• restarting comp.newprod, comp.simulation and comp.std.announce

    From Ivan Shmakov@21:1/5 to All on Sat Apr 26 18:42:42 2025
    XPost: comp.std.misc

    [Cross-posting to news:comp.std.misc , as suggested in
    news:vs73t3$3hlga$[email protected] , but setting Followup-To:
    news:comp.misc , as the former group doesn't seem to be active.]

    As some might know, back in March I've volunteered to take over
    moderation of comp.newprod and comp.std.announce newsgroups
    that are currently unusable due to the lack of a moderator.
    (My news.groups.proposals post is in the References:.)

    In my post I also suggested for comp.simulation (and a few other
    non-comp.* groups) to become unmoderated. I don't seem to see
    any response to that, so assuming the intent is still to remove
    it, I'm volunteering to take over its moderation as well.

    Like I've mentioned, I've never had much interest in moderated
    newsgroups before, much less been a moderator of one, so I can't
    promise smooth operation, especially right from the start.

    And regardless, I'd prefer not doing the thing entirely on my
    own; anyone here interested in taking part, perchance?

    My plan is to set up message submission email addresses aliased
    to my own mailbox, and approve (or not approve) them from there.
    Should there ever be any co-moderators, I'd move submissions to
    a password-protected portion of an https: website and provide
    an HTTP/1-based API (usable with curl(1) and compatible with
    Lynx, naturally) to approve the messages there.

    It's possible that I'd implement a whitelist of known-good
    posters for auto-approval, though that will require a way for
    the software to /identify/ such posters. Such as, e. g., a
    direct email message (Newsgroups: and all) to the submission
    address, with a valid DKIM signature for the MAIL FROM: domain.

    There're scarcely any mention of the groups at hand under
    http://ftp.isc.org/usenet/news.announce.newgroups/comp/
    (they might've been created outside of the Big-8 practices
    employed today): comp.windows.x.i386unix and comp.lsi.testing
    mention news:comp.newprod and news:comp.simulation ,
    respectively, while comp.remove-former-inet lists .std.announce
    among the groups proposed for deletion, though the group was
    kept in the end.

    My intent for the groups is to, obviously, keep them on-topic.
    Other than that, I'm going to check for what IETF / IESG
    announcement lists are around here and, if fitting, start
    forwarding messages from there to comp.std.announce.

    My secondary objective would be to keep the groups reasonably
    low-traffic, limiting the number of posts (so to save the
    readers from pressing "next, please" too many times, etc.)
    rather their cumulative size (data transfer is cheap these
    days, though my MX /does/ have rather tight message size
    limits.) In particular, I might, on occasion, ask users to
    repost their handful of articles of few lines each as a single
    larger one.

    Should any of the groups attract too much traffic (as unlikely
    as it is to happen), I'd be proposing some sort of quality
    guidelines, and, once those are enacted, start rejecting
    articles failing to meet them.

    By the by, I'd appreciate Message-Id:s of whatever periodic
    postings that were seen there. A cursory web search didn't
    reveal much to go by, and as the groups might've been created
    outside of the Big-8 process, there's no charter on file for
    them, either.

    That'd be all for now. Thoughts?

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Marco Moock@21:1/5 to All on Sun Apr 27 08:28:16 2025
    XPost: news.groups.proposals

    Follow-Up2: news.groups.proposals

    On 26.04.2025 18:42 Uhr Ivan Shmakov wrote:

    [Cross-posting to news:comp.std.misc , as suggested in
    news:vs73t3$3hlga$[email protected] , but setting Followup-To:
    news:comp.misc , as the former group doesn't seem to be
    active.]

    As some might know, back in March I've volunteered to take
    over moderation of comp.newprod and comp.std.announce newsgroups
    that are currently unusable due to the lack of a moderator.
    (My news.groups.proposals post is in the References:.)

    In my post I also suggested for comp.simulation (and a few
    other non-comp.* groups) to become unmoderated. I don't seem to see
    any response to that, so assuming the intent is still to
    remove it, I'm volunteering to take over its moderation as well.

    If so, please mention it in the RfD discussion and also CC that to [email protected], so we definitely not miss it.

    Like I've mentioned, I've never had much interest in moderated
    newsgroups before, much less been a moderator of one, so I
    can't promise smooth operation, especially right from the start.

    Depends on what you want to do. You will need moderation infrastructure
    (or ask others that already have), this normally works fine.

    And regardless, I'd prefer not doing the thing entirely on my
    own; anyone here interested in taking part, perchance?

    I don't see a problem here if you operate it on your own.

    My plan is to set up message submission email addresses
    aliased to my own mailbox, and approve (or not approve) them from
    there.

    This will probably not work if your have an inbox at Google/MS/Yahoo
    out of the box.

    Should there ever be any co-moderators, I'd move submissions to
    a password-protected portion of an https: website and provide
    an HTTP/1-based API (usable with curl(1) and compatible with
    Lynx, naturally) to approve the messages there.

    There is webstump for that.

    It's possible that I'd implement a whitelist of known-good
    posters for auto-approval, though that will require a way for
    the software to /identify/ such posters. Such as, e. g., a
    direct email message (Newsgroups: and all) to the submission
    address, with a valid DKIM signature for the MAIL FROM:
    domain.

    Webstump can use rules based on sender address, but DKIM will not be
    available by design. The message is being injected into the news server
    the poster is using and then sent to the moderation servers (e.g.
    mine). This will forward it to the submission address. DKIM signing
    needs the private key of the domain (e.g. example.org) and the mail
    never passes the MSA/MTA of example.org.

    Forging is possible here.

    My intent for the groups is to, obviously, keep them on-topic.
    Other than that, I'm going to check for what IETF / IESG
    announcement lists are around here and, if fitting, start
    forwarding messages from there to comp.std.announce.

    Fine in my eyes.




    --
    kind regards
    Marco

    Send spam to [email protected]

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Ivan Shmakov@21:1/5 to All on Fri Jun 20 15:45:41 2025
    XPost: news.misc

    So, it's happened; I've just posted announcements to the groups:

    news:[email protected] - news:comp.simulation news:[email protected] - news:comp.newprod news:[email protected] - news:rec.photo.moderated

    (Still have some research to do for the news:comp.std.announce
    announcement.)

    I'm posting this here in comp.misc for several reasons:

    * I believe that among the readers here might be those interested
    in joining these groups, or those who know those who might be;

    * I'm curious what issues with my "common submission guidelines"
    below an open discussion might reveal;

    * I'd rather not post said guidelines to each of the groups
    individually.

    I'm also cross-posting to news.misc because it seems more
    fitting, though it doesn't seem to have been active recently.
    Feel free to respond there (disregarding Followup-To:) instead
    if you like; if anything, I'll be monitoring both groups for
    this discussion.

    The guidelines below apply to comp.newprod, comp.simulation,
    comp.std.announce and rec.photo.moderated; and reflect both
    what I consider to be "good Usenet behavior," and how the
    things are currently set up on my side.


    Usenet is a public forum and as such, articles submitted
    to newsgroups are expected to be of interest to readers at
    large, not aimed at a particular person. If you believe a
    participant of the group might have an issue with something
    that is not strictly under the topic of the group, please
    communicate your considerations elsewhere, such as by email.

    The use of your "real" name in From: is not strictly required,
    but a /consistent/ identity is still expected. "Example Inc.
    Sales Department" might be suitable for comp.newprod
    submissions; and your Wikimedia Commons account name, for your
    rec.photo.moderated ones.

    It is recommended that an *.invalid domain email address is
    used in From:, both to thwart email address harvesters /and/
    to avoid your submission being rejected due to SPF mismatch.
    That said, there should be a valid email address, and one
    you're entitled to use, /somewhere/ in the message (say,
    Reply-To: or the signature.) Mangling is allowed, so long as
    the mail domain ends with .invalid /and/ it is obvious to me
    how to get the original address from there. Postings that
    lack a valid email address might be approved, but if they're
    rejected, you won't know why.

    It's possible to submit articles directly over email (see
    below.) In such a case, please use an appropriate valid
    address in the envelope (MAIL FROM:.)

    Submissions should not exceed 128 KiB in size (header + body.)
    While my MTA /will/ accept somewhat larger ones to account for
    transit overhead (Received: and such), one should not rely on that.

    As an aside, note that Usenet archives might truncate long
    articles. For example, http://al.howardknight.net/ IME only
    stores up to 200 lines of the body.

    By the same merit, if you forward an article from elsewhere
    (say, from a blog), be sure to include the original URI near
    the /top/ of your post: that way, even if the article gets
    truncated, someone reading it from archives will still be
    pointed to the full orignal instance. It's also nice as it
    allows the reader to stop reading the article early if they
    do not find the original site to their taste.

    Avoid using non-ASCII characters in the headers: while some
    newsreaders do support RFC 2047, others do not.

    Submission bodies should be "plain" plain text (MIME type
    "text/plain", no format=flowed, etc.), either ASCII (with or
    without MIME headers), or 8-bit UTF-8 (MIME headers required.)
    If your submission has other parts (say, text/html), I'll
    remove them, unless it's obvious to me that they are important
    to the message as a whole, in which case I'll reject and ask
    to resubmit. In exceptional cases, I might edit your submission
    to fit if it doesn't already.


    It's possible to submit articles by email, such as explained
    in http://www.big-8.org/wiki/Moderated_newsgroups
    #How_can_I_bypass_my_News_Service_Provider? (URI split for
    readability.)

    To use this option, first make sure your message has a proper
    Newsgroups: header. (At some point I'm going to configure my Exim
    instance to reject emails to the submission addresses lacking one.)

    Messages submitted this way should have a valid email address
    in the envelope (MAIL FROM:), must not fail SPF verification,
    and preferrably should also have a valid DKIM signature. Note
    that it's generally still possible (and advisable) to have an
    *.invalid address in the From: header.

    On a Unix-like system with appropriately configured mail delivery,
    the following example /usr/sbin/sendmail invocation can be used
    as a starting point (tested on Debian GNU/Linux "Bookworm", but
    if I'm reading sendmail(1) correctly, it should work on NetBSD
    just as well; refer to the individual announcements for the
    actual submission addresses):

    $ cat < newspost
    From: Alice <[email protected]d>
    Reply-To: Alice <[email protected]>
    Newsgroups: example.newsgroup.moderated
    Message-Id: <[email protected]d>
    Date: Fri, 20 Jun 2025 10:45:08 +0000
    Subject: test

    This is a test message.
    $ /usr/sbin/sendmail \
    -i -f [email protected] \
    [email protected] \
    < newspost

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Computer Nerd Kev@21:1/5 to Ivan Shmakov on Sat Jun 21 09:33:52 2025
    In comp.misc Ivan Shmakov <[email protected]d> wrote:
    It's possible to submit articles by email, such as explained
    in http://www.big-8.org/wiki/Moderated_newsgroups
    ^
    #How_can_I_bypass_my_News_Service_Provider? (URI split for
    readability.)

    Broken link - needs a capital 'N'.

    On a Unix-like system with appropriately configured mail delivery,
    the following example /usr/sbin/sendmail invocation can be used
    as a starting point (tested on Debian GNU/Linux "Bookworm", but
    if I'm reading sendmail(1) correctly, it should work on NetBSD
    just as well; refer to the individual announcements for the
    actual submission addresses):

    $ cat < newspost
    From: Alice <[email protected]d>
    Reply-To: Alice <[email protected]>
    Newsgroups: example.newsgroup.moderated
    Message-Id: <[email protected]d>

    I'd suggest including how the unique part of the Message-Id can be
    generated, or at least warning that this part must be generated for
    each message. Examples with unique IDs in them tempt people to
    re-use that unique ID.

    I've subscribed to comp.newprod and rec.photo.moderated since I'm
    interested in them, although I don't currently have cause to post
    in either.

    --
    __ __
    #_ < |\| |< _#

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Ivan Shmakov@21:1/5 to All on Tue Jun 24 16:30:54 2025
    On 2025-06-20, Computer Nerd Kev wrote:
    In comp.misc Ivan Shmakov wrote:

    I'm also cross-posting to news.misc because it seems more fitting,
    though it doesn't seem to have been active recently.

    (Should've rather cross-posted to news.groups, which is both
    a better fit /and/ active.)

    It's possible to submit articles by email, such as explained
    in http://www.big-8.org/wiki/Moderated_newsgroups
    ^
    #How_can_I_bypass_my_News_Service_Provider? (URI split for
    readability.)

    Broken link - needs a capital 'N'.

    Indeed, thanks.

    On a Unix-like system with appropriately configured mail delivery,
    the following example /usr/sbin/sendmail invocation can be used as
    a starting point:

    $ cat < newspost
    From: Alice <[email protected]d>
    Reply-To: Alice <[email protected]>
    Newsgroups: example.newsgroup.moderated
    Message-Id: <[email protected]d>

    I'd suggest including how the unique part of the Message-Id can be generated, or at least warning that this part must be generated for
    each message. Examples with unique IDs in them tempt people to
    re-use that unique ID.

    Fair enough. I shouldn't have included either that or Date:
    in the example at all, as the mail subsystem adds them if
    missing anyway. Though it makes sense to mention that the
    harvesters can gather mail domains from Message-Id:s just as
    well; if anything, they're right there in the overview,
    alongside From:s and References:.

    Also, pre-generating a Message-Id: might aid troubleshooting
    should the message be lost in transit, though with the low
    traffic I'm anticipating for these groups, knowing From: and
    the day it was submitted would probably be sufficient.

    I've subscribed to comp.newprod and rec.photo.moderated since I'm
    interested in them, although I don't currently have cause to post
    in either.

    I'm the least hopeful about .newprod, and certainly don't
    expect to have any reason to post there myself in the
    foreseeable future (aside of a periodic pointer to guidelines
    and such), but it sure would be nice to get it to work.

    I can think of things to post to rec.photo.moderated (and
    comp.simulation, FTM), though I'm not sure there's much reason
    to prefer it to rec.photo.digital at this point.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)