Question about usenet message propagation, control, etc
News Guy
explain something about how usenet messages are propagated, where the
control points lie, etc.
I would like to hear comments as to why or how these comments and
explanations are either correct or incorrect.
---------------------
There ARE master servers... whether you choose to call HOW messages get
transferred across the world MIRRORING or propagation or peering or
something else, USENET REQUIRES there be some place wherein information
/ postings can be collected, checked for duplicates, clarified, and
otherwise prepared.
Can you post and have messages appear OUTSIDE of the master services,,
yes,,, however, were it NOT for the EXTERNAL individual data bases and
localized transference, those messages would appear NO WHERE else EXCEPT
between whatever localized services might transfer amongst themselves.
Where it not for Google and other archival services, this might be more
clear as they have muddy the waters with the ability to pull/locate
messages OUTSIDE of master sites and the master servers/services, and
become, in essence a secondary master source.
Sites and services CAN refuse to let these archival activities occur,
and they CAN issues *take down notices* which MUST *by Law* be honored.
You may have also noticed this when using some group you located
somewhere in Usenet, attempting to post and never seeing your message
appear or if it appears [locally] never receiving a response... the
reason being there is no longer any MASTER to send the message to or
receive messages from...
Usenet CAN make that appear to NOT occur *IF* there are sufficient
numbers of SHARING among the services and *IF* other services then
mirror those postings/forums/messagings, but even then a NEW master has
essentially been created, that being wherever the originals come from
and are now collated.
Google and the like can NOW also make that appear to occur.... post
something and you MAY get a response from a Google user in its *Google
groups* archival services or from some secondary source USING Google to
receive its newsgroups... yet nothing has changed, Google is now the
MASTER..
Paul Tomblin
>The following are segments from a recent usenet post. They purport to
>explain something about how usenet messages are propagated, where the
>control points lie, etc.
>
>I would like to hear comments as to why or how these comments and
>explanations are either correct or incorrect.
Everything from the words "There are" on is 100% wrong.
--
Paul Tomblin <ptom...@xcski.com> http://blog.xcski.com/
"Man in the tower, this is the man in the bird, I'm ready to go, so give me
the word." "Man in the bird, this is the man in the tower, you sound funny,
delay's an hour." - Rod Machado
Russ Allbery
> There ARE master servers... whether you choose to call HOW messages get
> transferred across the world MIRRORING or propagation or peering or
> something else, USENET REQUIRES there be some place wherein information
> / postings can be collected, checked for duplicates, clarified, and
> otherwise prepared.
The English of this and the rest of the message that you quoted is
ambiguous and ill-defined, but this part certainly looks to be completely
incorrect and the incorrect assumption then infects the rest of the
document.
Any given posting to Usenet does originate at one server (well, one can
multiply inject postings, but almost no one does). However, there is no
concept of a "master" server in the sense that appears to be meant here.
Each server separately originates postings, and each server separately
collects, checks for duplicates, and presents all of the messages that it
receives from all of its peers. Usenet news servers generally have
sufficient peers (over 20 is typical) that the loss of any single peer is
completely unnoticable apart from postings originating at that peer.
The one partial exception is for moderated groups, where all postings
travel via e-mail to the moderator until the moderator approves them and
then all originate at the server or servers that the moderator uses to
approve and post submissions.
--
Russ Allbery (r...@stanford.edu) <http://www.eyrie.org/~eagle/>
News Guy
> The English of this and the rest of the message that you quoted is
> ambiguous and ill-defined, but this part certainly looks to be
> completely incorrect and the incorrect assumption then infects
> the rest of the document.
Just to give you some background.
Those comments were intended to explain the microsoft.public.* hierarchy
of newsgroups, and that microsoft has some "special" relationship with
them by virture that they were (presumably) originally created by
microsoft and that microsoft continues to play some sort of central role
in administering them or controlling, distributing the posts that are
posted to them to and from all NNTP servers that carry them.
Many people read and post to the microsoft.* hierarchy of groups
directly through the microsoft server, usually via it's http interface.
I'm speculating that most of those people do not know they are
interacting with usenet by doing so - if they even know what usenet is.
There is a small sub-set of people that do know what usenet is, but for
some reason they don't believe that their interaction with the
microsoft.* set of groups is (essentially) the same as reading / posting
to usenet, or they don't want to believe it because of a percieved sense
of inferiority they have with usenet.
Then there are others that know that the messages posted via the
microsoft server are somehow "leaked" to usenet, and vice-versa, and
they have strange ideas that microsoft does, in some way, have some
special authoritative or control relationship to the microsoft.* usenet
groups, a relationship that somehow makes their server a "master" server
or service.
The material that I posted in the OP was their attempt to explain that
relationship in terms of a "master" service, mirrors, and in general
some sort of colation or compilation and distribution of posts. That
is, that the microsoft server is the "primary" server for all posts to
the microsoft.* set of groups, and that all other nntp servers around
the world are "secondary" servers for those groups. The implication
being that if microsoft decided to deactivate or decouple their servers
with the rest of usenet, that the microsoft.* set of groups would cease
to function on the "world-wide" usenet.
I was wondering if there was a concise, precise, and brief way to
explain how that argument is just plain wrong.
Paul Tomblin
>being that if microsoft decided to deactivate or decouple their servers
>with the rest of usenet, that the microsoft.* set of groups would cease
>to function on the "world-wide" usenet.
No, if Microsoft cut their servers off from INN, then posts to the
microsoft.* newsgroups made to other servers will still trasmit to those
other server's peers and so on. Depending how well connected the servers
that were posted to were, those posts could very well end up at every
other usenet server in the world *except* Microsoft's. And in that
situation, you might find two disjoint sets of conversations going on, one
of the Microsoft servers and another on the non-Microsoft servers.
Microsoft can't control the spread of the microsoft.* newsgroups, because
there is no such thing as a "master" server.
--
Paul Tomblin <ptom...@xcski.com> http://blog.xcski.com/
"It's free they say, if you can get it to run, the geeks say `hey, that's half
the fun', but I've got a girlfriend and things to get done, the Linux OS sucks"
- Three Dead Trolls In A Baggie, "Every OS Sucks" http://www.deadtroll.com/
Tim Slattery
>Russ Allbery wrote:
>
>> The English of this and the rest of the message that you quoted is
>> ambiguous and ill-defined, but this part certainly looks to be
>> completely incorrect and the incorrect assumption then infects
>> the rest of the document.
>
>Just to give you some background.
>
>Those comments were intended to explain the microsoft.public.* hierarchy
>of newsgroups, and that microsoft has some "special" relationship with
>them by virture that they were (presumably) originally created by
>microsoft and that microsoft continues to play some sort of central role
>in administering them or controlling, distributing the posts that are
>posted to them to and from all NNTP servers that carry them.
But even that's not true. Yes, Microsoft owns a server that hosts its
microsoft.public.* hierarchy and they exert control over what appears
on that server. But zillions of NNTP servers independently host those
groups, and trade posts with each other and with the Microsoft
servers. I expect that there are many posts on the non-MS servers that
the MS servers screen out. MS can control only what's on their own
servers, not everybody else's.
--
Tim Slattery
Slatt...@bls.gov
http://members.cox.net/slatteryt
News Guy
the microsoft.* hierarchy are not really true usenet groups, and that
posts made to those groups (regardless where the injection point is) are
handled by Microsoft's "master" servers for redistribution to all other
servers - which are "secondary" servers according to his explanation.
PS: he's not a lawyer, but he thinks and writes like one, as you might
be able to tell as you read this:
----------------------------------
Just for an FYI, as previously noted within this group, I was part of
the originating of this Usenet aspect and I was part of the mirroring of
this Usenet indexing and distribution during those initial years of
establishing and expanding that "network" [that's before the Internet
and for a time afterwards]. That it now exists, for the most part, ON
the Internet, hasn't changed much in reality, its still a networking
system which relies upon parties [news servers if that's what helps
understand] and users to participate within it. If the businesses,
corporations, and other remove their support or no longer offer Usenet
access, it becomes again what it once was, just another offering for
discussion usage.
It was far more useful prior to the entry of Search Engines and their
locational abilities and archiving into the Internet, as they now
provide the indexing and locational abilities, and many now supply their
own versions of "Usenet" or more appropriately social networks without
the need for Usenet, save for historical purposes and to offer those
groups *distinct* within that network. Most groups and forums, however,
are not in Usenet, and the spiders and other now used by the Search
Engines are far more effective at locating and offering the groups for
"discovery" that a party might wish to locate and use.
Moreover, just as this forum has noticed decreases in usage not just
do to lack of OS use, "Usenet" [per se] has also suffered similar losses
within its own Networking system. As Twitter, Facebook, and the like
[distinct and separate social networks] expand their impact, Usenet will
also lose some or potentially all of its users and value. When "Cloud"
and other new aspects of Internet usage become the standard and fully
implemented, Usenet and many other present external networking features
and offerings will become unnecessary and likely fade away or become
somewhat niche like in nature
John F. Morse
Microsoft itself is just plain wrong.
That is not my opinion, but a well-known fact.
I don't have the time nor desire to teach you why, but if you want to
investigate it for yourself, be my guest.
There is Google, and there are many official court judgment transcripts
that are available, which label Microsoft as a liar and thief.
Wasting everyone's time to rehash this to try and counter Microsoft's
FUD is beyond my understanding.
Just say "no."
BTW, I do not carry any Microsoft group. They can carry them on their
server(s), and pay the electric bill themselves. Windows users are free
to connect there for their disinformation.
--
John
News Guy
> Microsoft itself is just plain wrong.
Huh?
> There is Google, and there are many official court judgment
> transcripts that are available, which label Microsoft as a
> liar and thief.
I don't think we were talking about Microsoft's corporate culture and
history here.
> Wasting everyone's time to rehash this to try and counter
> Microsoft's FUD is beyond my understanding.
Rehash? Was this subject already hashed once before?
Has Microsoft actually published any FUD regarding the microsoft.*
usenet hierarchy?
Russ Allbery
> The following is more material posted by the person that believes that
> the microsoft.* hierarchy are not really true usenet groups, and that
> posts made to those groups (regardless where the injection point is) are
> handled by Microsoft's "master" servers for redistribution to all other
> servers - which are "secondary" servers according to his explanation.
Well, I'm rather dubious about those bits, but the content of *this*
message that you quoted I completely agree with.
>> Moreover, just as this forum has noticed decreases in usage not just
>> do to lack of OS use, "Usenet" [per se] has also suffered similar
>> losses within its own Networking system. As Twitter, Facebook, and the
>> like [distinct and separate social networks] expand their impact,
>> Usenet will also lose some or potentially all of its users and
>> value. When "Cloud" and other new aspects of Internet usage become the
>> standard and fully implemented, Usenet and many other present external
>> networking features and offerings will become unnecessary and likely
>> fade away or become somewhat niche like in nature
And indeed this has already happened.
News Guy
> > The following is more material posted by the person that believes
> > that the microsoft.* hierarchy are not really true usenet groups,
>
> Well, I'm rather dubious about those bits, but the content of *this*
> message that you quoted I completely agree with.
You quoted the third paragraph of that message.
I'm more interested in what you think of the first paragraph.
Russ Allbery
> You quoted the third paragraph of that message.
>
> I'm more interested in what you think of the first paragraph.
Since I have no idea who the person is whose words you're requoting, I
can't comment on the claims about their background. The rest of the first
paragraph looked accurate to me.
News Guy
> > I'm more interested in what you think of the first paragraph.
>
> Since I have no idea who the person is whose words you're
> requoting, I can't comment on the claims about their background.
> The rest of the first paragraph looked accurate to me.
The author of the first paragraph, reproduced below, was Maurice Edward
Brahier. Can you (or anyone else) now make any comments regarding his
claim?
==================
Russ Allbery
> The author of the first paragraph, reproduced below, was Maurice Edward
> Brahier. Can you (or anyone else) now make any comments regarding his
> claim?
I haven't heard of him, but that doesn't mean anything. There were a lot
of people at Duke and UNC who could have been involved in the early days,
and even more once more UUCP sites got involved. I don't know most of the
names of the admins of early UUCP sites off the top of my head. He does
seem to have been around for a while.