here:
http://www.opensource.apple.com/darwinsource/Current/mailman-117/
And download a tarball of that stuff:
http://www.opensource.apple.com/darwinsource/tarballs/other/mailman-117.tar....
AFAICT, the -117 is an Apple internal number. Browsing the full list shows similar, non-upstream numbers for many other packages.
The full list of open source code for Mac OS X 10.4.8 for PPC is at:
http://www.opensource.apple.com/darwinsource/Current/
So, while it seems fair to say that many of the very knowledgeable folks here don't know the specifics of the Apple packaged mailman, it doesn't seem quite as fair to say that Apple hasn't shared their changes. They are available for anyone to download and check out.
Compare that to cPanel where the best you find if you're not a customer is a half-backed and outdated patch (which I was provided only after persistent requests).
Disclaimer: I don't know anything about the Apple packages and the only Apple product I own is an iPod. :)
-- Todd OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp
Don't take life seriously, you'll never get out alive.
![](https://secure.gravatar.com/avatar/334b870d5b26878a79b2dc4cfcc500bc.jpg?s=120&d=mm&r=g)
Todd Zullinger writes:
Brad Knowles wrote:
We're not talking about "shared" as you will find it defined on webster.com, this is "shared" as used by Richard Stallman.[1] "Sharing" is active contribution to the mainline, it's not merely posting your code for others to pick up and analyze.
This usage makes some sense, because the creators of a forked product have already done most of that analysis. That doesn't mean they're not liable to substantial additional work, of course, and open source doesn't oblige them to do it. But by the same token the Mailman developers are forced (by lack of resources to do such analysis themselves) to take the position that this code hasn't been contributed to the project, so they can't know much about it.
It's not "fair", but that's the customary usage of the verb "to share" in the parts of this community that I frequent.
Sure, Apple is doing the open source thing by the letter of the law, and that is no small thing. But the problem presented to the maintainers of the mainline by Apple's fork is of the same variety as the problem presented by cPanel.
Footnotes: [1] No, I don't approve of this usage, that's why I attribute it to a famous extremist. "Extremism in the defense of freedom is no vice."<wink>
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 5:32 PM -0500 12/14/06, Todd Zullinger wrote:
Apple does provide the source code for their mailman packages. You can browse it here:
Even if they have made the source code available for everything they've done with regards to Mailman (which includes all their proprietary management tools), this is not the same thing as contributing that code back to the Mailman project.
Likewise, webmin also includes a component for managing Mailman, and all their code is publicly available, but we don't say that webmin has contributed all their code back to the Mailman project.
Same any other group that takes our code and makes modifications to it -- even if they make all their source code publicly available, that's not the same thing as contributing it back to the Mailman project.
That is worse, yes. But that's also not the point.
Disclaimer: I don't know anything about the Apple packages and the only Apple product I own is an iPod. :)
Whereas I was an Apple fan from 1982, back in the original Apple ][ days, and before the introduction of the 16K language card, the Apple II+, the Apple IIe, the Apple //c, or any other more recent Apple product. I've been a MacFanatic since December of 1983 when I saw an early prototype behind closed doors. I've been a Mac owner since the day when I bought a Mac SE -- before the internal hard drive model was available.
I've got a long history with the company, but I'm not blind to things that they do poorly. At least, not totally blind.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/f10bf4ea18a58c83f99e6d5a4b12e322.jpg?s=120&d=mm&r=g)
Brad Knowles wrote:
But what value would MacOSX specific integrations be to the Mailman project? I think that they'd likely be of limited value. Just as Debian or Red Hat specific changes are. I know that John Dennis of Red Hat contributed the changes Red Hat made to make mailman fit in better with the FHS. But even those changes are not accepted and they are far more likely to be of general use than a Mac OSX service manage script. (I'm not arguing that they should have been, just using it as an example.)
If a vendor make changes to the functionality that improve upon things in Mailman, they yes, those changes would be great to see merged back in and it's good to chide those that don't do so. I am not aware that those are the sort of changes anyone has said Apple's made.
ISTM that often when a vendor makes such changes, they make them without the flexibility that would be needed to get them into the main code base. For example, cPanel has patched mailman to allow for lists of the same name and a different domain name within a single mailman install. But they have done it in a rather hacky and inelegant way so it's doubtful that even if they sent a patch to the developers list that it'd get accepted. It's only partially baked. :)
To me it makes a difference what sort of modifications you're talking about. Most of the modifications in this thread and similar threads deal with changes a vendor or distro make to integrate mailman into their specific way of doing things. And I think those changes aren't likely to be useful to the project as a whole. So that's the part where I think it's unfair to criticize vendors and distros over.
Obviously, those changes do impact the ability of folks here to offer help to those who ask for it. It is a good thing to remind those posters that they should check in with their vendor/distro to ensure that the problems they're having aren't vendor/distro specific. This is somewhat akin to the need to ask a poster where they installed mailman before anyone here can say "run this" or "edit that."
I certainly won't argue that you've got me handily beat in terms of experience and breadth of exposure to various systems. :-)
My point mainly is that it sometimes comes across that vendors/distros who install mailman and add to or change it to integrate it into their system gets painted as doing something wrong. I don't think that's the intent, but I felt it worth mentioning and pointing to the Apple code. For anyone interested in checking specifically what Apple has changed about the Mailman code, it's there.
Ideally a few Apple users would hang out on this list and could help to guide those who are asking for assistance and using the Apple installed Mailman. In that way, we'd all learn a little more about how to solve people's problems with Mailman.
-- Todd OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp
The limits of tyrants are prescribed by the endurance of those whom they oppress. -- Frederick Douglass
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 10:20 PM -0500 12/14/06, Todd Zullinger wrote:
But what value would MacOSX specific integrations be to the Mailman project?
Well, if they fed their changes back to us, that would allow us to incorporate that into future versions of the software, which would then be trivially easy for the vendor to upgrade to. Apple would no longer have to maintain their own proprietary modified version.
So, yes -- I think that there is a very strong benefit to the vendor, if they contribute their code back to the project.
I think there is also a very strong benefit to the user community (and the vendor), because that would allow us to better support the software on their boxes.
If there are things that need to be done to get Mailman to fit better into the FHS, and they aren't just the minimum changes hacked in but are appropriately and fully integrated into the software, I see no reason who those contributions would not be warmly welcomed.
But yes, if you just hack crap in, that's not likely to be accepted. I'm not saying that this is what John did -- I have no knowledge of his changes, one way or the other. I'm just saying that, in general, hackish type changes are much less likely to be accepted than ones where the added functionality is fully and properly integrated into the code.
Unfortunately, a lot of the changes most people submit are of the hackish variety.
Right, the hackish changes I was talking about.
Actually, I think it's just as valuable for a vendor to contribute those changes back to us as ones that actually have to do with core functionality.
This way, when someone comes to us with a problem on platform Y, we can look into the code and see where platform Y puts things, and we can give them a straight answer as to where the log files are, where the commands are, etc.... Otherwise, our only option is to say that the Mailman-standard location for a specific file is mumblefrotz, and that it's up to the questioner to figure out what this means for their platform.
When they do that without contributing back to us whatever changes they've made to the Mailman code or scripts, the locations of files, etc... then I think that they have done something wrong. Some could claim that they are in violation of the GNU copyleft, but at the very least I think that they are guilty of having failed to give adequate support back to the community from whom they are taking a given system.
Or maybe Apple could actually provide some actual support to their customers who've paid real money to get MacOS X Server.
I think your suggestion is more likely, because there are a couple of people in this situation who have done at least some of what you have suggested. I just don't think that it's fair to put on their shoulders the whole sum total of providing support for all MacOS X Server users who need help with the Apple-provided version of Mailman and who cannot get Apple to give them the support that they deserve.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/f10bf4ea18a58c83f99e6d5a4b12e322.jpg?s=120&d=mm&r=g)
Brad Knowles wrote:
It seems like that'd get to be a lot of work maintaining vendor specific integrations.
So, yes -- I think that there is a very strong benefit to the vendor, if they contribute their code back to the project.
Code is one thing, vendor specific integration is another.
They're part of the design goals for MM3. I'm not sure what the status is for 2.2 and I wasn't able to find any hits for FHS on wiki.list.org, but I may very well be doing something completely wrong there. :)
I don't think John's changes were of the hacked in variety, but the changes are enough that they haven't been integrated into the current 2.1 branch, though the patch was submitted in 10/2004. I think it's simply a case where those changes were very important to Red Hat (in that FHS compliance helped with other distro issues like integrating SELinux access controls) but they weren't important enough to the broader community of Mailman contributors to warrant making the changes in the 2.1 branch.
Nothing wrong with that. It's why open source is so nice. I have the choice to find vendors and projects that share more of my values than others may. If FHS compliance is really important to you, you'll like the Red Hat mailman packages. If you prefer simplicity and less scattering of your files, the stock mailman install will suit you better.
John's message and patch are in the mailman-developers archives here: http://www.mail-archive.com/mailman-developers@python.org/msg08110.html
Unfortunately, a lot of the changes most people submit are of the hackish variety.
That sounds like a good reason not to ache for too many more to me, especially if they're just vendor specific tweaks. ;)
Well, many of the changes that are made to the way mailman is installed aren't really in a format that a vendor or a distro can send to the Mailman project as a patch. These changes may be made in an rpm specfile or a debian control file or countless other distro and vendor specific methods. Being inundated with all of that doesn't help to support users better, IMO.
What helps is having people on the lists that know something about the various distro and vendor packages that can point out where something is to a new poster. Having used Red Hat systems for a number of years, I'll sometimes chime in when someone posts and says they have mailman-2.1.5-35 on Fedora or something like that, recognizing that they've got mailman installed via rpm. To help someone find out where mailman's parts are installed requires knowing a little about the package management system - unless mailman started shipping with a script in a standard path that you could point to and say, run "mailman_where_are_my_parts" or something. :)
That's a much broader view of the requirements of the GPL than I subscribe to. Changing install paths should be easy to do via configure. And if it's not and someone has to change the code to alter the paths, that's something that ought to be fixed in the mailman configure scripts (something akin to the FHS patch).
Changing the code in other ways is also allowed as long as those whom they ship the code to have the right to the source. That's the letter of the license as I understand it. And I agree with you that I don't much care for that as it does skirt the spirit of the license. I don't think that's the category that most vendor tweaks fall into though. They are likely in the former.
Well, yeah. That'd be even better. But since I don't give any money to Apple, I'm in no position to demand that they do anything for their customers. There's nothing wrong with making customers aware that they should expect support if they've paid for it, of course.
But if someone posts here and others can help them out, all the better as the knowledge goes into the list archives and FAQ.
I can't speak for Apple, but my experience using Red Hat and Fedora has been very good. I've found the community forums quite helpful at solving many problems. The users that get help in that way often never end up here, so we don't see how helpful the vendor/distro channel may be.
-- Todd OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp
The worth of a state, in the long run, is the worth of the individuals composing it. -- John Stuart Mill
![](https://secure.gravatar.com/avatar/334b870d5b26878a79b2dc4cfcc500bc.jpg?s=120&d=mm&r=g)
Todd Zullinger writes:
The problem is that most of the people who get referred to the "not our job, mon" FAQs chose Red Hat or Apple not for FHS compliance, but for turnkey installation convenience and system integration. If that fails, it's the vendor's responsibility, and the vendor is in the best position to answer questions.
I think it's important to insist on that point.
![](https://secure.gravatar.com/avatar/f10bf4ea18a58c83f99e6d5a4b12e322.jpg?s=120&d=mm&r=g)
Stephen J. Turnbull wrote:
True. Many times those folks are in over their head trying to configure an MTA, web server, and tuning mailman. They'd probably be better off reading up on their vendors package management system so that when someone here asked where the init script was, they could find it. :)
But to what extent? That's the point I was trying to get at in my original reply. The extreme is that there should be a FAQ and maybe even a section in the list welcome that says, "Use anything but mailman built from source? Don't ask questions here, ask your vendor." I don't believe that anyone here is in favor of that.
I agree that if someone comes here with questions that are obviously very dependent on some customization that their vendor has made that they should be directed to check with the vendor. (Same goes for users who need more basic help learning to use their OS of choice.) But many, if not most, questions from users who are using a Red Hat or even Apple package don't fall into that category, IMO.
-- Todd OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp
I got stopped by a cop the other day. He said, "Why'd you run that stop sign?" I said, "Because I don't believe everything I read." -- Stephen Wright
![](https://secure.gravatar.com/avatar/334b870d5b26878a79b2dc4cfcc500bc.jpg?s=120&d=mm&r=g)
Todd Zullinger writes:
I agree. But nobody's suggesting a "don't post here if you're using Red Hat" policy. The policy is that for installation problems, problems soon after installation, and missing functionality, ask the vendor *first*, unless you're sufficiently expert that you already know what a standard installation from source looks like and can explain the differences.
Note that without that expertise, you can't figure out what's "obviously very dependent." I think that if the user doesn't know, the default should be to go to the vendor, who knows more about installation issues and who has an interest in bullet-proofing their package.
![](https://secure.gravatar.com/avatar/746f7519ba02fb0d815e59f305c53fa2.jpg?s=120&d=mm&r=g)
Pierre Igot wrote:
I believe you did check the FAQ, and further, I don't believe you did anything wrong. If you feel we are picking on you, I apologize for my part in that. We're really only picking on Apple.
I actually didn't mean to send the partial reply at <http://mail.python.org/pipermail/mailman-users/2006-December/055006.html>. There was something in your reply at <http://mail.python.org/pipermail/mailman-users/2006-December/055011.html> that annoyed me. I still can't put my finger on exactly what, but I don't think my annoyance was warranted. Maybe I'm annoyed at Apple. I started to reply to that post three times, and each time, I thought better of it and canceled the message except the third time I accidently hit send instead of cancel.
Then having sent that obviously incomplete reply, I followed up with <http://mail.python.org/pipermail/mailman-users/2006-December/055007.html>
So that's my story, and again, I don't think you did anything wrong.
-- Mark Sapiro <msapiro@value.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 10:59 PM -0400 12/13/06, Pierre Igot wrote:
Okay, I split the MacOS X and MacOS X Server stuff into two separate FAQ entries. 1.21 now talks about using the Mailman-standard version of the software, while 1.29 now talks about the Apple-provided version of the software.
Please see <http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.021.htp> and <http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.029.htp>, respectively.
If there are any further clarifications or other things you'd like to add to these subjects, please feel free to edit the FAQ Wizard entry itself, and/or discuss that subject on this list.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/30ddf4e6b418c22b78fbe3ecdc0ec256.jpg?s=120&d=mm&r=g)
On 06-12-14, at 22:40, Brad Knowles wrote:
IMPORTANT: Please note that the correct spelling for Mac OS X and Mac
OS X Server is with a space between the "Mac" and the "OS".
I think part of the problem here stems from the very name of this
list. It's called "mailman-users." I am a "mailman user." So I feel
that this is the right list for me to get help about using Mailman.
There is nothing in the list description that warns users of non-
standard Mailman distributions away from the list. On the contrary,
this description puts the emphasis on users vs. developers. So
naturally "mailman users" such as myself feel that it's the natural
place to go and ask questions:
This mailing list is for users and other parties interested in the
Mailman mailing list management system. There is a separate mailing
list for people interested in discussion about development of the
system - mailman-developers.
(http://mail.python.org/mailman/listinfo/mailman-users]
That said, I am in full agreement that the FAQ needs to be read
first, and your updates are important--as long as you spell "Mac OS
X" properly. Otherwise, Mac OS X users might not even find the
relevant section :).
Pierre
-- Pierre Igot, administrateur des systèmes / Systems Administrator www.cprp.ca 902-837-7391
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 8:53 AM -0400 12/15/06, Pierre Igot wrote:
IMPORTANT: Please note that the correct spelling for Mac OS X and Mac OS X Server is with a space between the "Mac" and the "OS".
Those of us who have a long history with Mac OS (like, back to the days of the single floppy disk Mac Plus) will tend to have problems with this particular spelling issue.
That said, I will make an effort to find and correct all entries where this is misspelled. I probably wrote most of them, so I have the responsibility to make sure that they are correct.
There's two places where the mailing lists are described. One place is on the web pages at list.org (and mirrored at gnu.org), and one is within the Mailman listinfo description itself. Only Barry can change the description at list.org/gnu.org, but I will go in and update the listinfo description.
Not a problem.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 8:40 PM -0600 12/15/06, Brad Knowles wrote:
Found and corrected, in a surprising number of places.
I corrected the listinfo description at <http://mail.python.org/mailman/listinfo/mailman-users>. Please take a look and see if you have any further suggestions.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/01aa7d6d4db83982a2f6dd363d0ee0f3.jpg?s=120&d=mm&r=g)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Dec 15, 2006, at 9:40 PM, Brad Knowles wrote:
What kind of changes are you looking for? Do you have specific text
you'd like to change the online pages to use?
- -Barry
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin)
iQCVAwUBRYN6D3EjvBPtnXfVAQKsawP/by1uXjjSlijUTa3kAtHpzOeTL24WPfGJ mA4sSsmZVGsudSZu+0ElYVi9EhG7XjHVy97IZzdYww0rA8fTcY+MPhOM+XUraSof 0zLcRXQLY3CGLmGJMe0m+ruV/YOSxJf2WHWCl68Eh8u8GOynwxiypsCiGe2ft/C3 KF1/Q5ENzXU= =W4LZ -----END PGP SIGNATURE-----
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 11:46 PM -0500 12/15/06, Barry Warsaw wrote:
Well, I updated the listinfo page for mailman-users in three places. The first paragraph now reads:
This mailing list is for users and other parties interested
in the Mailman mailing list management system, as provided
for download via the resources shown at
http://www.list.org/download.html. There is a separate
mailing list for people interested in discussion about
development of the system - mailman-developers.
In particular, I added the reference to which version of the Mailman code we're talking about. I then added another paragraph, which reads:
Unofficially, we will try to provide what support we can
for versions of Mailman that come from other sources
(e.g., vendor-provided binary packages, vendor-provided
pre-installed software, etc...), but there will be a
limit to the level of support that we can provide for
versions of Mailman other than those which were built
directly from code downloaded directly from the
resources shown at http://www.list.org/download.html.
The idea being to make it clear to people that we will do as much as we can to support Mailman regardless of where they got the software and what modifications may have been made to it by the vendor, but to also recognize that there will be limitations to the level of support that we can provide for these versions. Oh, and to reinforce which version of Mailman we're talking about being the one that we officially support here.
Finally, I added another paragraph, which reads:
PLEASE MAKE SURE THAT YOU HAVE READ ALL THE RELEVANT
FAQ ENTRIES, ARCHIVE MESSAGES, ETC... BEFORE POSTING
ANY QUESTIONS TO THE LIST -- ESPECIALLY IF YOU HAVE
ANY QUESTIONS REGARDING A VERSION OF MAILMAN THAT
COMES FROM AN ALTERNATIVE SOURCE.
And yes, I went out of my way to make sure that this paragraph was in ALL CAPS. For the benefit of the whole Mailman community, I think it's really important to remind people that they are responsible for reading all the FAQ entries, archive messages, etc... that might be relevant to their question. This is doubly true for versions of Mailman that come from other sources.
Now, how much you might want to take from this and incorporate into the pages at list.org/gnu.org, I would say that would be up to you. If there are any other changes you'd like made to the listinfo pages, you're welcome to let me know and I'll be glad to make those for you -- if you like.
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/b273ab068bc220d17a3e4c710c401c4b.jpg?s=120&d=mm&r=g)
In a flurry of recycled electrons, Brad Knowles wrote:
Brad- I like those descriptions, although I might further clarify that "uesrs" doesn't mean people sending mail through mailman, it means people installing/managing/troubleshooting mailman.
Also, could we use a short faq page that lists some common search terms and their local format/style? This would cover things like the space in "MAC OS".
z!
![](https://secure.gravatar.com/avatar/7bdecdef03708b218939094eb05e8b35.jpg?s=120&d=mm&r=g)
At 9:38 AM -0800 12/18/06, Carl Zwanzig wrote:
Good point. I'll make that change.
Also a good idea. Do you want to get something started, and then let others fill in/correct as necessary?
-- Brad Knowles, <brad@shub-internet.org>
Trend Micro has announced that they will cancel the stop.mail-abuse.org mail forwarding service as of 15 November 2006. If you have an old e-mail account for me at this domain, please make sure you correct that with the current address.
![](https://secure.gravatar.com/avatar/01aa7d6d4db83982a2f6dd363d0ee0f3.jpg?s=120&d=mm&r=g)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Dec 16, 2006, at 1:05 PM, Brad Knowles wrote:
Brad, sorry for the delay in responding. Thanks for all the updates
here. As far as what I'll add to the static web pages, well,
probably not much. I really want to move more content into the wiki
and less in the static pages, so as that happens I'll definitely keep
this info in mind.
Thanks!
- -Barry
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin)
iQCVAwUBRZ7Ee3EjvBPtnXfVAQJjjwP/VI7BIUbFwqbf8AyRUyXc20vnEJJUc1kB 1oHBPIrJhjr2AuyNpxQpD4RwwIWdd7ENNnzIqv1/3cP04wXZOSufevfKFZhPDMOd 5GNkGA5VZ58xAcyMkVlvd7n7jThKSHsSXwPdBSa4M3kj/z2o2F37tdEqDXzCRE5g eAmJC44xszo= =O0k8 -----END PGP SIGNATURE-----
participants (8)
-
Barry Warsaw
-
Brad Knowles
-
cpz@tuunq.com
-
Mark Sapiro
-
Pierre Igot
-
Stephen J. Turnbull
-
Terry Allen
-
Todd Zullinger