From Newsgroup: news.admin.peering

Testing... testing... 123...

Looking at the PATH... is this a correct analysis of the various hops?

1. news.tcpreset.net is the injection point accepting the nntp post
2. .POSTED!not-for-mail confirms it's a direct NNTP POST
3. news-fra.pugleaf.net is likely one of tcpreset's primary peers.
4. srl.newsdeef.eu is a peering from a wider European mesh perhaps?
5. newsfeed.bofh.team Interesting. Do they share a peering relationship?
6. feeder.erje.net is a free-server feed node I think
7. 2.eu.feeder.erje.net another node, similar to what eternal-sept uses
8. news2.arglkargh.de now in the mid-tier general European mesh?
9. The final receiving peer (which varies per user)

Is this analysis of the path correct?
If not, what's the correct analysis?
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: news.admin.peering

Some nym <foo@bar.invalid> wrote:

Testing... testing... 123...

Looking at the PATH... is this a correct analysis of the various hops?

1. news.tcpreset.net is the injection point accepting the nntp post
2. .POSTED!not-for-mail confirms it's a direct NNTP POST
3. news-fra.pugleaf.net is likely one of tcpreset's primary peers.
4. srl.newsdeef.eu is a peering from a wider European mesh perhaps?
5. newsfeed.bofh.team Interesting. Do they share a peering relationship?
6. feeder.erje.net is a free-server feed node I think
7. 2.eu.feeder.erje.net another node, similar to what eternal-sept uses
8. news2.arglkargh.de now in the mid-tier general European mesh?
9. The final receiving peer (which varies per user)

Is this analysis of the path correct?
If not, what's the correct analysis?

Generally, there's no priority among peers. Two News sites could peer
directly, yet Path shows several intermediate peers. Peers flood each
other with articles not knowing what the other is missing.

!not-for-mail is a convention. You'll see that in a moderated newsgroup
in which the user's proto article isn't injected into Usenet till
approved by the moderator, and not directly by the user.
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: news.admin.peering

Am 20.06.26 um 16:33 schrieb Adam H. Kerman:

!not-for-mail is a convention. You'll see that in a moderated newsgroup
in which the user's proto article isn't injected into Usenet till
approved by the moderator, and not directly by the user.

You will see that in other articles too.
Dunno why that is included, as https://datatracker.ietf.org/doc/html/rfc1036#section-2.1.6 explicitly
says it should not be used for mail replies.
--
Gru|f
Marco

Junk-Mail bitte an trashcan@stinkedores.dorfdsl.de
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: news.admin.peering

Marco Moock <mm@dorfdsl.de> wrote:

Am 20.06.26 um 16:33 schrieb Adam H. Kerman:

You removed necessary context. This addresses an instance in which
not-for-mail appears on path even though the user did not submit his
proto article for injection himself.

I'm guessing Path is formed the same way even if part of the Path is
UUCP.

!not-for-mail is a convention. You'll see that in a moderated newsgroup
in which the user's proto article isn't injected into Usenet till
approved by the moderator, and not directly by the user.

You will see that in other articles too.
Dunno why that is included, as >https://datatracker.ietf.org/doc/html/rfc1036#section-2.1.6 explicitly
says it should not be used for mail replies.

I haven't read RFC 1036 in a while. I had not remembered that the path
could be formed with a choice of punctuation, not limited to a bang
path.

I think it's saying that older systems expected the path to begin with a
user's mailbox. Therefore, to force these systems to use Usenet in
followup and not Mail in reply, the path began with "not-for-mail".

I never used such a system so someone else will have to confirm.
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: news.admin.peering

This addresses an instance in which
not-for-mail appears on path even though the user did not submit his
proto article for injection himself.

AFAICT, the Path header only shows that the proto-article was injected by
the server news.tcpreset.net. The presence of the pseudo-sites POSTED and not-for-mail indicates that the article was posted directly by a user via
NNTP and that no email-based injection was involved.


LASTRELAY!news2.arglkargh.de!2.eu.feeder.erje.net!feeder.erje.net!newsfeed.bofh.team!srl.newsdeef.eu!news-fra.pugleaf.net!news.tcpreset.net!.POSTED!not-for-mail
LASTRELAY!news2.arglkargh.de!2.eu.feeder.erje.net!feeder.erje.net!news.mb-net.net!open-news-network.org!srl.newsdeef.eu!news-fra.pugleaf.net!news.tcpreset.net!.POSTED!not-for-mail


news.tcpreset.net - injecting server that accepted the post from the user POSTED - the article originated from a user session, not from a peer. not-for-mail - it was not injected via email or a mail-to-news gateway.

It seems to me, an outside observer, that the Path header does not reveal
the user's connection method. I don't see how it can prove or disprove
whether the posting was done via clearnet NNTP, Tor, I2P, VPN, or any other transport.

Mostly what I was asking about, given that news-fra.pugleaf.net and srl.newsdeef.eu appear identically in both paths, is whether it is
reasonable to infer that news.tcpreset.net uses a consistent upstream feed route through these peers when propagating articles toward those
destinations?
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: news.admin.peering

Some nym <foo@bar.invalid> wrote:

This addresses an instance in which
not-for-mail appears on path even though the user did not submit his
proto article for injection himself.

AFAICT, the Path header only shows that the proto-article was injected by >the server news.tcpreset.net. The presence of the pseudo-sites POSTED and >not-for-mail indicates that the article was posted directly by a user via >NNTP and that no email-based injection was involved.

That's not a concept. It's not possible. There has to be a gateway. I
just gave you an example in which the user does not submit his own
article for injection. The moderator approves the proto article then
submits it for injection. There is a gateway. There are mailing lists
gated to and from Usenet, which is possible thanks to the shared format.

There are mail2news servers; these are literal gateways.

If the proto article travelled some portion of the way to Usenet via
email relays, these are not shown on Path. Path begins with injection.

Since Marco had me reread RFC 1036, I remembered why not-for-mail is
used at all.

LASTRELAY!news2.arglkargh.de!2.eu.feeder.erje.net!feeder.erje.net!newsfeed.bofh.team!srl.newsdeef.eu!news-fra.pugleaf.net!news.tcpreset.net!.POSTED!not-for-mail
LASTRELAY!news2.arglkargh.de!2.eu.feeder.erje.net!feeder.erje.net!news.mb-net.net!open-news-network.org!srl.newsdeef.eu!news-fra.pugleaf.net!news.tcpreset.net!.POSTED!not-for-mail

news.tcpreset.net - injecting server that accepted the post from the user >POSTED - the article originated from a user session, not from a peer. >not-for-mail - it was not injected via email or a mail-to-news gateway.

Wrong. You'd see the News side of the gateway but not the gateway
itself, I think.

Uh, try it yourself? Doesn't dizum still have an associated mail2news
gateway?

It seems to me, an outside observer, that the Path header does not reveal
the user's connection method. I don't see how it can prove or disprove >whether the posting was done via clearnet NNTP, Tor, I2P, VPN, or any other >transport.

I ate a ham sandwich on the same day I posted to Usenet. That's not
revealed on Path either.

Mostly what I was asking about, given that news-fra.pugleaf.net and >srl.newsdeef.eu appear identically in both paths, is whether it is
reasonable to infer that news.tcpreset.net uses a consistent upstream feed >route through these peers when propagating articles toward those >destinations?

I already gave you an answer, which you clearly did not like. I shall
try harder to ignore ou from now on.
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: news.admin.peering

I shall try harder to ignore ou from now on.

I think the reason you misunderstood what I had said was that I didn't
explain it clearly enough for you to have fully understood it, so I
apologize for not repeating and being doubly clear using simple language.

It was my fault for not being as clear as I needed to have been, given
there is nothing wrong with what I said, which I believe and which I
believe others will concur.

I'm asking for help here, so I need to be very clear that others don't misunderstand, as you did, so I will apologize up front for my mistakes.

I wasn't denying the existence of gateways or moderated injections.
I was only noting that in this specific case, the Path shows
news.tcpreset.net as the injecting server and it shows that the article was
not email-injected.

I agree that earlier stages of a proto-article (moderation, mail2news,
UUCP, etc.) would not appear in the Path. My question was only about the visible upstream peers from the injecting server.

I apologize that you misunderstood that was the question, which is my fault
for not making it extremely clear to everyone as I'm asking for advice.
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: news.admin.peering

On 20 Jun 2026, "Adam H. Kerman" <ahk@chinet.com> posted some news:1117p8n$hhh8$3@dont-email.me:

Some nym <foo@bar.invalid> wrote:

This addresses an instance in which
not-for-mail appears on path even though the user did not submit his >>>proto article for injection himself.

AFAICT, the Path header only shows that the proto-article was injected
by the server news.tcpreset.net. The presence of the pseudo-sites
POSTED and not-for-mail indicates that the article was posted directly
by a user via NNTP and that no email-based injection was involved.

That's not a concept. It's not possible. There has to be a gateway. I
just gave you an example in which the user does not submit his own
article for injection. The moderator approves the proto article then
submits it for injection. There is a gateway. There are mailing lists
gated to and from Usenet, which is possible thanks to the shared
format.

There are mail2news servers; these are literal gateways.

If the proto article travelled some portion of the way to Usenet via
email relays, these are not shown on Path. Path begins with injection.

Since Marco had me reread RFC 1036, I remembered why not-for-mail is
used at all.

Chuckle...now there's a ruinous source.

LASTRELAY!news2.arglkargh.de!2.eu.feeder.erje.net!feeder.erje.net!newsf >>eed.bofh.team!srl.newsdeef.eu!news-fra.pugleaf.net!news.tcpreset.net!.P >>OSTED!not-for-mail >>LASTRELAY!news2.arglkargh.de!2.eu.feeder.erje.net!feeder.erje.net!news. >>mb-net.net!open-news-network.org!srl.newsdeef.eu!news-fra.pugleaf.net!n >>ews.tcpreset.net!.POSTED!not-for-mail

news.tcpreset.net - injecting server that accepted the post from the
user POSTED - the article originated from a user session, not from a
peer. not-for-mail - it was not injected via email or a mail-to-news >>gateway.

Wrong. You'd see the News side of the gateway but not the gateway
itself, I think.

Uh, try it yourself? Doesn't dizum still have an associated mail2news gateway?

One of the most reliable in Usenet history.

It seems to me, an outside observer, that the Path header does not
reveal the user's connection method. I don't see how it can prove or >>disprove whether the posting was done via clearnet NNTP, Tor, I2P,
VPN, or any other transport.

I ate a ham sandwich on the same day I posted to Usenet. That's not
revealed on Path either.

Mostly what I was asking about, given that news-fra.pugleaf.net and >>srl.newsdeef.eu appear identically in both paths, is whether it is >>reasonable to infer that news.tcpreset.net uses a consistent upstream
feed route through these peers when propagating articles toward those >>destinations?

I already gave you an answer, which you clearly did not like. I shall
try harder to ignore ou from now on.

lol, smh.

--- Synchronet 3.22a-Linux NewsLink 1.2