hugh wrote:
> The option does indeed exist in TP personalities' Properties. So far I
> haven't managed to get it to work. However I have come to the
> conclusion I am getting unnecessarily excited about something which
> doesn't seem to bother anyone else so I haven't spent any more time
> on it. Thank you all for your help -I've learnt a couple of more
> things about TP in the process - and I've been using it for about 20
> years I'm finding ES is working fine so that's one part of my
> transfer from Demon to BT sorted.
Ah, so the program is defective. Since demon no longer supports it, no
fixes will show up (from them). If it supports scripting to alter
behavior (rather than have to deal with the source code for the entire
program), you might be able to tell it to generate a MID value based on
your script that overrides what the defective code in the program wants
to use. I see from the manual there is a "You are using the wrong
script file" error message so maybe you can use scripts with Turnpike
but whether they can alter the MID header when you submit a message is
something I wouldn't know about. The definition of "script" may be
nothing more than a set of configuration settings or walking through
some configuration wizard and not some code you can write to alter the
program's behavior. Someone in the Turnpike newsgroup might know if
Turnpike supports scripts or not that alters its behavior.
Otherwise, I'd suggest disabling the MID option in a Personality setup
if that option is available and let the NNTP server assign its own MID
header.
>>> Has to do with the Message-ID (MID) header that *YOU* decided to
>>> insert into your outbound messages.
>> A false presumption.
>You willfully exhibit your ignorance almost like you're proud of it.
>There are only places where the MID header gets inserted: by the client
>or by the server. I've already mentioned what E-S inserts at their
>server for their MID value. It is NOT the poster's e-mail address. So,
>gee, if it isn't E-S server inserting a MID header with the poster's
>e-mail address, um, where else might it get added? What other source is
>left? Oh yeah, the client. Duh!
You wrote "that *YOU* decided". The person installing Turnpike will
have inserted his host-name or domain name on being asked to do so.
Turnpike will, unless instructed otherwise by a sufficiently
knowledgeable user, assign the message-ID right-hand-side from that.
The user probably does not even know, initially, that message-IDs exist,
or that they can be reconfigured.
>>> You configured your newsreader, Turnpike, to add its own MID header.
>> Another one.
>Guess you really want to show everyone that you're trying out for the
>troll-of-the-month award (a potted cactus).
No; merely one who can point out limitations in what you consider to be
your thinking. It's not difficult.
>>> I've never used Turnpike.
>> It shows. Do not write about what you are ignorant of.
>Ooh, I've been lambasted by a noob with all of 18 posts under his nym
>used here stretching into the distant past of 10 days ago.
This will be my third post in this thread.
>Heed your own advice.
>Oh yes, we've seen here how helpful you have been to the OP. Hmm, guess
>AIOE won't let you post about links to the manual for Turnpike, uh huh.
I do not need to help the OP here; I use another news group for that.
Like most Moroccans, Barry Margolin is more intelligent than you; but
that is a feeble and presumed-inadequate compliment to him.
-- (c) John Stockton, nr London UK. ?...@merlyn.demon.co.uk Turnpike v6.05 MIME.
Web <URL:http://www.merlyn.demon.co.uk/> - FAQish topics, acronyms, & links.
Check boilerplate spelling -- error is a public sign of incompetence.
Never fully trust an article from a poster who gives no full real name.
In message <jlakss$vm...@dont-email.me>, VanguardLH <V...@nguard.LH> writes
>hugh wrote:
>> The option does indeed exist in TP personalities' Properties. So far I
>> haven't managed to get it to work. However I have come to the
>> conclusion I am getting unnecessarily excited about something which
>> doesn't seem to bother anyone else so I haven't spent any more time
>> on it. Thank you all for your help -I've learnt a couple of more
>> things about TP in the process - and I've been using it for about 20
>> years I'm finding ES is working fine so that's one part of my
>> transfer from Demon to BT sorted.
In the turnpike newsgroup, I noticed some advice from Stocky about what
to use for a MID header. He states it is "courteous to make your the
right part of your M-ID recognisably yours" giving as a basis the
ability for others to recognize your posts. You'll find the MID header
rarely identifies a particular sender unless they have gone to the
trouble to register their own domain and use that in their MID header.
Everyone posting through Eternal-September is getting the same right-id
token (to the right of the ampersand character) with a random left-id
token, and that doesn't uniquely identify the poster at all.
You'll see what the RFC says to use, and that's a domain as the right-id
token (domain) portion of the MID header. Well, if you use a domain
then it better be one that you own since uniqueness of the left-id token
for the domain specified by the right-id token. When you were using
demon.co.uk to submit your messages, you were submitting to their server
and obviously they have permission to use their own domain. Some
newsreaders have established their own domains for their users that want
to generate MIDs in their clients rather than let the server generate
the MID header. For example, Forte registered the ax.com domain for
themselves for their Agent users to use in the right-id token of their
generated MID headers.
You should not be using someone else's domain because you cannot
guarantee that the left-id tokens your client generates will not
conflict with those the owner of that domain might use. When you used
demon.co.uk then their client's use of that domain was valid. Now that
you're not using them anymore means your posts should be using their
domain (or a subdomain of their domain). Now that you're using Eternal-
September as your newsgroups provider, you can let them use their domain
in their server-generated MID headers or you can have your client
specify its own MID headers but they shouldn't have Eternal's domain in
them (since you aren't coordinating your generated MID values with those
created by Eternal's server).
For users that want to generate their own MID headers, they need to use
a domain that they own in their right-id token in their MID headers. Do
you have any domains registered by you (i.e., some registrar leased you
a domain name)? As per the RFC, "The generator of the message
identifier MUST guarantee that the msg-id is unique." Using someone
else's domain means you can NOT guarantee you never generate MIDs that
won't conflict with those the domain owner might generate. When you use
someone else's domain (which is also without their permission) then the
servers have to resolve the conflicts.
VanguardLH <V...@nguard.LH> wrote:
>In the turnpike newsgroup, I noticed some advice from Stocky about what
>to use for a MID header. He states it is "courteous to make your the
>right part of your M-ID recognisably yours" giving as a basis the
>ability for others to recognize your posts. You'll find the MID header
>rarely identifies a particular sender unless they have gone to the
>trouble to register their own domain and use that in their MID header.
>Everyone posting through Eternal-September is getting the same right-id
>token (to the right of the ampersand character) with a random left-id
>token, and that doesn't uniquely identify the poster at all.
>RFC 5322: Internet Message Format
>http://www.rfc-editor.org/rfc/rfc5322.txt >Section 3.6.4: Identification Fields
Bloody hell. I never noticed before. Inclusion of Message-IDs are a SHOULD,
not a MUST. Why would that be?
>You'll see what the RFC says to use, and that's a domain as the right-id
>token (domain) portion of the MID header.
Why do I recall that it used to recommend fully-qualified domain, that
is, host+domain, to ensure uniqueness?
>You should not be using someone else's domain because you cannot
>guarantee that the left-id tokens your client generates will not
>conflict with those the owner of that domain might use.
That ought to be a MUST NOT, but I guess it's not.
> Bloody hell. I never noticed before. Inclusion of Message-IDs are a SHOULD,
> not a MUST. Why would that be?
Bear in mind that that RFC is describing email messages, not news
articles. Message-IDs have never been required for emails, and there
is no particular technical reason that they need to be.
>>You'll see what the RFC says to use, and that's a domain as the right-id
>>token (domain) portion of the MID header.
> Why do I recall that it used to recommend fully-qualified domain, that
> is, host+domain, to ensure uniqueness?
Perhaps because RFC 1036 section 2.1.5 says so.
>>You should not be using someone else's domain because you cannot
>>guarantee that the left-id tokens your client generates will not
>>conflict with those the owner of that domain might use.
> That ought to be a MUST NOT, but I guess it's not.
It is a requirement that the Message-ID be unique, but there are
many ways this could be achieved and the RFCs don't mandate any
particular one.
> >In the turnpike newsgroup, I noticed some advice from Stocky about what
> >to use for a MID header. He states it is "courteous to make your the
> >right part of your M-ID recognisably yours" giving as a basis the
> >ability for others to recognize your posts. You'll find the MID header
> >rarely identifies a particular sender unless they have gone to the
> >trouble to register their own domain and use that in their MID header.
> >Everyone posting through Eternal-September is getting the same right-id
> >token (to the right of the ampersand character) with a random left-id
> >token, and that doesn't uniquely identify the poster at all.
Jon Ribbens <jon+use...@unequivocal.co.uk> wrote:
>On 2012-04-02, Adam H. Kerman <a...@chinet.com> wrote:
>>VanguardLH <V...@nguard.LH> wrote:
>>>RFC 5322: Internet Message Format
>>>http://www.rfc-editor.org/rfc/rfc5322.txt >>>Section 3.6.4: Identification Fields
>>Bloody hell. I never noticed before. Inclusion of Message-IDs are a SHOULD,
>>not a MUST. Why would that be?
>Bear in mind that that RFC is describing email messages, not news
>articles. Message-IDs have never been required for emails, and there
>is no particular technical reason that they need to be.
You have a point. Given that, I wonder why they were included to begin with.
>>>You should not be using someone else's domain because you cannot
>>>guarantee that the left-id tokens your client generates will not
>>>conflict with those the owner of that domain might use.
>>That ought to be a MUST NOT, but I guess it's not.
>It is a requirement that the Message-ID be unique, but there are
>many ways this could be achieved and the RFCs don't mandate any
>particular one.
I wasn't commenting on uniqueness, just falsely claiming an association
with a particular domain.
Barry Margolin <bar...@alum.mit.edu> wrote:
>"Adam H. Kerman" <a...@chinet.com> wrote:
>>VanguardLH <V...@nguard.LH> wrote:
>>>In the turnpike newsgroup, I noticed some advice from Stocky about what
>>>to use for a MID header. He states it is "courteous to make your the
>>>right part of your M-ID recognisably yours" giving as a basis the
>>>ability for others to recognize your posts. You'll find the MID header
>>>rarely identifies a particular sender unless they have gone to the
>>>trouble to register their own domain and use that in their MID header.
>>>Everyone posting through Eternal-September is getting the same right-id
>>>token (to the right of the ampersand character) with a random left-id
>>>token, and that doesn't uniquely identify the poster at all.
>>>RFC 5322: Internet Message Format
>>>http://www.rfc-editor.org/rfc/rfc5322.txt >>>Section 3.6.4: Identification Fields
>>Bloody hell. I never noticed before. Inclusion of Message-IDs are a SHOULD,
>>not a MUST. Why would that be?
>>>You'll see what the RFC says to use, and that's a domain as the right-id
>>>token (domain) portion of the MID header.
>>Why do I recall that it used to recommend fully-qualified domain, that
>>is, host+domain, to ensure uniqueness?
>I think you're misunderstanding the terminology. host.domain is also a >domain.
No, Barry, that's why I asked a SPECIFIC question with UNAMBIGUOUS
terminology, given that "domain" can be ambiguous.
On 2012-04-02, Adam H. Kerman <a...@chinet.com> wrote:
> Jon Ribbens <jon+use...@unequivocal.co.uk> wrote:
>>Bear in mind that that RFC is describing email messages, not news
>>articles. Message-IDs have never been required for emails, and there
>>is no particular technical reason that they need to be.
> You have a point. Given that, I wonder why they were included to begin with.
My guess would be interoperability with other non-RFC724 mail systems.
That's reminded me of something amusing I had forgotten, that email
addresses (and Message-IDs) could originally optionally be literally
"local-part at domain" rather than "local-part@domain".
>>>>You should not be using someone else's domain because you cannot
>>>>guarantee that the left-id tokens your client generates will not
>>>>conflict with those the owner of that domain might use.
>>>That ought to be a MUST NOT, but I guess it's not.
>>It is a requirement that the Message-ID be unique, but there are
>>many ways this could be achieved and the RFCs don't mandate any
>>particular one.
> I wasn't commenting on uniqueness, just falsely claiming an association
> with a particular domain.
I know, but pretty much the only reason anyone cares about that is
that it makes guarantees of uniqueness impossible.
Part of the problem with mandating "an association with" a domain is
that it's extremely vague, and can't be enforced in any technological
manner anyway.
Jon Ribbens <jon+use...@unequivocal.co.uk> wrote:
>On 2012-04-02, Adam H. Kerman <a...@chinet.com> wrote:
>>Jon Ribbens <jon+use...@unequivocal.co.uk> wrote:
>>>Bear in mind that that RFC is describing email messages, not news
>>>articles. Message-IDs have never been required for emails, and there
>>>is no particular technical reason that they need to be.
>>You have a point. Given that, I wonder why they were included to begin with.
>My guess would be interoperability with other non-RFC724 mail systems.
But they wouldn't have used that style Message ID, and some of them had
different terms for the same concept.
>That's reminded me of something amusing I had forgotten, that email
>addresses (and Message-IDs) could originally optionally be literally
>"local-part at domain" rather than "local-part@domain".
On 2012-04-02, Adam H. Kerman <a...@chinet.com> wrote:
> Jon Ribbens <jon+use...@unequivocal.co.uk> wrote:
>>My guess would be interoperability with other non-RFC724 mail systems.
> But they wouldn't have used that style Message ID, and some of them had
> different terms for the same concept.
That wouldn't matter provided that you could define a two-way mapping
between Message-ID and whatever the other system used. Bear in mind
that RFC 822 et al deliberately placed very few restrictions on things
like addresses - in a "quoted string", almost any ASCII character is
allowable (including unprintable control characters).
> The manual
>says the user can set what domain gets used in the MID header.
Were you considerably smarter, you would have been aware that "can set"
and "did set" are significantly non-equivalent. And you might well have
thought that the manual that Hugh used (probably a nice printed one)
when installing Turnpike from new did not necessarily include all that
the electronic manual that you have seen has. See sig.
-- (c) John Stockton, nr London UK. ?...@merlyn.demon.co.uk Turnpike v6.05 MIME.
Web <URL:http://www.merlyn.demon.co.uk/> - FAQish topics, acronyms, & links.
Check boilerplate spelling -- error is a public sign of incompetence.
Never fully trust an article from a poster who gives no full real name.
> Barry Margolin <bar...@alum.mit.edu> wrote:
> >"Adam H. Kerman" <a...@chinet.com> wrote:
> >>VanguardLH <V...@nguard.LH> wrote:
> >>>In the turnpike newsgroup, I noticed some advice from Stocky about what
> >>>to use for a MID header. He states it is "courteous to make your the
> >>>right part of your M-ID recognisably yours" giving as a basis the
> >>>ability for others to recognize your posts. You'll find the MID header
> >>>rarely identifies a particular sender unless they have gone to the
> >>>trouble to register their own domain and use that in their MID header.
> >>>Everyone posting through Eternal-September is getting the same right-id
> >>>token (to the right of the ampersand character) with a random left-id
> >>>token, and that doesn't uniquely identify the poster at all.
> >>Bloody hell. I never noticed before. Inclusion of Message-IDs are a SHOULD,
> >>not a MUST. Why would that be?
> >>>You'll see what the RFC says to use, and that's a domain as the right-id
> >>>token (domain) portion of the MID header.
> >>Why do I recall that it used to recommend fully-qualified domain, that
> >>is, host+domain, to ensure uniqueness?
> >I think you're misunderstanding the terminology. host.domain is also a > >domain.
> No, Barry, that's why I asked a SPECIFIC question with UNAMBIGUOUS
> terminology, given that "domain" can be ambiguous.
When RFCs refer to domains, they mean any FQDN. Using the term to mean a container is an informal convention that should be avoided in formal documents.
> >>Bloody hell. I never noticed before. Inclusion of Message-IDs are a SHOULD,
> >>not a MUST. Why would that be?
> >Bear in mind that that RFC is describing email messages, not news
> >articles. Message-IDs have never been required for emails, and there
> >is no particular technical reason that they need to be.
> You have a point. Given that, I wonder why they were included to begin with.
Because they're useful.
One of the reasons why it's optional is so that a mail submission client can leave it out, and then the server can add it before forwarding.
Since it's often hard for clients to generate unique MIDs (because many client machines don't have a unique domain name for the RHS), it's likely to be better to leave it up to the server.
>> Bloody hell. I never noticed before. Inclusion of Message-IDs are a SHOULD,
>> not a MUST. Why would that be?
> Bear in mind that that RFC is describing email messages, not news
> articles. Message-IDs have never been required for emails, and there
> is no particular technical reason that they need to be.
The title of the RFC is "Internet Messsage Format". It is NOT just
about e-mail messaging. It's about Internet messaging and that includes
newsgroups. You will find references in RFCs for NNTP that will defer
to the Internet Message Format RFC. You'll find co-dependencies in many
RFCs since economy in RFC ratification is achieved through reuse rather
than rewrite.
For example, RFC 3977 for "Network News Transfer Protocol"
(http://tools.ietf.org/html/rfc3977) has its Appendix A section for
"Interaction with other Specifications" where it refers to Message-IDs
and which defers back to RFC 2822 (to older "Internet Message Format"
RFC).
>>>You'll see what the RFC says to use, and that's a domain as the right-id
>>>token (domain) portion of the MID header.
>> Why do I recall that it used to recommend fully-qualified domain, that
>> is, host+domain, to ensure uniqueness?
> Perhaps because RFC 1036 section 2.1.5 says so.
In RFC 5322 (and prior iterations of the RFC), section 3.6 denotes what
are the minimum and maximum counts for a header. For Message-ID, it may
appear a minimum of zero times or a maximum of 1 time but a comment that
remarks that the MID "should" be present. Despite whether the RFCs say
SHOULD, RECOMMENDED, MUST, or REQUIRED, they're still just specs which
client and server often don't heed or choose to disobey (deliberately or
due to ignorance).
The use of Message-ID is SHOULD or RECOMMENDED, not MUST or REQUIRED,
for what the RFC was written for (which is a general RFC on Internet
messaging, not solely for e-mail or newsgroups). However, without it
you don't have an indentifier with which to string the messages into a
threaded hierarchy. All hell would break loose in Usenet if you didn't
(or your server) include the MID header because Usenet would turn into
that worthless flat-file scheme used by web-based forums.
RFC 3977 for NNTP *does* demand the presence of the MID header. It says
"Every article handled by an NNTP server MUST have a unique message-id."
If the client doesn't insert one then the server MUST. If the client
inserts a MID header,
>> The manual says the user can set what domain gets used in the MID
>> header.
> Were you considerably smarter, you would have been aware that "can
> set" and "did set" are significantly non-equivalent.
"did set" was not in the quoted content you included in your reply;
i.e., you didn't provide the context. It wasn't anywhere in my posts.
It wasn't in the posts to which I replied. It wasn't in your reply to
hugh over in the Turnpike newsgroup. So you're making up the citation.
> And you might well have thought that the manual that Hugh used
> (probably a nice printed one) when installing Turnpike from new did
> not necessarily include all that the electronic manual that you have
> seen has.
I went and found a manual. hugh never said he had one. You never
divulged to hugh where was the manual (the abridged version to which you
allude or an expanded version).
> See sig.
Presumably I'm supposed to infer that you've now added the Turnpike
manual to your web site. Was hugh supposed to infer the same meaning
and similarly expected to dig around trying to find info (that isn't
there) on Turnpike? Just where (under which topic) on your web site at
http://www.merlyn.demon.co.uk/#LWTP might you have now added a copy of
the manual?
The only content that I found there was a short blurb on your "personal
views" regarding Usenet. There was some netiquette and terse tech info
there but hardly a reference for using Turnpike. I found
http://www.merlyn.demon.co.uk/misclinx.htm#DT which merely has a URL
link back to demon.co.uk (and not even to their Turnpike manual page).
Yes, I'm sure someone looking there for a copy of the Turnpike manual is
going to look under a miscellanous list of links to elsewhere. I only
found that web page by using Google's site search on "turnpike"
(http://www.google.com/search?q=site:www.merlyn.demon.co.uk+turnpike) to
see if it knew where you might've mentioned Turnpike since it's not one
of your Topics. http://www.merlyn.demon.co.uk/pc-links.htm also has
another external URL link to Turnpike.com but that's a dead site (it
redirects to demon.net).
Nope, didn't find a manual on Turnpike at your site. It might be there
but well buried; however, I really don't have a need to find out if you
have a copy of the manual since I already found it at demon.co.uk and
provided a link to it to hugh.
>>>> Has to do with the Message-ID (MID) header that *YOU* decided to
>>>> insert into your outbound messages.
>>> A false presumption.
>>You willfully exhibit your ignorance almost like you're proud of it.
>>There are only places where the MID header gets inserted: by the client
>>or by the server. I've already mentioned what E-S inserts at their
>>server for their MID value. It is NOT the poster's e-mail address. So,
>>gee, if it isn't E-S server inserting a MID header with the poster's
>>e-mail address, um, where else might it get added? What other source is
>>left? Oh yeah, the client. Duh!
> You wrote "that *YOU* decided". The person installing Turnpike will
> have inserted his host-name or domain name on being asked to do so.
> Turnpike will, unless instructed otherwise by a sufficiently
> knowledgeable user, assign the message-ID right-hand-side from that.
> The user probably does not even know, initially, that message-IDs exist,
> or that they can be reconfigured.
>>>> You configured your newsreader, Turnpike, to add its own MID header.
>>> Another one.
>>Guess you really want to show everyone that you're trying out for the
>>troll-of-the-month award (a potted cactus).
> No; merely one who can point out limitations in what you consider to be
> your thinking. It's not difficult.
>>>> I've never used Turnpike.
>>> It shows. Do not write about what you are ignorant of.
>>Ooh, I've been lambasted by a noob with all of 18 posts under his nym
>>used here stretching into the distant past of 10 days ago.
> This will be my third post in this thread.
>>Heed your own advice.
>>Oh yes, we've seen here how helpful you have been to the OP. Hmm, guess
>>AIOE won't let you post about links to the manual for Turnpike, uh huh.
> I do not need to help the OP here; I use another news group for that.
> Like most Moroccans, Barry Margolin is more intelligent than you; but
> that is a feeble and presumed-inadequate compliment to him.
> Jon Ribbens wrote:
>> Bear in mind that that RFC is describing email messages, not news
>> articles. Message-IDs have never been required for emails, and there
>> is no particular technical reason that they need to be.
> The title of the RFC is "Internet Messsage Format". It is NOT just
> about e-mail messaging. It's about Internet messaging and that includes
> newsgroups. You will find references in RFCs for NNTP that will defer
> to the Internet Message Format RFC.
Yes, thanks for that egg-sucking lesson.
> The use of Message-ID is SHOULD or RECOMMENDED, not MUST or REQUIRED,
> for what the RFC was written for (which is a general RFC on Internet
> messaging, not solely for e-mail or newsgroups).
Yes, you are repeating my point back to me.
> However, without it you don't have an indentifier with which to
> string the messages into a threaded hierarchy. All hell would break
> loose in Usenet if you didn't (or your server) include the MID
> header because Usenet would turn into that worthless flat-file
> scheme used by web-based forums.
Um, no, your message just would not appear or propagate since Usenet
relies fundamentally on Message-IDs.
Jon Ribbens wrote:
> On 2012-04-03, VanguardLH <V...@nguard.LH> wrote:
>> Jon Ribbens wrote:
>>> Bear in mind that that RFC is describing email messages, not news
>>> articles. Message-IDs have never been required for emails, and there
>>> is no particular technical reason that they need to be.
>> The title of the RFC is "Internet Messsage Format". It is NOT just
>> about e-mail messaging. It's about Internet messaging and that includes
>> newsgroups. You will find references in RFCs for NNTP that will defer
>> to the Internet Message Format RFC.
> Yes, thanks for that egg-sucking lesson.
>> The use of Message-ID is SHOULD or RECOMMENDED, not MUST or REQUIRED,
>> for what the RFC was written for (which is a general RFC on Internet
>> messaging, not solely for e-mail or newsgroups).
> Yes, you are repeating my point back to me.
>> However, without it you don't have an indentifier with which to
>> string the messages into a threaded hierarchy. All hell would break
>> loose in Usenet if you didn't (or your server) include the MID
>> header because Usenet would turn into that worthless flat-file
>> scheme used by web-based forums.
> Um, no, your message just would not appear or propagate since Usenet
> relies fundamentally on Message-IDs.
Wow, you certainly are super-sensitive on anyone that steps on your
coattails to expound on your statements.
>>In the turnpike newsgroup, I noticed some advice from Stocky about what
>>to use for a MID header. He states it is "courteous to make your the
>>right part of your M-ID recognisably yours" giving as a basis the
>>ability for others to recognize your posts. You'll find the MID header
>>rarely identifies a particular sender unless they have gone to the
>>trouble to register their own domain and use that in their MID header.
>>Everyone posting through Eternal-September is getting the same right-id
>>token (to the right of the ampersand character) with a random left-id
>>token, and that doesn't uniquely identify the poster at all.
>Bloody hell. I never noticed before. Inclusion of Message-IDs are a SHOULD,
>not a MUST. Why would that be?
It is only necessary to have an M-ID on the open Internet, where one has
no idea what other messages it might meet. While a message is on a
"private link" between its generator and the system which releases it
onto the open Net, that form of identification is not required.
Consider a postal message. You might write its initial form, and give
it to your secretary saying "Send this to Fat Fred at Smiths, please".
The secretary will prepare an envelope with a proper postal address,
"Attn: Mr F Brown, / Smith and Jones, Inc., / ... <Zipcode>" and get it
posted.
Turnpike can send mail in two distinct ways. For one, it needs to
generate an M-ID. For the other, it starts off in a different manner,
Demon assigning a unique M-ID. IIRC, the second way is no longer
supported at Demon, for being nowadays insufficiently secure.
Please remember that, at least in respect of Turnpike & Demon users, VLH
retains an inadequate understanding of the systems (as the /cognoscenti/
will see from what you quote above); considerably less, I suspect, than
yours. He should remember the words of the (well, a) great Huxley.
-- (c) John Stockton, nr London, UK. ?...@merlyn.demon.co.uk Turnpike 6.05 WinXP.
Web <http://www.merlyn.demon.co.uk/> - FAQ-type topics, acronyms, and links.
Command-prompt MiniTrue is useful for viewing/searching/altering files. Free,
DOS/Win/UNIX now 2.0.6; see <URL:http://www.merlyn.demon.co.uk/pc-links.htm>.
Once more, you are emitting from the wrong end of your alimentary canal.
>In the turnpike newsgroup, I noticed some advice from Stocky about what
>to use for a MID header. He states it is "courteous to make your the
>right part of your M-ID recognisably yours" giving as a basis the
>ability for others to recognize your posts. You'll find the MID header
>rarely identifies a particular sender unless they have gone to the
>trouble to register their own domain and use that in their MID header.
That sentence is sloppy and untrue. I wrote "recognisably yours",
nothing about owning domains. I'm discussing courtesy, not technical
requirements such as uniqueness, here (the word may be in Webster).
And, in context, that does not need to be "recognisable by all and
sundry", just recognisable by someone who has already seen some of his
M-IDs. At least one PITA uses fully-recognisable M-IDs - that enables me
to kill, unseen, not only all messages from him but also most messages
in threads descending from a message from him. That's useful. You, as
a reader, may lack that capability; it depends on the software you have
chosen.
By always using its own M-ID, E-S would defend all of its posters from
that treatment.
Then there is no need for the domain to be one that one owns. Mine is
owned by Demon, but leased by me. They have granted me full rights over
the use of that domain name, for so long as that pleases them. For
uniqueness, it is control, not ownership, that matters. Hugh has a
similar account. But we have not registered our domains; they are sub-
domains of one registered (presumbly) by Demon.
>Everyone posting through Eternal-September is getting the same right-id
>token (to the right of the ampersand character) with a random left-id
>token, and that doesn't uniquely identify the poster at all.
If that is true, it is only very recently true. The first article in
this thread has a Demon M-ID in Turnpike style, and its Path has E-S at
both ends.
But I hope ES-generated left parts are not random; that does not
guarantee uniqueness. They seem to be 12-character base-64 strings
Inspection shows that the probability of them being fully-random is
negligible. If they are mostly random, duplicates will be rare.
>You'll see what the RFC says to use, and that's a domain as the right-id
>token (domain) portion of the MID header. Well, if you use a domain
>then it better be one that you own since uniqueness of the left-id token
>for the domain specified by the right-id token. When you were using
>demon.co.uk to submit your messages, you were submitting to their server
>and obviously they have permission to use their own domain.
But Demon do not have the right to use subdomains that they have leased
out, unless that lease id terminated.
>You should not be using someone else's domain because you cannot
>guarantee that the left-id tokens your client generates will not
>conflict with those the owner of that domain might use.
You are not invariably wrong.
> When you used
>demon.co.uk then their client's use of that domain was valid. Now that
>you're not using them anymore means your posts should be using their
>domain (or a subdomain of their domain).
One can post (as I do) through Mozilla, E-S, AIOE without losing ones
rights over one's Demon sub-domain; and that includes the right to base
M-IDs on it.
Huxley.
-- (c) John Stockton, nr London, UK. ?...@merlyn.demon.co.uk Turnpike 6.05 WinXP.
Web <http://www.merlyn.demon.co.uk/> - FAQ-type topics, acronyms, and links.
Command-prompt MiniTrue is useful for viewing/searching/altering files. Free,
DOS/Win/UNIX now 2.0.6; see <URL:http://www.merlyn.demon.co.uk/pc-links.htm>.
Dr J R Stockton wrote, in
<4iyBRiUJb1ePF...@invalid.uk.co.demon.merlyn.invalid>
on Tue, 3 Apr 2012 21:00:09 +0100:
>But I hope ES-generated left parts are not random; that does not
>guarantee uniqueness. They seem to be 12-character base-64 strings
>Inspection shows that the probability of them being fully-random is
>negligible. If they are mostly random, duplicates will be rare.
"The generation of random numbers is too important to be left to
chance." Robert R. Coveyou
"Any one who considers arithmetical methods of producing random digits
is, of course, in a state of sin." John von Neumann
-- Nick Spalding
In message <81tnn7llq81vsuddlt17vda7ightupi...@4ax.com>, Nick Spalding <spald...@iol.ie> writes
>Dr J R Stockton wrote, in
><4iyBRiUJb1ePF...@invalid.uk.co.demon.merlyn.invalid>
> on Tue, 3 Apr 2012 21:00:09 +0100:
>>But I hope ES-generated left parts are not random; that does not
>>guarantee uniqueness. They seem to be 12-character base-64 strings
>>Inspection shows that the probability of them being fully-random is
>>negligible. If they are mostly random, duplicates will be rare.
>"The generation of random numbers is too important to be left to
>chance." Robert R. Coveyou
>"Any one who considers arithmetical methods of producing random digits
>is, of course, in a state of sin." John von Neumann
Just noticed on another group that the change I had previously attempted without apparent success in TP now seems to be working.
So the RHS of this M-ID should be [127.0.0.1]
Here goes
-- hugh
hugh <hugh@[127.0.0.1]> wrote:
> In message <81tnn7llq81vsuddlt17vda7ightupi...@4ax.com>, Nick Spalding > <spald...@iol.ie> writes
> >Dr J R Stockton wrote, in
> ><4iyBRiUJb1ePF...@invalid.uk.co.demon.merlyn.invalid>
> > on Tue, 3 Apr 2012 21:00:09 +0100:
> >>But I hope ES-generated left parts are not random; that does not
> >>guarantee uniqueness. They seem to be 12-character base-64 strings
> >>Inspection shows that the probability of them being fully-random is
> >>negligible. If they are mostly random, duplicates will be rare.
> >"The generation of random numbers is too important to be left to
> >chance." Robert R. Coveyou
> >"Any one who considers arithmetical methods of producing random digits
> >is, of course, in a state of sin." John von Neumann
> Just noticed on another group that the change I had previously attempted > without apparent success in TP now seems to be working.
> So the RHS of this M-ID should be [127.0.0.1]
> Here goes
