Not To James B. Byrne

Home » CentOS » Not To James B. Byrne
CentOS 27 Comments

1) You sent that to my email, not the list.
2) I already have “Filter before Junk Classification” selected in that filter’s Getting New Mail picklist.
3) You should look in your Spam folder and see if there aren’t some emails with [CentOS] in their Subject lines. If it’s completely empty, possibly you’re having TB delete emails it thinks are junk.

27 thoughts on - Not To James B. Byrne

  • 1) I replied privately to reduce the list load since it was a TB config issue that I was addressing and not particularly the topic being discussed, where you and I are doing something similar and I was interested to know why my solution works and yours doesn’t. But oh well :)

    2) The pertinent part of the TB filter was the “set junk to not-junk”, but that will only work if the filtering is applied before TB junking occurs, which is why I mentioned it to confirm your settings.

    3) I do not have any mailing list messages deposited in my spam boxes and do not have any “/dev/null” redirects either in gmail or in TB (and never will. I’m a sysad, therefore the word paranoid cannot be applied
    >:D). I can say with certainty that none of my mailing list emails have wound up in any of the 3 spam boxes that they could land in. I have checked them all. As I was mentioning, I have filters set up on all the mailing lists that I care about to not spam/junk any messages on those lists. And those filters have been working reliably for some time now. Which is why I am curious to know what is different between your filtering and mine.

    Thanks!
    Miranda

  • Just guessing, but it may be that you are using POP to retrieve the mail and getting an “uncategorized” view of new messages in the inbox, where if you use IMAP (with the possibility of syncing to multiple systems), gmail’s labels are mapped to imap folders before you get them.

  • No, actually it’s more like you have to get out of a bus –
    And you ask a person at the door to move a bit. Then suddenly some passengers turn to you and shout:
    “You don’t even know how to drive a bus.” :-))

    Greetings from Germany Alex

  • You may be onto something, because I *am* using IMAP (TB’s default during account setup) instead of POP3.

    I’ll be looking around in gmail next to see if there’s some way to pre-sort mail from CentOS.org (as Miranda implied) before whatever mail app I’m using at the time fetches it. I have never seen an online email interface I liked, so I don’t spend much time in gmail’s.

    One of the things that drove me to linux (and I liked rpm/yum better than dpkg/apt) was microsoft disabling Windows Mail in Win7 to force people to use their online ‘Live’ email.

  • I’ve used an assortment of methods over the years including using fetchmail to drop in my own imap server host before gmail had imap itself. But, I’ve given up on thinking I can do it better than google does.

    I’ve grown to like gmail’s web interface – or at least tolerate the parts I don’t like to get the part I do – which is mostly the raw search capability that I use in favor of trying to categorize anything, plus the fact that it is always in perfect sync with my android phone where I read a lot but rarely type long replies. Try going into configuration settings to tune it more to your liking. You may want some of the ‘labs’ options. In particular I don’t think I
    would like it without the auto-advance after archiving/deleting (that is, go to the next message instead of redrawing the index list).

    Given that you can use thunderbird on windows, that doesn’t make much sense, but linux isn’t a bad desktop these days.

    The main thing that keeps me from using an email ‘application’ for personal email is that I use an assortment of different machines during the day and don’t want to be tied to one or deal with the differences in the way the apps would work across windows/linux/mac versions even if I went to the trouble to install it and configure imap sync everywhere. Plus, I really don’t want to sync the huge ‘all mail’ archived folder but I want searches to go there. And, if you haven’t used gmail/yahoo/hotmail (now outlook.com) for a while you might be surprised at how well the new web interfaces work and how much you can customize them.

  • Ah, but I also use it on CentOS… I just don’t post as much from that copy. The point of that is to have at least 2 offline sources to my list subscriptions, since if the problem is with the network, having them all available only online is useless (my third full year of this list offline is close to complete).

    I fully expect this laptop to die anytime, since I bought it in 2006
    (its current HDD is ~4-years old)… when that happens my next laptop/notebook will have CentOS installed on it (probably 7, even though I dislike GNOME3 about the same as vista compared to XP), and my backup copy will probably be on an Android tablet. I’ve previously bought 2 copies of RedHat so I *have* contributed $$ to the cause, but I
    don’t care much for the whole subscription model; Community support works fine for me (if I was running a business on it, I’m certain I’d feel differently about that).

  • Offline? What’s that and why does it matter? I’ve already forgotten everything I can look up on google, so the world stops without access. But there is hardly anywhere I could go where I can’t access gmail, at least through the phone.

    Besides, when you dig up that offline email copy you were looking for it will tell you that to solve your problem you need to do an update –
    which you won’t be able to do without your network anyway.

  • I primarily use pop (and imap and gmail interface when I need to), but since I read email with different focus when I’m at home or at work, popping the msgs has turned out to be the most efficient way for me to operate. I also have gmail filters set up. My CentOS filter is dead simple:

    Matches: to:(CentOS@CentOS.org)
    Do this: Skip Inbox, Mark as read, Apply label “Lists/CentOS”, Never send it to Spam

    So in the gmail interface and in the gmail imap setup in TB, I have a nice little cubbyhole that has all the CentOS emails in it.

    Then when my TB clients pop the messages, they filter with this one:

    name=”CentOS”
    enabled=”yes”
    type=”17″
    action=”Move to folder”
    actionValue=”mailbox://nobody@Local%20Folders/08%20Lists/CentOS/CentOS”
    action=”Stop execution”
    condition=”OR (all addresses,contains,CentOS@CentOS.org) OR (all addresses,contains,mailman-owner@CentOS.org)”

    Interestingly, the filter in my home TB lacks the JunkScore=0 action, but seems to work just as well as the one at work that includes it.

    Hope this helps.

    Miranda

  • I don’t use the gmail interface for day-to-day email processing, for precisely that reason. It is why I resort to TB. When I’m at work I read all email with a work-centric focus. When I’m at home I read all email with a not-so-work-centric focus (unless I’m working from home, as I am today). But all email gets pulled to both locations. If something kills my work pc, I have a copy at home and vice versa. Which is handy when my email goes back 15+ years and google won’t let me keep it all there without paying for it which I’d rather not do. For the older email, those TB clients are the only copies I have. Even though I have backups, I still do this because recovery has been very quick this way (just replace the dead profile with the good one).

    And of course, when the apocalypse comes and gmail goes away, I’m all prepared! [/joke]

  • I don’t get it. Why auto-mark read in the first place?

    I have a completely separate work account. With its own restrictions and retention policies. It hasn’t always been that way but it seems easier now (someone else manages that server).

    I have 100+GB of google-space without paying extra, I think partly as a side effect of the android phone I use. And I don’t think there is any time-related restriction.

    I used to pull copies to my own server with fetchmail, and later imap-synced with thunderbird (sometimes including the All Mail folder). But the computers that used to do that have all died of old age so I gave up on being more reliable than google. Besides, with the work stuff in a separate account it is almost exclusively list mail that could be found in public archives anyway.

  • Marking it as read removes it from my gmail inbox for the times when I
    ~do~ need to read email using the gmail interface. I could do that I suppose, but I haven’t and probably wouldn’t have the time necessary to separate out the emails between the two accounts. I
    already have 6+ email accounts that I have to monitor so I’d rather not fork off another if I can help it. It’s not the time, just the byte volume. I get ~15GB of space for free per account, I think. The vast majority of my email unfortunately is not publicly archived, so I don’t have that option.

  • Writing as a humble programmer, why don’t you and Les write your own database application (using HTML, CSS, PHP and MariaDB (MySQL)) and store the important parts (or wholes) of emails in the database ?

    I do this. I can search on ‘text’, database entry descriptions, 6
    keyword fields, entry date, overdue date etc. and can email out from within the database system which has menu lists of email addresses. I
    can have 1 million topics and each topic can have 99 items of separate correspondence. Each separate item can link to 9 web items or stored items (PDFs, ODT, pictures etc.) stored on the server.

    Data can be retrieved in less than 2 seconds. The inbuilt links produce lists of related items. The system links into other databases
    (Names/addresses/emails/telephone numbers, information storage etc. etc.)

    Why keep masses and masses of irrelevant data in an unstructured format presided over by Google? Its not logical sense. Essentially, why store a lot of “rubbish” that will never ever be needed ?

    Happy Weekend.

  • Email is inherently unstructured and searches are over some set of words that I happen to remember. So you really need a full text indexer which google happens to be very good at. And the storage is their problem… Actually thunderbird is very nice at this too if you do have your own copy – I don’t remember if you have to enable indexing or if it is the default now.

  • Why? Because keeping “masses of irrelevant data” takes none of my time and when I need to dig for something there are ways to do so that cost me little or no time. Storing the irrelevant data is cheap, my time isn’t and I have very little time to spend on something like email. I
    converged on a solution that works for me and lets me do my job and take care of my family in a relatively efficient way. Sure there are all sorts of things I could do to be “better” but I have neither the time or resources to devote to making those happen. So I live with what I have.

    Les, I believe TB does index by default. I recall seeing a setting for that someplace in the menus. Thunderbird’s search is pretty good, pretty much like gmail’s in fact.

    Miranda

  • there are really 2 differnet things here, 1) what is the mailman included in a specific CentOS vesion doing and 2) what is the lists.CentOS.org machine doing with DKIM and what is the larger fix for each of those things.

  • The point is that mailman has the fix. I suppose you can look at the question of whether you solve the problem only for yourself or for all CentOS users as two different things but the solution is pretty much the same as any other bug that has been fixed (far) upstream.

  • I think it’s important to note that this actually isn’t a bug. This is failure to strip DKIM headers when forwarding a message. Note that when RHEL6 was released DKIM was still new and DMARC was pretty much unheard of. It’s not surprising that the version of Mailman in it does not take steps to remove DKIM headers as it’s simply a feature that would not have existed when that version was released.

    It’s also important to note that these headers *can* be removed in postfix (and probably other MTAs can as well) after the messages are submitted by mailman, so while it would be nice for mailman to do it it’s not strictly necessary, we can deal with the problem with the versions of mailman and postfix that are running on the server already. The trick is to simply set header_checks to match and remove the DKIM
    header which is quite easy.

    At that point we can have the server sign the message with its own DKIM
    signature and apply any relevant DMARC policy we want.

    I guess what I’m saying is you don’t *need* a new version of mailman to deal with this, you don’t need a new version of any software really, it can be dealt with the software we already have on the server with just a few config changes.

    Peter

  • the other important feature the new mailman has is to munge the From:
    field if the user’s DNS has the DMARC records indicating a draconian policy.

  • Grrr, yes, of course, DMARC likes to check the From: header now (utterly stupid). Anyways, it should be possible to rewrite this with header_checks as well. I would tack @CentOS.org onto the end of the domain and call it a day.

    Peter

  • Happily I never used Exchange but I have suffered from misconfigured Exchange installations.

    Definite don’t want anything to do with Micro$oft. I take pride in the work I do. That is incompatible with the M$ philosophy.

  • Noise removed.

    Is it too much to ask for that this thread, if not the list as a whole, return to being CentOS specific?

    John