Long Live Usenet

[Kestral Gaian]

I'm writing from the UK, where I'm preparing to teach some university
students here more about usenet in the new year, and am re-joining as a
result.

It's time we reclaimed the internet!

[Marco Moock]

Great!

Does your university still operate an NNTP server?

[Retro Guy]

Welcome back! Usenet still lives :)

[vort3]

On 2023-12-22 00:25, Kestral Gaian wrote:
> teach some university
> students here more about usenet

Did you mean the USENET or NNTP protocol in particular?

I have lots of questions about NNTP and hope to find people that could
explain everything to me.

I tried to read on the internet but all the info is as basic as «type
your server address and port into your news app and press this button to
post messages», which is not nearly enough for my understanding.

[Marco Moock]

Am 22.12.2023 um 02:13:30 Uhr schrieb vort3:

> I have lots of questions about NNTP and hope to find people that
> could explain everything to me.

I recommend reading the RFCs that describe that protocol:

https://www.rfc-editor.org/rfc/rfc3977
https://www.rfc-editor.org/rfc/rfc6048

[Kestral Gaian]

I wish - it's never had one. I might explore the security impact of
getting one up and running in the new year, but I'm very keen to support
the protocol and the 'net as a whole if I can.

Thank you for the warm welcome!

[vort3]

On 2023-12-22 02:21, Marco Moock wrote:
> I recommend reading the RFCs that describe that protocol:
>
> https://www.rfc-editor.org/rfc/rfc3977
> https://www.rfc-editor.org/rfc/rfc6048

Thanks for the links. I hope some day I'll have enough time to read all
of that.

I see that response code 411 «no such newsgroup» is not used when
posting articles, it's… Interesting. Trying to post to a newsgroup that
does not exist doesn't throw any errors, it just makes your post disappear.

This is certainly not what I expected.

[Marco Moock]

Am 22.12.2023 um 03:42:28 Uhr schrieb vort3:

> I see that response code 411 «no such newsgroup» is not used when
> posting articles, it's… Interesting. Trying to post to a newsgroup
> that does not exist doesn't throw any errors, it just makes your post
> disappear.

Please use ncat or telnet to verify that and show use the output.
When posting to multiple groups, servers usually accept that if only
one of them exist on the server.

See my test post here:
Message-ID: <um2bog$7b9v$1...@solani.org>

If I only give a group that doesn't exist on the server (testet with
solani here), it gives exactly that error message.

[2023-12-21 22:47:37] NNTP< 441 No valid newsgroups in "de.test.ungueltig"

[Sn!pe]

Welcome, that's very good news.

--
^Ï^. Sn!pe, PA, FIBS - Professional Crastinator

My pet rock Gordon likes the closure of Google's Web2News gateway.

[candycanearter07]

Awesome, can't wait for new users!
--
user <candycane> is generated from /dev/urandom

[yeti]

Maybe let's try to add a P2Pish NNTP dimension? Via DynDNS/I2P/Tor?
Everyone would be responsible for their own node only. That could be
chaos for a while with groups popping into and out of existence like
virtual particles, but after a while a rule set for an a bit more
fluid/flexible sibling of Usenet may crystallise out.

Smells like a nice experiment.

--
I do not bite, I just want to play.

[Sn!pe]

Wouldn't that be equivalent to each group having an owner in charge,
just like a web forum? I thought the main advantage of Usenet was
precisely that no one person is in charge.

[yeti]

Have the RFCs at hand and start with something like ...

------------------------------------------------------------------------
$ rlwrap nc news.eternal-september.org nntp
200 news.eternal-september.org InterNetNews NNRP server INN 2.8.0 (20231205 snapshot) ready (posting ok)
help
100 Legal commands
ARTICLE [message-ID|number]
AUTHINFO USER name|PASS password|SASL mechanism [initial-response]|GENERIC program [argument ...]
BODY [message-ID|number]
CAPABILITIES [keyword]
(((...)))
quit
205 Bye!
.
------------------------------------------------------------------------

Drop `rlwrap` if you cannot easily install it.

Coding simple things like fetching an article by group+number or
message-id sure raise your experience points too.

Try something like ...

------------------------------------------------------------------------
$ printf '%s\r\n' 'article <um269a$16thd$1...@dont-email.me>' quit . | nc
news.eternal-september.org nntp | less
------------------------------------------------------------------------

... too.

Sloppy code in BACON, MicroPython or C to do the same is surprisingly
short. Trying to care for every possible error sure blows up that a lot
and that's the level I too haven't reached yet.

Posting via `netcat` is easy too. Sure noone wants to do that instead
of using a newsreader, but sure it'll give you a smile:

<ujmkbl$1ne6b$1...@anonymous.dont-email.me>

How about a local group like e-s.nntp-kindergarten? (kindergar_d_en?)
There may be a better name...

Let's play!

...and later peer a swarm of own small servers.

--
3. Hitchhiker 10: (81) The Hitch Hiker's Guide to the Galaxy has this to
say on the subject of flying. (82) There is an art, it says, or rather
a knack to flying. (83) The knack lies in learning how to throw
yourself at the ground and miss.

[yeti]

snip...@gmail.com (Sn!pe) writes:

> Wouldn't that be equivalent to each group having an owner in charge,
> just like a web forum? I thought the main advantage of Usenet was
> precisely that no one person is in charge.

That just would be NNTP and not Usenet.

[Sn!pe]

yeti <ye...@tilde.institute> wrote:

> snip...@gmail.com (Sn!pe) writes:
>
> > Wouldn't that be equivalent to each group having an owner in charge,
> > just like a web forum? I thought the main advantage of Usenet was
> > precisely that no one person is in charge.
>
> That just would be NNTP and not Usenet.

I just don't grok it; maybe enlightenment will come later.

[David Dalton]

On Dec 21, 2023, Marco Moock wrote
(in article <um2brb$7b9v$1...@solani.org>):

> [2023-12-21 22:47:37] NNTP< 441 No valid newsgroups in “d thinke.test.ungueltig”
I think E-S fails posting and gives an error if the first group
in a list of groups on the Newgroups line does not exist
on E-S, but not if it is listed later in that list.

--
David Dalton dal...@nfld.com https://www.nfld.com/~dalton (home page)
https://www.nfld.com/~dalton/dtales.html Salmon on the Thorns (mystic page)
"And if I shed a tear I won’t cage it; I won't fear love;
And if I feel a rage I won’t deny it; I won't fear love" (Sarah McLachlan)

[Ray Banana]

Thus spake David Dalton <dal...@nfld.com>

> I think E-S fails posting and gives an error if the first group
> in a list of groups on the Newgroups line does not exist
> on E-S, but not if it is listed later in that list.

This is demonstrably wrong, please see <8mjzp6v...@raybanana.net>
in alt.test.


--
Пу́тін — хуйло́
https://www.eternal-september.org

[Julieta Shem]

vort3 <sheri...@gmail.com> writes:

> On 2023-12-22 02:21, Marco Moock wrote:
>> I recommend reading the RFCs that describe that protocol:
>> https://www.rfc-editor.org/rfc/rfc3977
>> https://www.rfc-editor.org/rfc/rfc6048
>
> Thanks for the links. I hope some day I'll have enough time to read
> all of that.

Print it out. Read a page per day. You'll be done in a month if you
begin with RFC 977. In a few months if you begin with RFC 3977.

[Grant Taylor]

On 12/21/23 15:07, Kestral Gaian wrote:
> I wish - it's never had one. I might explore the security impact of
> getting one up and running in the new year, but I'm very keen to support
> the protocol and the 'net as a whole if I can.

Please feel free to ask questions. Also feel free to email me directly
if you would rather not talk publicly.

I'm also happy to help peer with new Usenet servers. Spread the joy.

> Thank you for the warm welcome!

:-)



--
Grant. . . .

[Grant Taylor]

On 12/21/23 14:13, vort3 wrote:
> I have lots of questions about NNTP and hope to find people that could
> explain everything to me.

Start asking questions and we can try to answer them.

> I tried to read on the internet but all the info is as basic as «type
> your server address and port into your news app and press this button to
> post messages», which is not nearly enough for my understanding.

Ya....

[Julieta Shem]

snip...@gmail.com (Sn!pe) writes:

> yeti <ye...@tilde.institute> wrote:
>
>> Kestral Gaian <k...@kesgai.com> writes:
>>
>> > On 21/12/2023 18:28, Marco Moock wrote:
>> >> Does your university still operate an NNTP server?
>> >
>> > I wish - it's never had one. I might explore the security impact of
>> > getting one up and running in the new year, but I'm very keen to
>> > support the protocol and the 'net as a whole if I can.
>>
>> Maybe let's try to add a P2Pish NNTP dimension? Via DynDNS/I2P/Tor?
>> Everyone would be responsible for their own node only. That could be
>> chaos for a while with groups popping into and out of existence like
>> virtual particles, but after a while a rule set for an a bit more
>> fluid/flexible sibling of Usenet may crystallise out.
>>
>> Smells like a nice experiment.
>
> Wouldn't that be equivalent to each group having an owner in charge,
> just like a web forum? I thought the main advantage of Usenet was
> precisely that no one person is in charge.

I think the USENET is for people like us. Many folks can't easily
handle this no-one-is-in-charge. So, one thing that I think
universities and organizations can do is to set up an authenticated NNTP
server for their internal discussions --- something closed. (Abusers
would be easily suspended by system administrators. No privacy. Just a
local NNTP forum.) But, this local NNTP server would also carry some
USENET groups that could be of some interest to the local people. The
locals wouldn't be able to post to the USENET groups, but they could
read it --- and this way the USENET would be presented to students or
whoever they are. People would likely ask --- why can't we post to
these groups? And we would instruct people on how to join the USENET.

Universities send and receive a lot of mail. A local NNTP server would
serve them a lot better. Threaded, archived, public. The mail box
would get a lot quieter. Maybe that's precisely what people don't want
--- but I would really like that.

[immibis]

What would be the difference from current Usenet?

[Daniel65]

Grant Taylor wrote on 22/12/23 5:07 pm:

> On 12/21/23 15:07, Kestral Gaian wrote:
>> I wish - it's never had one. I might explore the security impact of
>> getting one up and running in the new year, but I'm very keen to
>> support the protocol and the 'net as a whole if I can.
>
> Please feel free to ask questions.  Also feel free to email me directly
> if you would rather not talk publicly.

If Kestral is about to start teaching students AND wants to introduce
them to UseNet, I'm guessing it would be desirable that Kestral CAN
access a REFERENCE SOURCE without the students knowing so, whilst I
think most here would help out, being able to get to the REFERENCE
SOURCE undiscovered would be useful!!

>
> I'm also happy to help peer with new Usenet servers.  Spread the joy.
>
>> Thank you for the warm welcome!
>
> :-)

--

Daniel

[immibis]

Usenet may reclaim the internet from large corporations - but spammers
and trolls reclaimed the Usenet from users. There isn't any good
solution to the spam problem. And even if veteran usenet users may not
see persistent trolls as spammers, I do.

If I hooked up ChatGPT to respond to articles at random, that would be
spam, right? So why isn't it spam when a human writes in an equally
useless way?

Most successful solutions to spam so far, across all systems, have
involved appointing moderators to remove it from all users' feeds.

[immibis]

On 12/22/23 07:17, Julieta Shem wrote:
> The
> locals wouldn't be able to post to the USENET groups, but they could
> read it --- and this way the USENET would be presented to students or
> whoever they are. People would likely ask --- why can't we post to
> these groups? And we would instruct people on how to join the USENET.

Or that server could just join the USENET.

[Daniel65]

Kestral Gaian wrote on 22/12/23 5:25 am:

A Uni course introducing students to UseNet in Scandinavia this year!
A Uni course introducing students to UseNet in U.K. next year!

UseNet refuses to lie down and die!! ;-P
--
Daniel

[Daniel65]

candycanearter07 wrote on 22/12/23 11:05 am:

> On 12/21/23 12:25, Kestral Gaian wrote:
>> I'm writing from the UK, where I'm preparing to teach some university
>> students here more about usenet in the new year, and am re-joining as
>> a result.
>>
>> It's time we reclaimed the internet!
>
> Awesome, can't wait for new users!

The more, the merrier!!
--
Daniel

[Marco Moock]

Current usenet servers mostly use NNTP over the normal IPv4/IPv6
internet instead of TOR or I2P.

[yeti]

Tor isn't thought as main feature, sust to have a constand provider
independent address. Iirc I mentioned DynDNS too... because I think
more of leaf systems. For the NATed ones an overlay network is the
easiest way to be reachable without too many digging through the router
config if there is one at all. The NAT may be on the provider side.
DynDNS only solves parts of the problem.

[yeti]

Tor isn't thought as main feature, it just helps to have a constant
provider independent address. Iirc I mentioned DynDNS too ... because I

think more of leaf systems. For the NATed ones an overlay network is
the easiest way to be reachable without too many digging through the
router config if there is one at all. The NAT may be on the provider
side. DynDNS only solves parts of the problem.

[yeti]

Tor isn't thought as main feature, it just helps to have a constant
provider independent address. Iirc I mentioned DynDNS too ... because I

think more of systems at home. For the NATed ones an overlay network is

the easiest way to be reachable without too many digging through the
router config if there is one at all. The NAT may be on the provider
side. DynDNS only solves parts of the problem.

[Retro Guy]

There are some. I run two nntp servers inside I2P and one in Tor. I have
two nntp peers in I2P also.

v77lu6t26velvaddm4gibyzd5sogskuczp7vwoc4vxmukvvaucva.b32.i2p
tk4pwu574u4zrndofwx7dtlh2g2swauhtfblyrbhzxypznomgxxhgqyd.onion:119

[immibis]

On 12/22/23 09:55, Marco Moock wrote:
> Current usenet servers mostly use NNTP over the normal IPv4/IPv6
> internet instead of TOR or I2P.
>

Because each NNTP link is configured separately and negotiated by
humans, they can use any protocol that works for them. Apparently there
are still UUCP links in use on Usenet. Tor and I2P tunnels are no problem.

[Marco Moock]

With IPv4-NAT on the provider-side, your server can't be reached with
IPv4. Although, you can use IPv6 because there is no NAT at the ISP.
You can of course only peer with IPv6-capable servers.

[Sn!pe]

yeti's article (that you are responding to):
Message-ID: <87plyye...@tilde.institute>
is missing from my E-S feed, although I see from Howard Knight's article
lookup service that it was posted on E-S; I wonder why I can't see it?
<http://al.howardknight.net/?ID=170327271700>. How odd.

Whatever.
It seems to me that adding complication to a system that is already
working well is in part gilding the lily and part reinventing the wheel.
What is the expected benefit from these changes? Perhaps I'm just
misunderstanding something; convince me, please.

[Julieta Shem]

immibis <ne...@immibis.com> writes:

> On 12/22/23 09:55, Marco Moock wrote:
>> Current usenet servers mostly use NNTP over the normal IPv4/IPv6
>> internet instead of TOR or I2P.
>>
> Because each NNTP link is configured separately and negotiated by
> humans, they can use any protocol that works for them. Apparently

> there are still UUCP links in use on Usenet. [...]

Can you point us to some documentation on at least one? I'd like to
know more about them.

[Julieta Shem]

You're right. These solutions seem to boil down to the benevolent
dictator. But it seems to be a pattern than malevolent NNTP sysadmins
are not the majority because, if they were, we'd be in a mess right now
and USENET would be destroyed. Google Groups was effectively a
malevolent one --- it wouldn't remove the spam. So, although not what
we would like to call a solution, it seems that with this great
community work, USENET continues. Maybe that's a solution? Maybe we
should just continue this way until some nontrivial strategy appears to
ease the burden?

[Julieta Shem]

Or they could keep a low profile telling their users --- yes, some of
these groups belong to the USENET, a nontrivial forest you should not go
in without preparation. --- Yes, if you want to talk to such complex
and complicated anarchic outside world, step out of the boundaries of
our peaceful respectful sane orderly world, you could get an NNTP
account with some of the major players and we wish you luck.

Sounds like a marketing campaign. We mystify it a little. Show the tip
of the iceberg, teach them to identify what is USENET and what is just
quiet sane orderly local group and leave the rest up to them. This
sounds interesting. Besides, it kinda trivializes the sysadmin duties
of this local NNTP server.

[Sn!pe]

Wise words.

--
^Ď^. Sn!pe, PA, FIBS - Professional Crastinator

[Jim Jackson]

On 2023-12-22, Julieta Shem <js...@yaxenu.org> wrote:
> ... So, one thing that I think

> universities and organizations can do is to set up an authenticated NNTP
> server for their internal discussions --- something closed. (Abusers
> would be easily suspended by system administrators. No privacy. Just a
> local NNTP forum.) But, this local NNTP server would also carry some
> USENET groups that could be of some interest to the local people. The
> locals wouldn't be able to post to the USENET groups, but they could
> read it --- and this way the USENET would be presented to students or
> whoever they are. People would likely ask --- why can't we post to
> these groups? And we would instruct people on how to join the USENET.
>
> Universities send and receive a lot of mail. A local NNTP server would
> serve them a lot better. Threaded, archived, public. The mail box
> would get a lot quieter. Maybe that's precisely what people don't want
> --- but I would really like that.

We did that at University of Leeds School of computing deptartment
between the early 1990's to 2010-ish.

We had "official" newsgroups for each Module, one only postable by
lecturer and helpers along with an associated discuss group for the
students. Andf there were a load of dept./topic discussion newsgroups.
I believe that the UK university networking team ran a standard news
server, which everyone also had access to.

worked well, until the university mandated a uniform web-based student
forum thingy across the university. in the early 2000's. The non-mudule
based groups existed for a while but withered away. I'd retired by then.

[Julieta Shem]

Not surprised. It's a very uneducational decision, though. So,
educators should oppose. And I'm sure they do, but they're so few.

I think we should not fight it. I don't even mind pretending that
USENET is history and showing it as some kind of history class. People
can be so clueless --- we must be patient. What I mean is --- we could
introduce the news groups as some kind of esoteric thing.

Students and teachers these days set up Discord servers, Whatsapp
groups, Telegram whatever. This will continue. For sure. But perhaps
one or two teachers can gather themselves in an NNTP servers as if they
were saying --- here's an alternative (a lot more suitable for technical
matters). --- That's how we ``used to do it'' in the 80s and 90s.
History is, after all, an academic thing and bloody useful in technical
subjects (and every other one).

[Kestral Gaian]

It'll be great to work out how to get something set up, so any advice is
gladly appreciated!

[Daniel65]

Sn!pe wrote on 23/12/23 7:02 am:

> Marco Moock <mm+s...@dorfdsl.de> wrote:
>> Am 22.12.2023 um 10:49:04 Uhr schrieb yeti:
>>> Marco Moock <mm+s...@dorfdsl.de> writes:
>>>> Am 22.12.2023 um 09:08:02 Uhr schrieb immibis:
>>>>> What would be the difference from current Usenet?
>>>>
>>>> Current usenet servers mostly use NNTP over the normal
>>>> IPv4/IPv6 internet instead of TOR or I2P.
>>>
>>> Tor isn't thought as main feature, it just helps to have a
>>> constant provider independent address. Iirc I mentioned DynDNS
>>> too ... because I think more of leaf systems. For the NATed ones
>>> an overlay network is the easiest way to be reachable without too
>>> many digging through the router config if there is one at all.
>>> The NAT may be on the provider side. DynDNS only solves parts of
>>> the problem.
>>
>> With IPv4-NAT on the provider-side, your server can't be reached
>> with IPv4. Although, you can use IPv6 because there is no NAT at
>> the ISP. You can of course only peer with IPv6-capable servers.
>
> yeti's article (that you are responding to): Message-ID:
> <87plyye...@tilde.institute> is missing from my E-S feed,
> although I see from Howard Knight's article lookup service that it
> was posted on E-S; I wonder why I can't see it?
> <http://al.howardknight.net/?ID=170327271700>. How odd.

Neither can I (sort of) .... I see

immibis
Message-ID: <um3g52$1g4h3$1...@dont-email.me>

Marco replying to immibis
Message-ID: <um3iup$85mt$1...@solani.org>

Then Marco 'replying' to Marco
Message-ID: <um41e8$85mt$2...@solani.org>

Then your (Snipe) reply to Marco
Message-ID: <1qm5muw.186oqbk170d91tN%snip...@gmail.com>

*BUT* .... I do see yeti's reply (that you miss) to Marco
Message-ID: <87msu2e...@tilde.institute>

So, for me, it's all there ... just not threading properly!!
--
Daniel

[Marco Moock]

Am 22.12.2023 um 20:02:11 Uhr schrieb Sn!pe:

> What is the expected benefit from these changes? Perhaps I'm just
> misunderstanding something; convince me, please.

TOR provides a way were a computer that is behind CGNAT or DS-Lite (not
reachable via IPv4 from outside) is reachable.
There are ISPs that only provide IPv4 with CGNAT or even implement SPI
firewalls (common in cellular networks in Germany).

With TOR, you can still access a server there.

[Marco Moock]

Am 22.12.2023 um 09:12:21 Uhr schrieb immibis:

> Usenet may reclaim the internet from large corporations - but
> spammers and trolls reclaimed the Usenet from users. There isn't any
> good solution to the spam problem. And even if veteran usenet users
> may not see persistent trolls as spammers, I do.

The vast amount of spam comes from a small amount of Usenet servers.
One of them will close down in Feb.

Many users in de.* killfiled Google, aioe and mixmin. There is/was a
reason for that.

[Sn!pe]

Daniel65 <dani...@nomail.afraid.org> wrote:

> Sn!pe wrote on 23/12/23 7:02 am:

[...]

> > yeti's article (that you are responding to): Message-ID:
> > <87plyye...@tilde.institute> is missing from my E-S feed,
> > although I see from Howard Knight's article lookup service that it
> > was posted on E-S; I wonder why I can't see it?
> > <http://al.howardknight.net/?ID=170327271700>. How odd.
>
> Neither can I (sort of) .... I see
>
> immibis
> Message-ID: <um3g52$1g4h3$1...@dont-email.me>
>
> Marco replying to immibis
> Message-ID: <um3iup$85mt$1...@solani.org>
>
> Then Marco 'replying' to Marco
> Message-ID: <um41e8$85mt$2...@solani.org>
>
> Then your (Snipe) reply to Marco
> Message-ID: <1qm5muw.186oqbk170d91tN%snip...@gmail.com>
>
> *BUT* .... I do see yeti's reply (that you miss) to Marco
> Message-ID: <87msu2e...@tilde.institute>
>
> So, for me, it's all there ... just not threading properly!!
>

And now I see it too, but misthreaded. I guess there may be a flaky
'reader in the subthread that has somehow messed up the References.

[Sn!pe]

[disclaimer: what follows is probably outdated]

Yes, I'm fairly familiar with Tor. Are you talking about running
servers on consumer-grade connections? I don't know of an ISP that
is happy to do that. I'm not sure that an ISP would be particularly
pleased about a Tor node being hosted either. I imagine that a higher
grade connection would come with a fixed IP if requested.

I experimented with dynamic DNS a decade or so ago for the dynamic IP
I then had. That worked quite well and is somewhat simpler than what is
proposed. I never ran a serious server using it though.

[Retro Guy]

On Sat, 23 Dec 2023 14:57:46 +0000, Sn!pe wrote:

> Daniel65 <dani...@nomail.afraid.org> wrote:
>
>> Sn!pe wrote on 23/12/23 7:02 am:
> [...]
>>> yeti's article (that you are responding to): Message-ID:
>>> <87plyye...@tilde.institute> is missing from my E-S feed,
>>> although I see from Howard Knight's article lookup service that it
>>> was posted on E-S; I wonder why I can't see it?
>>> <http://al.howardknight.net/?ID=170327271700>. How odd.
>>
>> Neither can I (sort of) .... I see
>>
>> immibis
>> Message-ID: <um3g52$1g4h3$1...@dont-email.me>
>>
>> Marco replying to immibis
>> Message-ID: <um3iup$85mt$1...@solani.org>
>>
>> Then Marco 'replying' to Marco
>> Message-ID: <um41e8$85mt$2...@solani.org>
>>
>> Then your (Snipe) reply to Marco
>> Message-ID: <1qm5muw.186oqbk170d91tN%snip...@gmail.com>
>>
>> *BUT* .... I do see yeti's reply (that you miss) to Marco
>> Message-ID: <87msu2e...@tilde.institute>
>>
>> So, for me, it's all there ... just not threading properly!!
>>
>
> And now I see it too, but misthreaded. I guess there may be a flaky
> 'reader in the subthread that has somehow messed up the References.

The article in question seems to have a Superseds header:
Supersedes: <87sf3ue...@tilde.institute>

Maybe some viewed the original message, have the MID and shared it, but
others are seeing the new MID.

Just my guess.

[vort3]

On 2023-12-22 07:29, yeti wrote:
> How about a local group like e-s.nntp-kindergarten? (kindergar_d_en?)
> There may be a better name...

Yeah that would actually be great.

[Sn!pe]

That figures, thanks, RG.

--
^Ï^. Sn!pe, PA, FIBS - Professional Crastinator

My pet rock Gordon just is.

[Marco Moock]

Am 23.12.2023 um 15:22:49 Uhr schrieb Sn!pe:

> Yes, I'm fairly familiar with Tor. Are you talking about running
> servers on consumer-grade connections? I don't know of an ISP that
> is happy to do that.

They are an internet service provider and not a "you are
only allowed to retrieve the content we want" company.
If they are, cancel the contract there and choose another.

> I'm not sure that an ISP would be particularly pleased about a Tor
> node being hosted either.

That is a hidden service then, so it is hard for them to detect that
TOR node.

If they don't want that amount of traffic, they must not sell flat
rates.

> I imagine that a higher grade connection
> would come with a fixed IP if requested.

Depends on the ISP. Some new ISPs can't provide that because they don't
have enough IPv4 addresses.

[Sn!pe]

Not all ISPs are created equal, that's for sure. As it happens, I have
a fixed IP but I very rarely use it barefoot. Almost all of my Internet
use is via VPN, just because.

BTW, don't be so sure that ISPs can't detect Tor. While I don't know
much about it, there is a thing called 'deep packet inspection'.

--
^Ï^. Sn!pe, PA, FIBS - Professional Crastinator

My pet rock Gordon just is.

[yeti]

snip...@gmail.com (Sn!pe) writes:

> Yes, I'm fairly familiar with Tor. Are you talking about running
> servers on consumer-grade connections?

No.

I'm not running a Tor node extending the Tor network.

I'm running Tor and HiddenServices. No "dark" stuff, see signature.

Please don't mix up those two setups and scare away potentially future
users of HiddenServices.

[Sn!pe]

It's quite a while since I played with that stuff, it seems I'm behind
the times. I'm not trying to scare anybody away, I'm just trying to
grok this proposal myself. I certainly see no good reason to dissuade
those who want to try it. I still can't see how it might be widely
useful but if there's a valid usage case then go for it; it will be
interesting to watch.

--
^Ï^. Sn!pe, PA, FIBS - Professional Crastinator

My pet rock Gordon just is.

[Grant Taylor]

On 12/22/23 14:34, Julieta Shem wrote:
> Can you point us to some documentation on at least one? I'd like to
> know more about them.

I've had test UUCP-over-SSH links pass Usenet articles for the fun of it.



--
Grant. . . .

[yeti]

Grant Taylor <gta...@tnetconsulting.net> writes:

> I've had test UUCP-over-SSH links pass Usenet articles for the fun of
> it.

I played with port-540-UUCP-over-Tor. That needed a config for very
peer on every peer and soon got too much (complete graph complexity).

Then I switched to SMTP-over-Tor to have a mailbox open to everyone just
knowing the address(es). I'll care about restrictions when the need
shows up. So far it did not.

[David Dalton]

On Dec 22, 2023, Ray Banana wrote
(in article <8mfrzuv...@raybanana.net>):

> Thus spake David Dalton<dal...@nfld.com>
>
> > I think E-S fails posting and gives an error if the first group
> > in a list of groups on the Newgroups line does not exist
> > on E-S, but not if it is listed later in that list.
>
> This is demonstrably wrong, please see<8mjzp6v...@raybanana.net>
> in alt.test.

OK, though I thought I observed that behaviour in a followup
to a crosspost to alt.atheism some months ago, but I don’t
remember the details.

--
David Dalton dal...@nfld.com https://www.nfld.com/~dalton (home page)
https://www.nfld.com/~dalton/dtales.html Salmon on the Thorns (mystic page)
"And if I shed a tear I won’t cage it; I won't fear love;
And if I feel a rage I won’t deny it; I won't fear love" (Sarah McLachlan)

[immibis]

On 12/22/23 21:34, Julieta Shem wrote:

> immibis <ne...@immibis.com> writes:
>
>> Because each NNTP link is configured separately and negotiated by
>> humans, they can use any protocol that works for them. Apparently
>> there are still UUCP links in use on Usenet. [...]
>
> Can you point us to some documentation on at least one? I'd like to
> know more about them.
>

Try comp.mail.uucp

[immibis]

As I understand it, it's very uncommon to get penalized by your ISP for
running a server on your residential connection, unless it somehow
causes a problem for them. Many people run Minecraft servers at home
without any issue. However, they aren't obligated to make it easy for
you, either.

[immibis]

I read that groups are only created if there's enough volume of traffic
to justify a separation from another group. Obviously most of them don't
have that volume any more, but they did at some point.

"kindergarten" is used in English.

[immibis]

Well, locking people out of having an account until they jump through
some pointless hoops seems to have worked for Blue Sky.

[immibis]

UUCP, like NNTP, is a graph of individually-configured point-to-point
links. Trying to make a complete graph is a waste of time.

[Daniel65]

immibis wrote on 24/12/23 10:07 pm:

"... only created if there's enough volume ..."?? How would you/anyone
know if there would be enough traffic in a NG unless there actually was
that NG??
--
Daniel

[Chris Jacobs]

Op 24-12-2023 om 13:57 schreef Daniel65:

Traffic that was not quite ontopic in other newsgroups but that would
have been totally ontopic in the new one.

[Grant Taylor]

On 12/24/23 05:24, immibis wrote:
> UUCP, like NNTP, is a graph of individually-configured point-to-point
> links. Trying to make a complete graph is a waste of time.

I actually suspect that a complete graph / full mesh would be a very low
ROI.

If you plan out the network well, distributing things can be more efficient.

T1: S1 -> S2
T2: S1 -> S3 while S2 -> S4
T3: S1 -> S5 while S2 -> S6 while S3 -> S7

Three time slots sends to seven systems.

This is super-linear growth. But I think that it's not exponential growth.

The point being that you don't /need/ the complete graph / full mesh to
send articles this way.

Admittedly, this may be an artifact of which system the articles
originate form and how they would feed to each other.

[candycanearter07]

Hey, there's still plenty of old-tech clubs at schools, maybe we can
normalize a USENET club?
--
user <candycane> is generated from /dev/urandom

[yeti]

vort3 <sheri...@gmail.com> writes:

> On 2023-12-22 07:29, yeti wrote:
> Yeah that would actually be great.

You're on ~nntp too or is the user with the same name there just
accidentally someone using the same nick?

[vort3]

On 2023-12-25 06:27, yeti wrote:
> You're on ~nntp too or is the user with the same name there just
> accidentally someone using the same nick?

Do you mean news.tilde.club NNTP? What ~nntp?

If yeas, then yeah, it's me. I got really interested in NNTP when I
started using tildes (I'm on ~team, ~club and rawtext.club too) so I'm
there, reading news, but also I'm here trying to find more active groups.

[Daniel65]

Chris Jacobs wrote on 25/12/23 12:03 am:

Yes, sure .... but if that was only, say, five or ten posts a day in
each of, say, a hundred newsgroups out of the, what, 100,000++ possible
newsgroups, how would any ISP realise how popular a new newsgroup might be??
--
Daniel

[yeti]

vort3 <sheri...@gmail.com> writes:

> Do you mean news.tilde.club NNTP? What ~nntp?

Yip!

> If yeas, then yeah, it's me. I got really interested in NNTP when I
> started using tildes (I'm on ~team, ~club and rawtext.club too) so I'm
> there, reading news, but also I'm here trying to find more active
> groups.

Let's switch over to there. Group tilde.projects? If you have a better
idea just ping. In their IRC I'm only in #institute. In the long(er)
run I can dedust my stuck INN on PIs experiment, so we'd have something
to trash/crash as often as we like.

[immibis]

They don't, which is why there haven't been *any* new groups in a very
long time. The one that broke the spell is comp.infosystems.gemini.

Also, ISPs don't create groups.

[yamo']

Hi,

Kestral Gaian a tapoté :
>
> I wish - it's never had one. I might explore the security impact of
> getting one up and running in the new year, but I'm very keen to support
> the protocol and the 'net as a whole if I can.
>
> Thank you for the warm welcome!

You can ask to the Tsukuba newsmaster.

Path : gama.is.tsukuba.ac.jp

I can give you by mail the contact I have.

Happy christmas!

--
Stéphane 🔗<http://pasdenom.info/fortune/>
On ne peut pas transporter partout avec soi le cadavre de son père.
-+- Guillaume Apollinaire -+-

[Daniel65]

immibis wrote on 25/12/23 11:40 pm:

Yes, you are correct .... but they are the ones that include the 'new'
group on their feeds, aren't they?? ;-P
--
Daniel