> From: "Rick Moen" <rick(a)linuxmafia.com>
> Subject: Re: [BALUG-Admin] list settings? ...
> Date: Fri, 18 Aug 2017 06:20:27 -0700
>> Should decide before lists.balug.org. moves over. But other than that
>> I think either sounds fine for BALUG-Test. What say ye?
>> Preference/recommendation on BALUG-Test? I'm totally open on that
>> one.
>
> Yearly, definitely.
>
> If the unlikely happens and there's a huge 2017 chunk, all you need do
> is change to a different chunk size and regenerate. Nobody is going to
> have permalinks, so it doesn't matter.
And done - setting changed on BALUG-Test to yearly.
If there's any archive bit 'o that that doesn't automagically
switch over to yearly, can deal with that sometime before
lists.balug.org moves over ... no great rush before then.
Just a From test - ignore
From postmaster(a)example.com Sat Aug 12 08:04:19 2017
From postmaster(a)example.com Sat Aug 12 08:04:19 2017
> From postmaster(a)example.com Sat Aug 12 08:04:19 2017
And yes, the full raw mbox archive for BALUG-Test
is publicly available:
$ curl -I https://temp.balug.org/pipermail/balug-test.mbox/balug-test.mbox
HTTP/1.1 200 OK
Date: Tue, 11 Jul 2017 14:29:02 GMT
Server: Apache/2.4.10 (Debian)
Last-Modified: Tue, 11 Jul 2017 14:23:34 GMT
ETag: "a046-5540b71d48512"
Accept-Ranges: bytes
Content-Length: 41030
Content-Type: application/mbox
$
Procedure documented on:
https://www.wiki.balug.org/wiki/doku.php?id=balug:mail_and_lists
notably:
o Y full mbox archive publicly available; procedure:
add:
PUBLIC_MBOX = Yes
to file:
/etc/mailman/mm_cfg.py
restart mailman
for existing lists, toggling archive from public to private and
back again seems sufficient to then create the needed link
Also making it available via public rsync is on the todo list, but
not towards the higher/highest priority bits.
"Not worse than" migration off of DreamHost.com is higher on the priorities.
Additional improvements generally come later.
[Bcc: BALUG-Test list]
So ... BALUG-Test list ... not an official announcement ... yet, but,
it's up and running, and mostly functional ... with some noteworthy
caveats/issues still to be addressed (and probably some I've not yet
discovered).
Policy, etc. - haven't really written that up yet, but shall.
In the meantime, it goes approximately as follows:
Fairly similar to BALUG-Talk - generally need to be subscribed to post,
no job postings (no reason to have exceptions on that on the Test list)
It *is* a test list, so some things are and generally will remain
different:
Recommended convention: if you *do not* want a response to your posting
to the list, include both of these two strings:
test
ignore
in your Subject: header
and they can be upper and/or lower case (testing should be done case
insensitive). If you don't include both test and ignore, you may get
human and/or automated responses, e.g. received your posting fine at ...
and here's full headers as received. No guarantees you'll get such
response(s), but if you don't include test and ignore, you may get such
response(s).
The list may wink in and out of existence, but for the most part will be
persistent, have preserved archives, etc. Archives may be reloaded
and/or altered (posts added/dropped) - notably for testing - but for the
most part the archives will generally contain what was posted to the
list.
Anyway, that's at least approximation of the policy (will write it up
more later, and have it on the list page).
Caveats - at present the archives are effectively broken. It's a web
server configuration issue - they're otherwise fine and being archived
(and it's not filesystem or file permissions or the like). I'll get
that fixed (one of the relatively higher priorities).
At present, emails with an empty body are being rejected. I still need
to fix that (as legitimate mailman commands may have command in Subject
header and have empty body).
The mail server has a whole lot 'o anti-spam stuff on it. Fairly likely
some of that may be too aggressive and/or annoying. I like how it's
rejecting tons 'o cr*p spam :-) ... but I want to keep the nuisance
factor sufficiently low on legitimate email. It will probably need some
adjusting. If you get a "hard" / "permanent" failure (or the mail
server automagically blacklist or bans you on what should be legitimate
email), shoot me an email note: Michael.Paoli(a)cal.berkeley.edu - and
include full headers and the complete bounce message contents.
Some of the greylisting and flood protection may be a bit annoying -
especially on first messages. E.g. a legitimate email doesn't
necessarily get an "instant" response/acceptance - I've tuned some of
that down so initial delays are less annoying (e.g. dropped to 2
minutes) while still effectively thwarting spammers. Anyway, much of
that may yet be subject to further adjustments, etc., as reasonably
appropriate. There are also a lot of block lists that are used ... at
least some of those may be a bit too aggressive in their blocking and
block some legitimate email. Anyway, let me know where issues are
encountered.
Also, the site is neither high bandwidth nor high availability. So,
sometimes it may not be available ... but for the most part it generally
should be up and available.
So, where is it, how to get started?
List page, here:
https://temp.balug.org/cgi-bin/mailman/listinfo/balug-test
It, and the other BALUG lists, will temporarily migrate from
lists.balug.org to temp.balug.org - and later settle back on
lists.balug.org (part of our migration away from
DreamHost.com). BALUG-Test list first, then BALUG-Admin,
then BALUG-Talk, and last, BALUG-Announce.
And how can you tell if your post was successful without being able to
see the archives?
Looks like with the default:
Receive your own posts to the list? Yes
One should be able to tell/distinguish by headers, etc., in received
posting.
Looks like also there's default:
Avoid duplicate copies of messages? Yes
One might possibly want to change to No to help ensure all copies are
received.
And ... well, not exactly step-by-step, but tracking fair bit of
migration and testing here:
https://www.wiki.balug.org/wiki/doku.php?id=balug:mail_and_lists
(it's not fully current ... but not that far from current either)