From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Schmidt Date: Thu, 14 Jan 2010 05:00:31 +0000 Subject: Re: 'Not metoo' feature Message-Id: <4B4EA4EF.8090303@yahoo.com.au> List-Id: References: <4B4DDEAC.5030103@yahoo.com.au> In-Reply-To: <4B4DDEAC.5030103@yahoo.com.au> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: mlmmj@mlmmj.org >> As I think through the things in Mailman that I used to use >> before deciding I'd like to switch, a feature I'd be interested >> in having is a 'not metoo' feature. This means when the manager >> sends out an email it omits the sender from the distribution list. > > Just checked, and I believe the changes for this need to be done in > mlmmj-send.c, in the functions send_mail_verp and send_mail_many_list, > both in the loop > > for(i = 0; i < addrs->count; i++) { > > => comparing "from" with addrs->strs[i] and just "continue" when these > are equal should be ok ... > > Of course then the new tunable needs to be checked etc ... Sounds like a good start. Ideally this would be selectable on a per-subscriber basis. Perhaps best would be a notmetoo version of the list (much like the nomail and digest versions), then subscribers could have their preferred delivery method by subscribing to the right version; a tunable could make the default +subscribe action a notmetoo subscription. Hmm. As I further my investigations into things that I have found useful but which mlmmj doesn't have, I discover another, which is moderation being choosable on a per-subscriber basis. You can get an approximation of this using access rules, but after a little while and as list sizes grow, this will quickly become cumbersome and inefficient. Making another 'version' of the list, though, just starts to get ridiculous, as combinations of different versions start to become possible and desirable. I think to do this properly somehow the idea of per-subscriber options needs to be given to mlmmj, and that's both a sizeable job, and a bit of an upheaval to current practice. Perhaps something to consider long term. In the meantime, my most pressing need is what I originally wrote about, and perhaps an easier short-term solution to this is just to allow an exception list. Smartest may be to have two tunables: 'metoo' and 'notmetoo'; if neither exist, behaviour is as normal; if notmetoo exists, senders are included in the distribution unless listed in the notmetoo file; if metoo exists, senders are omitted from distribution unless listed in the metoo file. I would usually use the metoo file, then, but others may prefer the other. Perhaps we could have some discussion about how these other per-subscriber things could be implemented satisfactorily long-term later. I've pretty much finished my survey of things I'd like/find useful in mlmmj now, and everything else is easier than the stuff mentioned in this email. I'll raise the issues over time in new threads and hopefully help make some patches. Ben.