Can I get a reality-check on the following comments?
News Guy
--------------------
MS has several different categories of newsgroups, and operates them in
two different ways. Some Usenet groups are identified as 'managed'.
This means that MS staff participate in those groups, but, as the name
implies, some sort of additional 'management' might also be going on.
It appears that MS staff participation in these groups is being wound
down, but this might only be because they will only deal with people
they can reliably identify as qualifying for elevated support, and they
can only do this reliably through the sign-on function provided in the
http forums. Other Usenet groups are, as far as I can tell, neither
monitored not censored, although all groups have spam filters and
messages can be deleted by special request. Of course, deleted messages
might have escaped into the network before deletion.
Then there are the http forums. The rules here are completely
different, subject to frequent change, and not very well advertised. I
wouldn't be at all surprised that some posts to a Usenet group don't
make it to the
corresponding forum, and the reverse is quite possible. I would also
expect that the forum postings are monitored, although it can't be in
real time as they appear very promptly after posting.. Perhaps the
monitoring occurs after they appear in the forum but before they make it
to Usenet. We have seen cases of failures in the connection that put the
two sets of messages out of sync. In some cases this is repaired with a
batch synchronization, and in other cases just left to re-align by
themselves.
Monitoring and censorship in Usenet is wider than you seem to think.
There are the profanity filters already mentioned. Many servers limit
the size of postings to non-binary groups. Many servers check for
things like copyright breaches in specialised groups where it's likely,
plus there's the checking that happens for sensitive material (eg,
terrorist-related) that the server operators see as exposing them to
legal issues.
Pete Ashdown
>Monitoring and censorship in Usenet is wider than you seem to think.
>There are the profanity filters already mentioned. Many servers limit
>the size of postings to non-binary groups. Many servers check for
>things like copyright breaches in specialised groups where it's likely,
>plus there's the checking that happens for sensitive material (eg,
>terrorist-related) that the server operators see as exposing them to
>legal issues.
My server limits all postings to 128K because I'm only interested in text
discussion. If you need more than 128K to get your message across, you should
probably be writing books and not squandering your talents on Usenet. That
isn't censorship as much as it is a technical limit. Binaries are a PITA and
require a huge amount of resources to effectively handle. I don't have any
profanity filters, but I do have filters that specifically target spammers.
--
Pete Ashdown pash...@xmission.com http://pashdown.org Salt Lake City, Utah
XMission Internet Access - http://xmission.com - Voice: 801 539 0852
News Guy
> > Monitoring and censorship in Usenet is wider than you seem to
> > think.
> > There are the profanity filters already mentioned.
> > Many servers limit the size of postings to non-binary groups.
> > Many servers check for things like copyright breaches
> > There's the checking that happens for sensitive material
> > (eg, terrorist-related) that the server operators see as
> > exposing them to legal issues.
>
> My server limits all postings to 128K because ...
Regarding size limitations, I was under the impression that most text
servers limit posts to 19kb or less. But post size is somewhat
irrelavent to the theme of this thread.
The point I was exploring was this notion that usenet is (or is not) a
censored or even a censorable communications channel.
Specifically, it has been pointed out (in another thread in another
newsgroup) that specifically the Microsoft NNTP servers and HTTP
interface to those groups seem to experience some degree of active
censorship, and that one might want to experience the microsoft.*
newsgroups without the hindrence of post removal due to censorship. The
rejoinder being that "censorship in Usenet is wider than you seem to
think" (ie there is little difference between usenet servers in this
regard).
The examples being offered are:
- profanity filters
- copyright checking
- checking for criminal or terrorism-related material
I'm unaware of the extent that any of the above are implimented on any
significant number of NNTP servers (be they institutional, corporate,
ISP, or hobby / complimentary servers). I don't consider filtering
based on size as being censorship.
The idea of censoring is more significant if a human being (acting as
NNTP or IT admin) has the power / ability / authority / responsibility
to delete a post on his/her server after reading and judging it -
regardless if any third party brings the post to their attention.
The idea of an automated program deleting (or not accepting) posts that
contain certain keywords is slightly more palatible (from a free-speech
POV) but it does still smack of censorship.
I will repeat the first few paragraphs of my previous post, because I am
interested to know if they contain factual information or is more
speculation or opinion:
====================
MS has several different categories of newsgroups, and operates them in
two different ways. Some Usenet groups are identified as 'managed'.
This means that MS staff participate in those groups, but, as the name
implies, some sort of additional 'management' might also be going on.
It appears that MS staff participation in these groups is being wound
down, but this might only be because they will only deal with people
they can reliably identify as qualifying for elevated support, and they
can only do this reliably through the sign-on function provided in the
http forums. Other Usenet groups are, as far as I can tell, neither
monitored not censored, although all groups have spam filters and
messages can be deleted by special request. Of course, deleted messages
might have escaped into the network before deletion.
Then there are the http forums. The rules here are completely
different, subject to frequent change, and not very well advertised. I
wouldn't be at all surprised that some posts to a usenet group don't
make it to the corresponding forum, and the reverse is quite possible. I
would also expect that the forum postings are monitored, although it
can't be in real time as they appear very promptly after posting.
D. Stussy
news:h6f6q7$1rg$1...@news.xmission.com...
> News Guy <Ne...@Guy.com> writes:
>
> >Monitoring and censorship in Usenet is wider than you seem to think.
> >There are the profanity filters already mentioned. Many servers limit
> >the size of postings to non-binary groups. Many servers check for
> >things like copyright breaches in specialised groups where it's likely,
> >plus there's the checking that happens for sensitive material (eg,
> >terrorist-related) that the server operators see as exposing them to
> >legal issues.
>
> My server limits all postings to 128K because I'm only interested in text
> discussion. If you need more than 128K to get your message across, you
should
> probably be writing books and not squandering your talents on Usenet.
That
> isn't censorship as much as it is a technical limit. Binaries are a PITA
and
> require a huge amount of resources to effectively handle. I don't have
any
> profanity filters, but I do have filters that specifically target
spammers.
Note that there are some checkgroups control messages that exceed 128k in
size. However, they are smaller than 256k.
(The "microsoft" and "big-8" hierarchies have the oversized checkgroups
messages)
Sunny
"News Guy" <Ne...@Guy.com> wrote in message
news:4A8B32CC...@Guy.com...
<snip>
> Specifically, it has been pointed out (in another thread in another
> newsgroup) that specifically the Microsoft NNTP servers and HTTP
> interface to those groups seem to experience some degree of active
> censorship, and that one might want to experience the microsoft.*
> newsgroups without the hindrence of post removal due to censorship. The
> rejoinder being that "censorship in Usenet is wider than you seem to
> think" (ie there is little difference between usenet servers in this
> regard).
I was/am one of the participants in this thread (in <
microsoft.public.win98.gen.discussion >
The original thread subject was about win 98 being abandonware, this was
changed to " USENET verses Microsoft forums".
Despite four posters supplying references, one poster (who dislikes Usenet
and "Usenetters") keeps insisting that;
a. Microsoft controls all posts to Usenet on their NNTP Servers,
b. Microsoft is a "Master" Usenet server and has signed agreements with
every ISP News server,
c. Free news servers (AIOE was mentioned) are only available to those who
have no access to news servers through their ISP, and
d. Is convinced that all Microsoft public NGs are monitored for profanity
and content.
Julien ÉLIE
> Note that there are some checkgroups control messages that exceed 128k in
> size. However, they are smaller than 256k.
>
> (The "microsoft" and "big-8" hierarchies have the oversized checkgroups
> messages)
And also the largest one (190k), which is for "fido7".
--
Julien �LIE
� Cette obscure clart� qui tombe des �toiles. � (Pierre Corneille)
Tim Slattery
>Despite four posters supplying references, one poster (who dislikes Usenet
>and "Usenetters") keeps insisting that;
>a. Microsoft controls all posts to Usenet on their NNTP Servers,
Well, they would certainly control posts made directly to their
servers, as well as what comes to those servers via peering
agreements. But *all* posts? Nohow.
>b. Microsoft is a "Master" Usenet server and has signed agreements with
>every ISP News server,
They are master only of their own server. They would have peering
agreements with some others, nothing else that I know of. Of course, I
wouldn't know.
>c. Free news servers (AIOE was mentioned) are only available to those who
>have no access to news servers through their ISP, and
Ridiculous.
>d. Is convinced that all Microsoft public NGs are monitored for profanity
>and content.
They do have filters in place. They screen for spam, probably for
profanity. Not sure about "content", I guess it depends on what
content. There are a large number of extremely flaky posts in
ms.public.vista.general, for example. (IS evangelism, personal
attacks, etc). But they don't make details of those filters public.
--
Tim Slattery
Slatt...@bls.gov
http://members.cox.net/slatteryt
Charles Lindsey
>The point I was exploring was this notion that usenet is (or is not) a
>censored or even a censorable communications channel.
>Specifically, it has been pointed out (in another thread in another
>newsgroup) that specifically the Microsoft NNTP servers and HTTP
>interface to those groups seem to experience some degree of active
>censorship, and that one might want to experience the microsoft.*
>newsgroups without the hindrence of post removal due to censorship. The
>rejoinder being that "censorship in Usenet is wider than you seem to
>think" (ie there is little difference between usenet servers in this
>regard).
Yes, but you need to understand that rules vary between the various
different hiererchies. Some hierarchies started out as private
hierarchies, local to some site, or locally interconnected group of sites.
But then they "leaked" out onto the main Usenet, and became stored on many
publicly accessible servers. It is questionable whether such hierarchies
are then "public" or "private".
Microsoft.* is one such hierarchy, and Microsoft can still claim (well,
they would, wouldn't they) that it is a private hierarchy which they have
a right to control, and that its availability on the wider Usenet is
simply a "service provided by them to the wider public".
The same is, in principle, true of many other hierarchies devoted to the
products of particular vendors (opera.*, for example).
Ideally, those vendor-specific hierarchies ought to have been
sub-hierarchies of biz.* (which has, or used to have, some degree of
control exercised by Ed Hew). But that does not seem to have worked out as
intended :-( .
--
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Web: http://www.cs.man.ac.uk/~chl
Email: c...@clerew.man.ac.uk Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5
Pete Ashdown
>> My server limits all postings to 128K because I'm only interested in text
>> discussion. If you need more than 128K to get your message across, you
>Note that there are some checkgroups control messages that exceed 128k in
>size. However, they are smaller than 256k.
xmission\
:!*,control.checkgroups:AC,Tm,<262144:innfeed!
News Guy
> Yes, but you need to understand that rules vary between the
> various different hiererchies.
If I understand this correctly, then there are no rules outside of the
"big-8" hirearchies.
And we are talking about a newsgroup (or set of groups) that are not
part of the big-8.
> Some hierarchies started out as private hierarchies, local to some
> site, or locally interconnected group of sites. But then they
> "leaked" out onto the main Usenet,
I question your use of the word "leaked", as if it was unplanned,
unwanted, or uncontrolled. An act of nature. Random.
I don't know the history of the microsoft.* hierarchy - ie - did it
exist for some time only on microsoft servers before microsoft decided
to peer it's servers with the outside world?
> and became stored on many publicly accessible servers.
> It is questionable whether such hierarchies are then "public" or
> "private".
If any entity decides to operate their own, isolated, NNTP server, and
create a set of groups on that server for others to access, then there
is no question that is a "private" server, hosting private groups. It
may be publically accessible, or it may require user login.
If the operator of that server seeks out and establishes a peering
arrangement with one (or more) usenet operators, then how can there be
any question that the notion of "public" or "private" server or
newsgroup has been effectively erased? If you peer your once private
server with the external usenet, how can you possibly claim that your
server (or at least your groups) are anything but public?
> Microsoft.* is one such hierarchy, and Microsoft can still claim
> (well, they would, wouldn't they) that it is a private hierarchy
> which they have a right to control, and that its availability on
> the wider Usenet is simply a "service provided by them to the
> wider public".
Microsoft can only claim that their server is private and offered as a
service. In the context of the world-wide usenet, the groups themselves
are not private or their's to control. I don't know how you can claim
the groups themselves are still tied to microsoft in some functional,
operational way.
I can make a post to microsoft.public.something via the aioe.org news
server, and you can read and reply to that post on the acme.com news
server, and that entire transaction did not pass through or was
influenced or moderated by any person or machine employed or operated by
microsoft.
> The same is, in principle, true of many other hierarchies devoted
> to the products of particular vendors (opera.*, for example).
How would you describe the operational or functional link that those
hierarchies have with their corporate name-sakes?
Are you saying that if Microsoft ended their peering with the rest of
usenet or "pulled the plug" on their nntp server, that the microsoft
hierarcy would simply disappear from all nntp servers world-wide? Or
that microsoft could compel all nntp server operators to remove the
microsoft hirearchy from their servers?
Adam H. Kerman
>Charles Lindsey wrote:
>>Yes, but you need to understand that rules vary between the
>>various different hiererchies.
>If I understand this correctly, then there are no rules outside of the
>"big-8" hirearchies.
>And we are talking about a newsgroup (or set of groups) that are not
>part of the big-8.
You don't understand it correctly. You didn't understand it correctly
earlier. You won't understand it correctly tomorrow.
Usenet is distributed. Each server makes its own rules. Some newsgroups
may exist only on one server or within one network. These aren't Usenet
newsgroups. With regard to institutional newsgroups that exist only on
the News server within the institution's network, then whatever rules
apply on that server are the rules that apply in the hierarchy which
happen to be one and the same.
That's not exactly the case with Microsoft.
>>Some hierarchies started out as private hierarchies, local to some
>>site, or locally interconnected group of sites. But then they
>>"leaked" out onto the main Usenet,
>I question your use of the word "leaked", as if it was unplanned,
>unwanted, or uncontrolled. An act of nature. Random.
Didn't I just tell you this last week? Microsoft distributes the
newsgroups within a select number of News servers. One of these News
servers distributes messages beyond this small network making them
available to any Usenet News server.
It's not Microsoft doing it but it's not leaked, I suppose, as it's done
deliberately with Microsoft's toleration.
>If any entity decides to operate their own, isolated, NNTP server, and
>create a set of groups on that server for others to access, then there
>is no question that is a "private" server, hosting private groups. It
>may be publically accessible, or it may require user login.
That wouldn't be NNTP then.
>If the operator of that server seeks out and establishes a peering
>arrangement with one (or more) usenet operators, then how can there be
>any question that the notion of "public" or "private" server or
>newsgroup has been effectively erased?
It's quite simple. You make an agreement amongst yourself to create a
small network and not to distribute messages beyond the network.
>Microsoft can only claim that their server is private and offered as a
>service.
A private server is a server that supports a specific set of users.
Microsoft's server is public. Anyone in the community may use it.
News Guy
> You don't understand it correctly. You didn't understand it
> correctly earlier. You won't understand it correctly tomorrow.
>
> Usenet is distributed.
I never said it wasn't.
> Each server makes its own rules.
I never said they didn't.
> Some newsgroups may exist only on one server or within one
> network. These aren't Usenet newsgroups.
I never mentioned or intended to discuss those types of groups.
> With regard to institutional newsgroups that exist only on
> the News server within the institution's network,
> then whatever rules apply on that server are the rules
> that apply in the hierarchy which happen to be one and
> the same.
So far you haven't said anything that I argued against or was even
talking about.
> That's not exactly the case with Microsoft.
And what - you're just going to leave that statement as-is? Without
saying why that's "not not exactly the case with Microsoft" ?
> > I question your use of the word "leaked", as if it was
> > unplanned, unwanted, or uncontrolled. An act of nature.
> > Random.
>
> Didn't I just tell you this last week? Microsoft distributes
> the newsgroups within a select number of News servers.
Are you saying that microsoft acts as some sort of central clearing
house for all posts made to the microsoft.* set of newsgroups?
Are you saying that all posts made to the microsoft.* set of groups,
regardless of where those posts were injected, are flowed directly to
microsoft where they are (possibly or potentially) moderated, filtered,
censored, and then re-distributed back to the world-wide usenet?
> One of these News servers distributes messages beyond this
> small network making them available to any Usenet News server.
Are you saying that the flow of posts made to any of the microsoft.*
groups (regardless which server the poster was actually in contact with)
are ALWAYS routed through microsoft's servers?
And that this is fundamentally different than posts made to most any
other non-moderated usenet newsgroup?
> It's not Microsoft doing it but it's not leaked, I suppose, as
> it's done deliberately with Microsoft's toleration.
And you can't, or won't explain why posts made to the microsoft.* groups
are handled in a radically different manner than posts to any other
usenet newsgroup.
> > If any entity decides to operate their own, isolated, NNTP
> > server, and create a set of groups on that server for others
> > to access, then there is no question that is a "private"
> > server, hosting private groups. It may be publically
> > accessible, or it may require user login.
>
> That wouldn't be NNTP then.
An NNTP server can operate in isolation or disconnected from usenet.
Usenet client software can still function to read and post messages to
that server. There may be no "transport" happening between that server
and any other server, but it's still an NNTP server from the point of
view of the clients interacting with it.
> > If the operator of that server seeks out and establishes a
> > peering arrangement with one (or more) usenet operators,
> > then how can there be any question that the notion of
> > "public" or "private" server or newsgroup has been
> > effectively erased?
>
> It's quite simple. You make an agreement amongst yourself to
> create a small network and not to distribute messages beyond
> the network.
That is not the case with Microsoft or the set of microsoft.* groups,
unless you define a "small network" as a network confined to planet
earth.
> > Microsoft can only claim that their server is private and
> > offered as a service.
>
> A private server is a server that supports a specific set of
> users. Microsoft's server is public. Anyone in the community
> may use it.
Charles Lindsey stated that Microsoft can (or does) claim that the set
of microsoft.* groups is a private hierarchy which they have a right to
control, and that its availability on the wider Usenet is simply a
"service provided by them to the wider public".
I countered by saying no - only Microsoft's server machines could be
considered private, and that the groups themselves are not private by
any sense of the word - certainly not from a control or distribution
point of view.
Charles Lindsey
>Charles Lindsey wrote:
>> Yes, but you need to understand that rules vary between the
>> various different hiererchies.
>If I understand this correctly, then there are no rules outside of the
>"big-8" hirearchies.
There are many hierarcies (uk.* for example) which have rules every bit as
complete and consistent as those governing the Big-8 (see
www.usenet.org.uk for that particular case). The vast majority of Usenet
sites worldwide acknowledge and enforce those rules.
Then there are hierarchies (alt.* and free.* for example) where the rules
are either minimal or non-existent, and Usenet sites worldwide accept or
ignore those groups as they see fit.
And then there are hierachies at every stage in between, with varying
degrees of uniformity in the manner in which Usenet sites worldwide
recognize them.
>I question your use of the word "leaked", as if it was unplanned,
>unwanted, or uncontrolled. An act of nature. Random.
You may question it, but that is exactly how it happens. AIUI (though I
may be wrong) that is more or less what happened to ms.*.
>I don't know the history of the microsoft.* hierarchy - ie - did it
>exist for some time only on microsoft servers before microsoft decided
>to peer it's servers with the outside world?
AIUI, that is how it started. Whether its subsequent leak was due to a
deliberate act by Microsoft, or to the actions of some maverick who simply
grabbed it and propagated it, I would not know.
>If the operator of that server seeks out and establishes a peering
>arrangement with one (or more) usenet operators, then how can there be
>any question that the notion of "public" or "private" server or
>newsgroup has been effectively erased? If you peer your once private
>server with the external usenet, how can you possibly claim that your
>server (or at least your groups) are anything but public?
Quite simple. You simply stand up and shout for all the world to hear that
your server/hierarchy/whatever is "private", and that you have the right
to control it.
That is the easy bit. The hard bit, of course, is to persuade the rest of
the world to believe you (an art which Microsoft seems to have mastered in
many other fields, if not in this one :-) ).
>Microsoft can only claim that their server is private and offered as a
>service. In the context of the world-wide usenet, the groups themselves
>are not private or their's to control. I don't know how you can claim
>the groups themselves are still tied to microsoft in some functional,
>operational way.
I don't claim it (who cares what I claim?). But that does not prevent
Microsoft from claiming it it if suits their Corporate Image to do so.
>I can make a post to microsoft.public.something via the aioe.org news
>server, and you can read and reply to that post on the acme.com news
>server, and that entire transaction did not pass through or was
>influenced or moderated by any person or machine employed or operated by
>microsoft.
That is true, and as a consequence the microsoft hierarchy may become
(indeed propably has become) fragmented, with different servers carrying
slightly different sets or articles. But nothing like the degree of
fragmentation that you see with alt.* and free.*.
>> The same is, in principle, true of many other hierarchies devoted
>> to the products of particular vendors (opera.*, for example).
>How would you describe the operational or functional link that those
>hierarchies have with their corporate name-sakes?
The link is ill-defined, but usually works quite well if people don't try
to muck about with it. For example, I am not aware that opera.* is at all
fragmented - it seems to be spam free (at least from the server I take it
from). It serves a useful purpose for Opera to communicate with its User
Community - in short, it ain't broke and is not in need of fixing.
>Are you saying that if Microsoft ended their peering with the rest of
>usenet or "pulled the plug" on their nntp server, that the microsoft
>hierarcy would simply disappear from all nntp servers world-wide? Or
>that microsoft could compel all nntp server operators to remove the
>microsoft hirearchy from their servers?
No, I am not saying any of those things.
Adam H. Kerman
>News Guy <Ne...@Guy.com> writes:
>>Charles Lindsey wrote:
>>>Yes, but you need to understand that rules vary between the
>>>various different hiererchies.
>>If I understand this correctly, then there are no rules outside of the
>>"big-8" hirearchies.
>There are many hierarcies (uk.* for example) which have rules every bit as
>complete and consistent as those governing the Big-8 (see
>www.usenet.org.uk for that particular case). The vast majority of Usenet
>sites worldwide acknowledge and enforce those rules.
Rules on what that are acknowledged and enforced at the vast majority of
Usenet sites? I thought you were talking about institutional hierarchies
on a single News server, but it's apparent that you're talking about
something else.
>Then there are hierarchies (alt.* and free.* for example) where the rules
>are either minimal or non-existent, and Usenet sites worldwide accept or
>ignore those groups as they see fit.
Rules for what?
>>I question your use of the word "leaked", as if it was unplanned,
>>unwanted, or uncontrolled. An act of nature. Random.
>You may question it, but that is exactly how it happens. AIUI (though I
>may be wrong) that is more or less what happened to ms.*.
I think that after 10 or 15 years, it can be assumed that Microsoft is
well aware of what's happening and therefore, there is no leaking taking
place.
Suppose a set of servers agree to exchange a certain group of articles.
They each agree upon a common Distribution. So articles with a
Distribution are sent only to sites that recognize that Distribution and
have created the newsgroup on the Newsgroups header.
Leaking occurs when somebody isn't doing what he agreed to do. If the
Distribution is misspelled it's not a restriction because it won't be
recognized. One of the servers may be sending articles to servers that
don't recognize the Distribution so there won't be an effect.
If all articles posted to a specific newsgroup are supposed to have a
restricted Distribution and one of the servers accepts an article
without a Distribution (or without adding the header), then the article
is sent worldwide.
As far as I know, there is no leaking of articles posted to microsoft.*
newsgroups from the Microsoft server. I've never noticed that it adds a
Distribution header.
Now, without adding the header, they can all agree to send no article
posted to a specific set of newsgroups beyond the servers they have a
specific arrangement with. Clearly, that agreement doesn't exist.
>>I don't know the history of the microsoft.* hierarchy - ie - did it
>>exist for some time only on microsoft servers before microsoft decided
>>to peer it's servers with the outside world?
>AIUI, that is how it started. Whether its subsequent leak was due to a
>deliberate act by Microsoft, or to the actions of some maverick who simply
>grabbed it and propagated it, I would not know.
Surely if the peer in question wasn't holding up its end of the
agreement, Microsoft would have stopped peering. As that's not the case,
it's only fair to say that Microsoft goes along with it. The only thing
we don't know is why Microsoft refuses to arrange peering itself.
>>If the operator of that server seeks out and establishes a peering
>>arrangement with one (or more) usenet operators, then how can there be
>>any question that the notion of "public" or "private" server or
>>newsgroup has been effectively erased? If you peer your once private
>>server with the external usenet, how can you possibly claim that your
>>server (or at least your groups) are anything but public?
>Quite simple. You simply stand up and shout for all the world to hear that
>your server/hierarchy/whatever is "private", and that you have the right
>to control it.
>That is the easy bit. The hard bit, of course, is to persuade the rest of
>the world to believe you (an art which Microsoft seems to have mastered in
>many other fields, if not in this one :-) ).
Why do you have to inform the world? You control who uses your server
and you decide whether certain articles are sent beyond your server. If
your server isn't offered to the public, it's not a public server. You
don't have to give notice.
News Guy
> > I question your use of the word "leaked", as if it was unplanned,
> > unwanted, or uncontrolled. An act of nature. Random.
>
> You may question it, but that is exactly how it happens.
As Adam Kerman has tried to expain to you, there can never be a
"leakage" of nntp messages out of or into an NNTP server without the
intentional establishment of a peering relationship to at least one
other server.
> AIUI (though I may be wrong) that is more or less what
> happened to ms.*.
And the idea that Microsoft would establish some sort of closed or
limited peering relationship with just one other external (and similarly
usenet-disconnected) NNTP server is absurd. Who would be the operator
of such a server? Sun Microsystems? IBM? Novell? Acme? Is that what
is required of your "leakage" theory?
> AIUI, that is how it started. Whether its subsequent leak
> was due to a deliberate act by Microsoft,
Why you would define a deliberate, intentional peering relationship as a
"leak" is, well, strange. I guess that in your lexicon, peering =
leakage.
> > If you peer your once private server with the external
> > usenet, how can you possibly claim that your server (or
> > at least your groups) are anything but public?
>
> Quite simple. You simply stand up and shout for all the
> world to hear that your server/hierarchy/whatever is
> "private", and that you have the right to control it.
If you want your set of newsgroups to be "private", then what are you
doing establishing a peering relationship such that the contents of
those groups are propagated to the rest of usenet?
What sense would it make to establish such a peering relationship, and
then "stand up and shout for all world to hear that your groups are
private" ??? What kind of logic is that?
> That is the easy bit. The hard bit, of course, is to persuade
> the rest of the world to believe you
Again, your logic is bizarre. If Microsoft wanted to contain their
hierarchy to their own server, then all they had to do was to NOT
establish any peering relationship with the rest of usenet. Why is that
such a hard concept to understand?
Once a newsgroup exists on the general, world-wide usenet, I'm not aware
of any argument that an "entity" can make to remove the group on the
basis that it simply does not want the group to exist.
> > I can make a post to microsoft.public.something (...) and that
> > entire transaction did not pass through (...) any person or
> > machine employed or operated by microsoft.
>
> That is true, and as a consequence the microsoft hierarchy may
> become (indeed propably has become) fragmented,
Please explain what you mean by "fragmented". Are you saying that a
survey of many different NNTP servers that carry the microsoft.* groups
might show (or would show?) that they do not all contain the same set of
postings?
That is true if only because of various automated spam-removal systems
in place on various servers, but beyond that I highly doubt that "legit"
messages are at least not making it to every corner of usenet, where
some individual servers may choose to delete them upon receipt for one
reason or another.
Are you saying that the reason for the lack of uniformity is that
something has "broken" with regard to the operation of the ms groups?
Are you saying that the normal or intended function of microsoft and
their servers is to act as a hub or central node for those posts, and
that all NNTP servers world-wide have a means to "obey" and operate with
microsoft's servers in that regard?
> > How would you describe the operational or functional link that
> > those hierarchies have with their corporate name-sakes?
>
> The link is ill-defined, but usually works quite well if people don't
> try to muck about with it.
Your statement implies that you know that such links exist, and that you
know how they work. Therefore, please describe what you know about them
and how they work.
> > Are you saying that if Microsoft ended their peering with the rest
> > of usenet or "pulled the plug" on their nntp server, that the
> > microsoft hierarcy would simply disappear from all nntp servers
> > world-wide? Or that microsoft could compel all nntp server
> > operators to remove the microsoft hirearchy from their servers?
>
> No, I am not saying any of those things.
Then please explain what you know about the relationship between the
microsoft.* usegroup hierarchy and the corporate entity known as
Microsoft as it pertains to the propagation and circulation of posts to
those microsoft.* newsgroups.
Adam H. Kerman
>Charles Lindsey wrote:
>>>I question your use of the word "leaked", as if it was unplanned,
>>>unwanted, or uncontrolled. An act of nature. Random.
>> You may question it, but that is exactly how it happens.
>As Adam Kerman has tried to expain to you, there can never be a
>"leakage" of nntp messages out of or into an NNTP server without the
>intentional establishment of a peering relationship to at least one
>other server.
Aargh. I didn't say that.
If two servers exchange messages in a specific newsgroup agree not to
send those messages to any other server, but of the servers accidentally
send the messages to other servers, that's a leak.
The servers Microsoft peers with do not appear to have an agreement not
to send messages to other servers that Microsoft doesn't peer with, so
there can't be any leaking taking place.
Bruce Esquibel
> Again, your logic is bizarre. If Microsoft wanted to contain their
> hierarchy to their own server, then all they had to do was to NOT
> establish any peering relationship with the rest of usenet. Why is that
> such a hard concept to understand?
> Once a newsgroup exists on the general, world-wide usenet, I'm not aware
> of any argument that an "entity" can make to remove the group on the
> basis that it simply does not want the group to exist.
I don't know the current status of the microsoft groups but many moons ago
when they first started to circulate, they were NOT being "fed" into usenet
using peering.
From what I remember the microsoft news servers were open for reading (not
sure on posting) and some people were grabbing the articles using shell
scripts and programs like suck then injecting them into usenet.
There was no peering agreement with MS and anyone else, at that time anyway,
yet nearly all their groups appeared to be part of usenet.
This was a problem because from this side of things (the usenet side) they
all appeared to be legit groups since if two usenet systems carried like
microsoft.support.xp, if either system posted to it, the article would show
up on both servers, but the post never ended up on microsofts servers.
So from the usenet side you would see posts in a microsoft.* group coming
from both usenet and microsoft but if you used the microsoft server, only
posts coming from them were in the group.
Another example of this was the gmane groups.
I also think this is the reason google doesn't really run an nntp server,
they want the google.* groups to be only available internally via the web
interface so they don't leak out.
I agree with you though, once the group exists on usenet there is little
they can do about it but it doesn't change the intent of the purpose of the
group. If people want to shovel posts from the sanctioned servers into
usenet, they would have to figure a way to stop that practice, not expect
everyone to drop the groups.
-bruce
b...@ripco.com
News Guy
> > As Adam Kerman has tried to expain to you, there can never be a
> > "leakage" of nntp messages out of or into an NNTP server without
> > the intentional establishment of a peering relationship to at
> > least one other server.
>
> Aargh. I didn't say that.
>
> If two servers exchange messages in a specific newsgroup agree
> not to send those messages to any other server,
In this hypothetical case where microsoft is the first server, then who
would have operated the second server? Is there any evidence that
microsoft peered their server with a server operated by a second entity,
with the intention of such a closed or limited sharing as described -
and what would have been the reason for it?
Do we know if Microsoft had a close technical relationship with another
corporation 10 to 20 years ago that would or could have operated this
hypothetical second server?
And if the closed aspect of group distribution was so important to
Microsoft, then wouldn't this "leakage" by the second server be noticed
and corrected fairly quickly, at the risk of Microsoft ending the
relationship with the second server?
> but of the servers accidentally send the messages
> to other servers, that's a leak.
Wouldn't a "group-create" message need to be sent to those "other
servers" in order for the postings to be propagated to them and
elsewhere to the rest of usenet?
> The servers Microsoft peers with do not appear to have an
> agreement not to send messages to other servers that
> Microsoft doesn't peer with, so there can't be any leaking
> taking place.
The premis that Microsoft would peer it's server with a "special"
external server with the intent of a closed distribution is the biggest
flaw in this argument. If Microsoft's intent is that their groups stay
private, then there would have to be an important reason to extend their
groups to this second server. But beyond that, I question that a server
can just start to emit posts to their peers in groups that don't exist
without additional proceedures to establish the new groups on the peers.
Is there no one reading this thread that knows the early history of the
microsoft.* hierarchy on usenet, and Microsoft's role and intentions for
them? Or do I have to post such a question to news.admin.technical?
News Guy
> From what I remember the microsoft news servers were open for
> reading (not sure on posting) and some people were grabbing
> the articles using shell scripts and programs like suck then
> injecting them into usenet.
If posts were being grabbed from the MS server and injected into usenet
by a third party, wouldn't that create a one-way conversation? Because
any replies to those posts on usenet would not be seen on the MS server
- unless a similar injection was done in the reverse direction?
And even if the above was happening, surely Microsoft was in a position
to prevent it from happening if they didn't want it to happen - by using
some sort of secure login into their nntp server -?
> There was no peering agreement with MS and anyone else, at that
> time anyway, yet nearly all their groups appeared to be part of
> usenet.
Perhaps a set of duplicate microsoft newsgroups were created on usenet,
to match the names exactly with the names created by microsoft on their
server? Even if no peering existed between microsoft and the rest of
usenet, those groups could exist and be functional on usenet without
microsoft's permission or involvement.
> This was a problem because from this side of things (the usenet
> side) they all appeared to be legit groups since if two usenet
> systems carried like microsoft.support.xp, if either system
> posted to it, the article would show up on both servers, but
> the post never ended up on microsofts servers.
That's not necessarily a problem. It's Microsoft's choice to peer their
server with usenet, but it's not up to them to control whether or not a
parallel set of groups with exactly the same names appear on usenet. It
then becomes a user choice to decide which microsoft "world" they want
to participate in. Microsoft's closed little world, or the wider usenet
world, or even both (but not simultaneously).
> So from the usenet side you would see posts in a microsoft.*
> group coming from both usenet and microsoft but if you used
> the microsoft server, only posts coming from them were in the
> group.
Assuming the one-way injection was indeed happening as you described
above.
> I agree with you though, once the group exists on usenet there
> is little they can do about it but it doesn't change the intent
> of the purpose of the group. If people want to shovel posts
> from the sanctioned servers into usenet, they would have to
> figure a way to stop that practice, not expect everyone to
> drop the groups.
But it doesn't appear that any such "sanctioned shovelling" is happening
today. It does appear that microsoft does in fact have a working
peering relationship with usenet. But I still haven't heard an
explanation as to why they name their servers along the lines of
TK2MSFTNGP02.phx.gbl.
I think I came across a explanation that phx stood for phoenix.
Clearly, MSFT can be seen in the above FQDN (and I use FQDN loosely).
Some path lines I see have the next machine in the sequence as msrtrans,
followed by msrn-out, followed by a sequence of readnews.com machines,
followed by one or more glorb.com machines.
Adam H. Kerman
>"Adam H. Kerman" wrote:
>>>As Adam Kerman has tried to expain to you, there can never be a
>>>"leakage" of nntp messages out of or into an NNTP server without
>>>the intentional establishment of a peering relationship to at
>>>least one other server.
>>Aargh. I didn't say that.
>>If two servers exchange messages in a specific newsgroup agree
>>not to send those messages to any other server,
>In this hypothetical case . . .
Be so kind as to NOT change my hypothetical. You can damn well write
your own hypothetical.
>Wouldn't a "group-create" message
Do you mean a newgroup message?
>need to be sent to those "other servers" in order for the postings to
>be propagated to them and elsewhere to the rest of usenet?
No.
Some servers file article for groups not created locally in a
pseudonewsgroup called "junk" and that can also be propagated to other
servers. Some servers create groups locally based on the Newsgroups
header of a message that ends up in "junk".
Newgroup messages are an unnecessary part of administering a News server
as group creation policies are entirely local, something you've been
told hundreds of times. All that being said, Microsoft has NEVER sent a
control message, not newgroup, rmgroup, nor checkgroups. Let me repeat
that so you don't rewrite what I just said and pretend to misunderstand:
Microsoft has NEVER sent a control message.
>The premis that Microsoft would peer it's server with a "special"
>external server with the intent of a closed distribution is the biggest
>flaw in this argument.
It's not my premise. It's not anybody's premise. No one has claimed any
of Microsoft's peers are "special", so knock off the straw man crap.
Generally, no one cares what you believe because no matter how many
times you are corrected, you'll continue to make statements based on one
false premise or another.
>Is there no one reading this thread that knows the early history of the
>microsoft.* hierarchy on usenet, and Microsoft's role and intentions for
>them? Or do I have to post such a question to news.admin.technical?
No one will stop you from trolling in another newsgroup, asking silly
questions, then accidentally misunderstanding what answers you were
given because you constantly rewrite what you were told till it loses
its original meaning.
Adam H. Kerman
>I don't know the current status of the microsoft groups but many moons ago
>when they first started to circulate, they were NOT being "fed" into usenet
>using peering.
>From what I remember the microsoft news servers were open for reading (not
>sure on posting) and some people were grabbing the articles using shell
>scripts and programs like suck then injecting them into usenet.
Pardon me for my error and thanks for the correction.
Bruce Esquibel
> If posts were being grabbed from the MS server and injected into usenet
> by a third party, wouldn't that create a one-way conversation? Because
> any replies to those posts on usenet would not be seen on the MS server
> - unless a similar injection was done in the reverse direction?
Yes.
> And even if the above was happening, surely Microsoft was in a position
> to prevent it from happening if they didn't want it to happen - by using
> some sort of secure login into their nntp server -?
Well, off hand I don't see what a "secure login" would have to do with it.
Anyone who had incentive to do it, just did it.
I really don't understand why you are so hyper about this, there are
probably hundreds if not thousands of groups with similar "who knows where
they came from" origins.
In case you are trying to do a historical account, I remember two things that
came up a long time ago when people questioned if they were "legit
newsgroups" or not.
First off, remember that MS was late to the ballgame with the internet. Bill
"its just a passing fad" Gates said it was just a passing fad but would keep
an eye on it. This was somewhat documented if you look for a news story
(best guess 1998-2000) where MS forgot to renew the domain microsoft.com and
some guy in Kentucky or Tennessee shelled out the $35 for them and did it
himself.
When MS realized the internet wasn't going away anytime soon and little if
none of their software was internet ready (it's still in beta today), they
decided to put some effort in and get with the program.
Back in those days, "the internet" mostly consisted of email, web (sort of)
and usenet. Basically if you were an ISP, offering those 3 items put you in
the ballpark to call yourself that. This meant you went with something that
worked, namely unix.
This put MS at a crossroads since developement requires a two way street.
They thought they needed to create a usenet software package because it was
one of the basic elements of "the internet". But how do you test out
software that needs two to tango when you can't really "give it away"
because it was destined for making millions o'dollars as a propriatary
package?
You set up a server and let the developers kick the tires.
And thats it, the first story I heard of why "their" newsgroups were
floating around on usenet was because news.microsoft.com (or whatever they
called it then) was publically open. Users of it just set their newsreaders
to use nntp (a wild concept in those days over reading from a spool) and
pointed it to them. You had direct access to the minds behind the
programming at MS. It's why it's all support groups.
Their server(s) didn't propagate, they didn't exchange news, it's as simple
as that. People just configured the news reader to access them directly.
Like I mentioned earlier, someone got the idea to just take the posts and
feed them into a server somewhere. Was a good source of information to pass
along. Just seems to have stuck with all this time. I don't know if they are
currently peering legit or if the tradition of stealing their stuff lives
on.
The 2nd version of "why microsoft.* exists" is similar to the first except a
little later in the time line. At some point MS decided to get into the ISP
game themselves (pre MSN days). These were the days where even Kmart was
selling internet access (bluelight.net I think) and NetZero was "free today,
free forever"... another broken promise (net-zero, get it, net zero, no
cost, hmm, never mind)..
Anyway the big dog on the block was AOL, which at that time didn't have that
much to do with the internet because to make money, you needed to be a
content provider.
So version 2 of the story is the news.microsoft.com server(s) (again if that
is what it was called) was one of MS's EXCLUSIVE content features. I do
think this might of been closer to reality but since I didn't make notes
back then, personally I think the timeline is wrong.
I do remember the server was not accessible via nntp unless you were on one
of their dialup pools (again, remember, unless you had access to a dedicated
line, you were on dialup). This still wouldn't stop anyone from using the
dialup to grab the articles and re-distribute to usenet. Even if they had to
use a login, there isn't any identifiable marks in post to show it was taken
from any one account.
There was some speculation that whoever it was, was using something like
SLRN which is ancient, worked, common and can easily "grab" all the new
posts from a server into a batch, which was easily injected into another
server with only a few tweaks.
The point is, story #1, #2 or some combination of both, it was clear from
the beginning MS was not peering with anyone. I don't know what the current
situation is but historically, those groups were being redistributed "under
the table" and they just didn't care or didn't see any way to stop it.
The same is not true for the gmane.linux.* groups. If you check most servers
that still have them in the active file, they are all dead, no posts. They
used to be quite active. Like with the ms.* ones, someone decided "for the
public good" to scoop them up from gmane and redistribute on usenet. This
was maybe 4 to 6 years ago but someone at gmane pointed out it's more of a
public dis-service since posts made to those groups over usenet would never
return back to gmane.
So it stopped due to logic more than legal threats.
> Perhaps a set of duplicate microsoft newsgroups were created on usenet,
> to match the names exactly with the names created by microsoft on their
> server? Even if no peering existed between microsoft and the rest of
> usenet, those groups could exist and be functional on usenet without
> microsoft's permission or involvement.
Ok, shorter answer, I promise. I don't think there were microsoft.* control
messages for the longest time. I don't think they exist now but at least
someone is distributing a list of them with the descriptions.
We (ripco) added them for one simple reason, they kept showing up in the
"unwanted" section of the daily report from the news server. When it shows
you like the "top 30 unwanted groups" and the first 29 are
microsoft.something, you figure you are missing out on something and go with
the flow, just add them.
So we add them, a few of my peers see the same thing, they add them, then
they exist, no other reason.
> That's not necessarily a problem. It's Microsoft's choice to peer their
> server with usenet, but it's not up to them to control whether or not a
> parallel set of groups with exactly the same names appear on usenet. It
> then becomes a user choice to decide which microsoft "world" they want
> to participate in. Microsoft's closed little world, or the wider usenet
> world, or even both (but not simultaneously).
I dunno how they think, so can't answer authoritively. What I do know is the
first version of the story I heard "why they are here" I always beleived
because someone I knew had a copy of that usenet server software and nearly
busted a gut when I found out it was INN re-written for NT5 or whatever was
their "server platform" at the time. They even left some of the comments in
from Rich Salz in the code.
> But it doesn't appear that any such "sanctioned shovelling" is happening
> today. It does appear that microsoft does in fact have a working
> peering relationship with usenet. But I still haven't heard an
> explanation as to why they name their servers along the lines of
> TK2MSFTNGP02.phx.gbl.
I dunno, ask glorb or wherever else those paths seem to be injected into
where the feed is coming from. Never really seen an admin try to "protect
his sources". There must be some logic behind it.
Off hand, historically, that gbl was used by several companies in the past,
off the top of my head one of the last was either Global Crossings or just
plain Global Communications.
It's possible it's just a circuit ID number for a T1 circuit, at least that
first part with the TK2, the phx is probably phoenix az and gbl is the
people they got the line from, either now or at one time.
I can see completely naming a machine after the circuit id it was on. We
used to do that somewhat when 56k dedicated lines were popular. Was easier
to keep track of when a customer called with a problem, you at least knew
what direction to look into, which vendor to call and bitch at.
Back when we were nearly all shell boxes we used to name them after secret
service agents. Meant something to us, not anyone else. Seemed more
appropriate than planets, greek gods or whatever else was popular in those
days.
-bruce
b...@ripco.com
Bruce Esquibel
> What sense would it make to establish such a peering relationship, and
> then "stand up and shout for all world to hear that your groups are
> private" ??? What kind of logic is that?
I guess it depends on how the word "private" is defined.
I was just killing some time and figured out that:
is a public server, point your newsreader to it, full read and post access.
Now of course they only carry the microsoft.* groups, but if you post a
message on it then look at the headers...
From: Bruce Lang <bl...@ripco.com>
Subject: test12345
User-Agent: tin/1.8.2-20060425 ("Shillay") (UNIX) (SunOS/5.10 (i86pc))
Message-ID: <Ooh3uaqI...@TK2MSFTNGP05.phx.gbl>
Newsgroups: microsoft.public.nntp.test
Date: Fri, 21 Aug 2009 14:31:01 -0700
NNTP-Posting-Host: shell3.ripco.com 74.8.102.118
Path: TK2MSFTNGP01.phx.gbl!TK2MSFTNGP05.phx.gbl!not-for-mail
Lines: 1
Xref: TK2MSFTNGP01.phx.gbl microsoft.public.nntp.test:87392
It's pretty obvious those phx.gbl machines are theirs and like I said,
are probably just archaic names if anything.
I mean it still leaves the question if glorb and the others are doing a
direct peering with them but the point I wanted to make is, since the server
seems open to anyone, how can it be "private".
Since both read and post access is available on it, no authentication is
needed, I can't really see what difference it makes if the posts end up on
usenet or not.
They free and open, so privacy doesn't enter into the equation.
-bruce
b...@ripco.com
News Guy
> I was just killing some time and figured out that:
>
> msnews.microsoft.com
>
> is a public server, point your newsreader to it, full read and
> post access.
>
> They free and open, so privacy doesn't enter into the equation.
It's one thing to operate an individual (read: isolated) NNTP server to
allow full public read and write posting ability with no
authentication. I can understand how some might call that a private
server, hosting private groups, but allow public access to it.
If another party decides to do all the legwork to form a bidirectional
bridge between that private server and the rest of usenet, then I can
understand how that bridge runs counter to the intention of the
owner/operator of the server that the server be isolated from usenet.
But that is not really an aspect of the microsoft.* hierarchy I wanted
to know more about in this thread.
To restate my question, first my observation:
There are those that believe that the microsoft.* set of usenet
newsgroups have a unique relationship to microsoft, a relationship along
the lines of message coordination, control, aggregation and
distribution. A relationship that places the microsoft servers in a
role as "master" or "primary" servers, and all other usenet servers as
"slave" or "secondary" servers.
This role for microsoft and their servers has implications (so some
people claim) for the continued presence of the microsoft.* groups on
the internet at large, and that if microsoft so desired, that they could
force their removal from usenet either by "pulling the plug" (so to
speak) on the "master" role of their servers, or by exerting (or
threatening to exert) legal force on individual NNTP server operators.
My question being simply - is any of that true?
In asking this question, the conversation has taken several tangents,
some of them exploring the origins of these groups, the concepts of
peering and control messages as they apply to the microsoft.* hierarcy,
and the past and present connectivity that microsoft's nntp servers have
Adam H. Kerman
>Bruce Esquibel wrote:
>>I was just killing some time and figured out that:
>>is a public server, point your newsreader to it, full read and
>>post access.
>>They free and open, so privacy doesn't enter into the equation.
>It's one thing to operate an individual (read: isolated) NNTP server to
>allow full public read and write posting ability with no
>authentication. I can understand how some might call that a private
>server, hosting private groups, but allow public access to it.
Would you please not pull this on Bruce, who has genuine expertise as a
News administrator?
Private would mean "not offered to the public". It's not private. Gah.
Bruce Esquibel
> There are those that believe that the microsoft.* set of usenet
> newsgroups have a unique relationship to microsoft, a relationship along
> the lines of message coordination, control, aggregation and
> distribution. A relationship that places the microsoft servers in a
> role as "master" or "primary" servers, and all other usenet servers as
> "slave" or "secondary" servers.
> This role for microsoft and their servers has implications (so some
> people claim) for the continued presence of the microsoft.* groups on
> the internet at large, and that if microsoft so desired, that they could
> force their removal from usenet either by "pulling the plug" (so to
> speak) on the "master" role of their servers, or by exerting (or
> threatening to exert) legal force on individual NNTP server operators.
> My question being simply - is any of that true?
No, it's not true. None of it.
There is little doubt the origin of the microsoft.* groups on usenet is a
bit cloudy, but I can guarantee there was no conspiracy behind them. A lot
of stupidity probably but the door could of been locked and secured a long
time ago, they just never made any effort to do it.
You really need to study up on the Sciencetologist vs. Netcom case, besides
the current day RIAA vs. illegal uploads stuff, I think that case was the
only successful (mostly) usenet take down that was attempted. And like the
RIAA stuff, it only stopped where and who the posts were coming from, there
was no attempt to "go after" other servers where the posts ended up at.
I don't understand why people like you want to see a conspiracy where one
doesn't exist. Microsoft can easily "secure" that server with little to no
effort if in fact it's supposed to be private. It's obviously a public
server now even if it wasn't supposed to be in the past.
From the test post I made yesterday, it appears they are peering with at
least 3 of the same peers I do from examing the path headers, so the concept
of the posts not supposed to be passed around goes out the window.
Maybe the microsoft.* groups as a collective are the best example of how not
to run a server but if you count the thousands of other groups that fall
into the same mis-handling, they aren't really in a majority.
It's a public server that peers, maybe in one direction only but the only
thing fishy about it is why MS isn't charging for it's use. Seems
uncharacteristic of them.
-bruce
b...@ripco.com
News Guy
> > My question being simply - is any of that true?
>
> No, it's not true. None of it.
> I don't understand why people like you want to see a conspiracy
> where one doesn't exist.
I'm surprised you can't see my pov here. I'm NOT looking or arguing
*for* any sort of conspiracy.
I've summarized and posted here the beliefs and POV of others as a way
to expose how they are incorrect. I do not and have never believed that
microsoft has any special relationship to the microsoft.* hierarcy in
any way other than perhaps their initial, original creation.
> Microsoft can easily "secure" that server
This thread is not so much about microsoft's server as it is about how
(or if) Microsoft does or can influence the current operation (and even
existance) of the microsoft.* hierarchy on usenet. Again, some people
believe Microsoft can and does exert this influence. I believe they
don't because they can't. Everthing I've read here so far seems to be
in agreement with that.
> It's a public server that peers, maybe in one direction only
No - I don't believe it's only in one direction. I see posts end up on
the http version of their usenet service that originate from an outside
server and vice-versa.
> but the only thing fishy about it is why MS isn't charging
> for it's use. Seems uncharacteristic of them.
They also don't charge for a hotmail / msn account, so I don't think
it's fishy that they also don't charge to access their nntp server.
Besides, who would pay a fee to access a server that only carries a few
dozen newsgroups?
Bruce Esquibel
> This thread is not so much about microsoft's server as it is about how
> (or if) Microsoft does or can influence the current operation (and even
> existance) of the microsoft.* hierarchy on usenet. Again, some people
> believe Microsoft can and does exert this influence. I believe they
> don't because they can't. Everthing I've read here so far seems to be
> in agreement with that.
Well, you got me, I really don't understand what you are after.
So good luck with your quest for truth.
By the way, some of the hidden meaning to the path they use for the news
server seems to be answered in the arin record...
# nslookup msnews.microsoft.com
Server: 127.0.0.1
Address: 127.0.0.1#53
Non-authoritative answer:
Name: msnews.microsoft.com
Address: 207.46.248.16
# whois -h whois.arin.net 207.46.248.16
OrgName: Microsoft Corp
OrgID: MSFT
Address: One Microsoft Way
City: Redmond
StateProv: WA
PostalCode: 98052
Country: US
NetRange: 207.46.0.0 - 207.46.255.255
CIDR: 207.46.0.0/16
NetName: MICROSOFT-GLOBAL-NET
NetHandle: NET-207-46-0-0-1
Parent: NET-207-0-0-0-0
NetType: Direct Assignment
NameServer: NS1.MSFT.NET
NameServer: NS5.MSFT.NET
NameServer: NS2.MSFT.NET
NameServer: NS3.MSFT.NET
NameServer: NS4.MSFT.NET
Comment:
RegDate: 1997-03-31
Updated: 2004-12-09
Notice the OrgID, net name and the names of the listed named servers.
The plot thins.
-bruce
b...@ripco.com
Charles Lindsey
>As Adam Kerman has tried to expain to you, there can never be a
>"leakage" of nntp messages out of or into an NNTP server without the
>intentional establishment of a peering relationship to at least one
>other server.
On the contrary, that is simply not true. Anybody who has the capability
to read articles on a server has the ability to repost/reinject them on to
some other server (whether with the approval of the first server owner,
or not). That sort of a feed is usually known as a "suck feed" and usually
(but not necessarily) makes use of the NNTP NEWNEWS command.
For example, I run my own private server for my own use. I don't formally
"peer" with anyone (though I could if I so wished). If you look at the
Path of this article you may (or may not) find the host "clerew" at the
far end (that was certainly the Path when it left here).
>> AIUI (though I may be wrong) that is more or less what
>> happened to ms.*.
>Why you would define a deliberate, intentional peering relationship as a
>"leak" is, well, strange. I guess that in your lexicon, peering =
>leakage.
I don't define any such thing. But others have recounted the early history
of the microsoft groups in this thread, and they confirm my recollection
that it indeed started out as a "leak". Faced with that, Microsoft could
do little else than go along with it.
Charles Lindsey
>But that is not really an aspect of the microsoft.* hierarchy I wanted
>to know more about in this thread.
>To restate my question, first my observation:
>There are those that believe that the microsoft.* set of usenet
>newsgroups have a unique relationship to microsoft, a relationship along
>the lines of message coordination, control, aggregation and
>distribution. A relationship that places the microsoft servers in a
>role as "master" or "primary" servers, and all other usenet servers as
>"slave" or "secondary" servers.
For every statement, however bizarre, you may be sure that someone,
somewhere "believes" it; but that does not make it true.
In the case in question, then yes, there are a lot of people who believe
it. Maybe even Microsoft believe it. It may even have a grain of truth in
it.
>This role for microsoft and their servers has implications (so some
>people claim) for the continued presence of the microsoft.* groups on
>the internet at large, and that if microsoft so desired, that they could
>force their removal from usenet either by "pulling the plug" (so to
>speak) on the "master" role of their servers, or by exerting (or
>threatening to exert) legal force on individual NNTP server operators.
Again, there are undoubtedly people who believe that, but that does not
make it true (and it almost certainly isn't). And even if Microsoft COULD
make it happen, it would involve far more than "pulling the plug". OTOH,
Microsoft have access to many megadollars, and "money talks" :-) .
>My question being simply - is any of that true?
What is truth? You have to form your own opinion. Usenet has been
carefully and cunningly designed so as to make "absolute truth" a very
slippery concept. It is an "adhocracatic anarchy" - but it works.
John Eddy
laid off during some re-org work.
I've only had time to browse the thread (since I'm at work) and can
give the following (tho some is clouded by memory).
No, there isn't any conspiracy involved with the microsoft.* hierarchy
leaving msnews.microsoft.com. Basically, the server was set up for
public, anonymous access, with no feeds to usenet, but some
enterprising usenet admin who thought his users would like to access
the groups added them to his server decided to slurp all the posts.
Things being what they are, they slowly worked their way into usenet
proper.
Eventually, MS simply allowed everyone to pull posts from the server,
but, shortly before I was forced out, a dedicated feed was set up
with, IIRC, 2-3 providers.
There were, basically, three levels of post removal.
One that blocked posts from being submitted to the master server from
the slaves, one that trolled through posts after they were made and a
manual 'get rid of this message id' type process that was for one-off
work (although that second process could easily be used for one-off
work, for instance, a post to every newsgroup that contained an
unlikely phrase could be removed on an ad hoc basis).
I tried to block any outbound cancels. I believed (still do) that we
were well in our rights to remove posts from our own server, or even
users from directly accessing our server, but I didn't want to block
people from posting to the groups off of our servers.
Did I miss anything?
Adam H. Kerman
>Eventually, MS simply allowed everyone to pull posts from the server,
>but, shortly before I was forced out, a dedicated feed was set up
>with, IIRC, 2-3 providers.
Interesting history. So, at this point, when it sounds like they became
Usenet groups, why didn't you start issuing checkgroups?
News Guy
> I was the NG Admin for MS from, oh, 2003-ish to 2005 before being
> laid off during some re-org work.
>
> Did I miss anything?
I'll come back and post more questions, but for starters:
Why does microsoft name their NNTP servers along the lines of
something.phx.gbl ?
Why phx.gbl? What is the significance or meaning of that suffix?
Why no "X-complaints-to:" or "X-Abuse: " header line?
Why no "Organization: Microsoft Corporation" header line?
Also, why does Microsoft allow posts from external domains to "flow
through" their server, eg:
------------------
From: "N. Miller" <anon...@msnews.aosake.net>
Subject: Re: Selling license keys remover!
User-Agent: 40tude_Dialog/2.0.15.41
Organization: Microshaft
Date: Mon, 12 Oct 2009 11:48:51 -0700
Message-ID: <ozhbm48i9e3p$.d...@msnews.aosake.net>
Newsgroups: microsoft.public.windowsxp.general
NNTP-Posting-Host: 68.125.49.40
Path: (something) ! TK2MSFTNGP01.phx.gbl ! TK2MSFTNGP04.phx.gbl
------------------
Note the above message ID (not generated by MS) yet the path clearly
indicates the post was injected by an MS server. How would you explain
that? There are many similar examples where the message ID seems to be
created by the "banter.com" organization.
News Guy
> I was the NG Admin for MS from, oh, 2003-ish to 2005 before being
> laid off during some re-org work.
>