You're right, Dave, I didn't read the original license language -- I
should have. I personally don't see any real issues with the language
quoted below, as it grants a license to copy, distribute, and modify,
which are the three rights that make the most sense in a software
context, as well as "use".

However, on a point of extreme pedantry, the copyright notice itself is
technically flawed. You're OK here because you added some redundancy,
but under US law a copyright notice must either fully spell out the word
Copyright, abbreviate it as "Copr.", or use the c-in-a-circle symbol --
the "(c)" construct isn't a valid substitute.

(While under US law, any of the three forms suffice, if you include the
real c-in-a-circle, it provides additional protection under one
international treaty, the Universal Copyright Convention.)

I'm not aware of any cases on point, but because the relevant laws and
treaties specifically require the c-in-circle symbol, there is a broad
consensus that c-in-parentheses does *not* suffice in a copyright
notice, and that adding it does no good at all.

The US Copyright office circular on copyright notices is at
http://www.copyright.gov/circs/circ03.html#form

Sorry for that diversion, and as noted your use of the full word
"Copyright" solves the problem -- but the use of "(c)" is so widespread
that I wanted to use this opportunity to denounce it.

John
----

David L. Mills said the following on 06/15/2008 12:05 AM:
> John,
>
> I can tell you haven't read the statement. Please do. Perhaps a lawyer
> would have phrased it differently and maybe modern times are different
> than mellow old MIT days, but I am loath to change it on advice, as this
> might compromise prior use and interpretation of the statement.
>
> Dave
>
> ************************************************** *********************
> * *
> * Copyright (c) David L. Mills 1992-2008 *
> * *
> * Permission to use, copy, modify, and distribute this software and *
> * its documentation for any purpose with or without fee is hereby *
> * granted, provided that the above copyright notice appears in all *
> * copies and that both the copyright notice and this permission *
> * notice appear in supporting documentation, and that the name *
> * University of Delaware not be used in advertising or publicity *
> * pertaining to distribution of the software without specific, *
> * written prior permission. The University of Delaware makes no *
> * representations about the suitability this software for any *
> * purpose. It is provided "as is" without express or implied
> *
> * warranty. *
> * *
> ************************************************** *********************John
> Ackermann N8UR wrote:
>
>> David Woolley said the following on 06/14/2008 04:34 PM:
>>
>>
>>> To the extent that copyright agreements are optional, their purpose is
>>> normally to give you a copy only on the condition that you do not
>>> exercise some of the rights that you would have if given the copy
>>> without making the agreement.
>>>
>>> I suspect Open BSD's problem is that the statement of permission doesn't
>>> actually give all the rights that are restricted by copyright. In
>>> particular, if it says "use", that has a technical meaning that doesn't
>>> include copying to give to others.
>>
>> Good point, David. The US copyright law lists five exclusive rights
>> that the copyright holder possesses. "Use" isn't one of them, and that
>> does make software copyright issues more interesting. The five rights
>> are the right to: copy, create derivative works, distribute, perform
>> publicly (applies only to certain types of works), and to display
>> publicly (also applies only to certain types of works).
>>
>> The reason that "use" normally works in the software context is that to
>> make software function you almost always need to copy it (for example,
>> from hard disk to RAM). In some situations, "use" would also trigger
>> the display right. But it's also possible to conceive of a computing
>> device with nonvolatile memory such that there is no copying involved in
>> running the the software. So, a right to "use" is inherently ambiguous
>> from a copyright perspective.
>>
>> I believe (pretty strongly) that "use" would not normally be interpreted
>> to include the right to distribute copies to third parties. However,
>> that right could be implied by the circumstances (if I publish my
>> software on a web site and encourage downloads with no restrictions, it
>> might well be that I've granted an implied license to those who download
>> to further distribute.
>>
>> By the way -- relying on copyright notices drafted in the early '80s
>> isn't necessarily a good idea. At least in the US, there was still
>> significant debate on whether software could even be copyrighted up
>> until 1983, when Apple v. Franklin pretty much laid that to rest.
>> Rights notices of the time were experimental, to say the least...
>>
>> John
>
> _______________________________________________
> questions mailing list
> questions@lists.ntp.org
> https://lists.ntp.org/mailman/listinfo/questions
>
>

"David L. Mills" writes:

>David, Anton,

>As I have reminded folks many, many times, please read the copyright
>statement in the distribution. Note very carefully that there is no
>license, only a copyright statement that clearly specifies no fee, no
>restictions on use other than including the copyright statement itself
>in derived products.

A license is a legal statement about giving permission to copy the program.
If there were really no license, noone else would have the right to copy
the program. You give them that right. That is a license.

It is a slightly bizarre one, since the copyright is claimed to owned by
you, and you put in a term saying that the University of Deleware name is
not to be used. Can I ask you to put in that my name is also not to be used
in any advertising? :-). (In the case of Project Athena I suspect that the
copyright is actually owned by MIT while in this case it is owned by you.)


>Except for one minor word change, the copyright statement is identical
>to the one used by MIT in Project Athena and has not chenged since 1988.

>Be advised, lawyers have already stalked me about restictive licenses; I
>threw the one from IBM, who showed up unannounced, right back out the door.

>Dave

>David Woolley wrote:
>> Anton Persson A wrote:
>>
>>> Is this requirement also valid for ntp v3, I can't quite find it
>>> but I'm unsure? The reason I'm asking is that we have quite
>>> a blood thirsty testing squad here and I need to clarify
>>> if this is a compliance requirement or not.
>>
>>
>> Why aren't you just using the reference implementation code? The only
>> issue I'm aware of about the licence is that that OpenBSD claim the
>> wording means it doesn't meet their free for commercial use requirements
>> whereas Dave Mills claims that the licence already meets that
>> requirement, and doesn't need rewording.

I have no idea what the OpenBSD claim means. I would think that "any
purpose with or without fee" is clear enough to cover "commercial use".
What is their legal theory about "any purpose with or without fee" that would
exclude "commercial use".

David Woolley writes:

>David L. Mills wrote:

>> statement in the distribution. Note very carefully that there is no
>> license, only a copyright statement that clearly specifies no fee, no

>If that were really true (that there was no licence), almost no-one
>would be able to use the code. A copyright notice says what is
>copyright and who owns the copyright. It just removes doubt as to
>whether a licence is needed (not much software was written by people who
>have been dead for 70 years, so most software needs a licence; software
>written by the US government, at least within the USA, is a special
>case), and tells you who is able to give that licence.

>A licence is a statement of permission to do things which would
>otherwise be illegal. Copying the ntpd code would be illegal without a
>licence.

>> restictions on use other than including the copyright statement itself
>> in derived products.

>The statement that there are no restrictions is actually the licence.

>You may be confused by commercial software, which doesn't use a bare
>licence, but uses a licence agreement, which is actually a contract.
>Note that some people argue that bare licences are not possible in many
>countries. Some people even argue that bare licences are not valid in
>the USA.

>To the extent that copyright agreements are optional, their purpose is
>normally to give you a copy only on the condition that you do not
>exercise some of the rights that you would have if given the copy
>without making the agreement.

>I suspect Open BSD's problem is that the statement of permission doesn't
>actually give all the rights that are restricted by copyright. In
>particular, if it says "use", that has a technical meaning that doesn't
>include copying to give to others.

To quote the license
* Permission to use, copy, modify, and distribute this software and *
* its documentation for any purpose with or without fee is hereby *
* granted

I find it hard to develope a legal theory which makes this restrictive.

>People in the ntpd community generally assume that the copyright owners
>will act as though the licence you intended exists, but others may want
>to act on the basis of what the licence actually says.

It then helps to actually read the license.


>Normally the following would be taken as implied, but I will make them
>explicit, as an example. I'm not a lawyer, so they may well not be
>watertight. You should independently verify any statements about the
>law made here.

>Copyright notice: This article is Copyright 2008 David J Woolley of
>London England.
>Parts prefixed ">" are Copyright 2008 Doctor David L Mills and/or the
>University of Delaware (and are included under the doctrine of fair use
>for criticism or review).

>Licence: David J Woolley grants permission for all acts covered by
>copyright law which contributors to a USENET newsroup gatewayed to a
>public mailing list would normally expect to be permitted.

Yee gads. That would be terrible. It would create a legal morass in trying
to define "normally expect to be permitted" Normal for whom, expect by
whom? contributors are an extremely varied bunch. I might expect David
Mills to jump and modify the code at my whim, and that I can sue him if he
does not do so. Does this make it "normally expect"? Does one have to poll
all usenet users to find out what is "normal"?

The words David actually uses in his license are far far clearer.

jra@febo.com (John Ackermann N8UR) writes:

>David Woolley said the following on 06/14/2008 04:34 PM:

>> To the extent that copyright agreements are optional, their purpose is
>> normally to give you a copy only on the condition that you do not
>> exercise some of the rights that you would have if given the copy
>> without making the agreement.
>>
>> I suspect Open BSD's problem is that the statement of permission doesn't
>> actually give all the rights that are restricted by copyright. In
>> particular, if it says "use", that has a technical meaning that doesn't
>> include copying to give to others.

>Good point, David. The US copyright law lists five exclusive rights

Except that Mills uses the words "use copy modify distribute" Now you could
argue tht "use" is inappropriate since nothing in copyright law gives him
any ability to restrict the use, but since he is giving blanket permission,
that is irrelevant. copy=copy, modify=create derivative work,
distribute=distribute. You might argue that use=perform publicly.


>that the copyright holder possesses. "Use" isn't one of them, and that
>does make software copyright issues more interesting. The five rights
>are the right to: copy, create derivative works, distribute, perform
>publicly (applies only to certain types of works), and to display
>publicly (also applies only to certain types of works).

>The reason that "use" normally works in the software context is that to
>make software function you almost always need to copy it (for example,
>from hard disk to RAM). In some situations, "use" would also trigger
>the display right. But it's also possible to conceive of a computing
>device with nonvolatile memory such that there is no copying involved in
>running the the software. So, a right to "use" is inherently ambiguous
>from a copyright perspective.

Only if the license attempts to restrict the right to use. If it grants it,
it is irrelevant. It could also give you the right to breathe, and the fact
that copyright law cannot be used to restrict that right would be equally
irrelevant.



>I believe (pretty strongly) that "use" would not normally be interpreted
>to include the right to distribute copies to third parties. However,
>that right could be implied by the circumstances (if I publish my
>software on a web site and encourage downloads with no restrictions, it
>might well be that I've granted an implied license to those who download
>to further distribute.

>By the way -- relying on copyright notices drafted in the early '80s
>isn't necessarily a good idea. At least in the US, there was still
>significant debate on whether software could even be copyrighted up
>until 1983, when Apple v. Franklin pretty much laid that to rest.
>Rights notices of the time were experimental, to say the least...

>John

Unruh wrote:
> David Woolley writes:
>
> To quote the license
> * Permission to use, copy, modify, and distribute this software and *
> * its documentation for any purpose with or without fee is hereby *
> * granted
>
> I find it hard to develope a legal theory which makes this restrictive.

It has already been established that the licence has changed since
OpenBSD made their objections, so they were not objecting to the exact
licence from which you are quoting.

>
>> Licence: David J Woolley grants permission for all acts covered by
>> copyright law which contributors to a USENET newsroup gatewayed to a
>> public mailing list would normally expect to be permitted.
>
> Yee gads. That would be terrible. It would create a legal morass in trying
> to define "normally expect to be permitted" Normal for whom, expect by

But that is exactly what one is doing by posting to USENET without an
explicit licence. I was simply trying to state the implied licence,
which is very subjective.

jra@febo.com (John Ackermann N8UR) writes:

>You're right, Dave, I didn't read the original license language -- I
>should have. I personally don't see any real issues with the language
>quoted below, as it grants a license to copy, distribute, and modify,
>which are the three rights that make the most sense in a software
>context, as well as "use".

>However, on a point of extreme pedantry, the copyright notice itself is
>technically flawed. You're OK here because you added some redundancy,
>but under US law a copyright notice must either fully spell out the word
>Copyright, abbreviate it as "Copr.", or use the c-in-a-circle symbol --
> the "(c)" construct isn't a valid substitute.

>(While under US law, any of the three forms suffice, if you include the
>real c-in-a-circle, it provides additional protection under one
>international treaty, the Universal Copyright Convention.)

>I'm not aware of any cases on point, but because the relevant laws and
>treaties specifically require the c-in-circle symbol, there is a broad
>consensus that c-in-parentheses does *not* suffice in a copyright
>notice, and that adding it does no good at all.

>The US Copyright office circular on copyright notices is at
>http://www.copyright.gov/circs/circ03.html#form

>Sorry for that diversion, and as noted your use of the full word
>"Copyright" solves the problem -- but the use of "(c)" is so widespread
>that I wanted to use this opportunity to denounce it.

Yes, it is pedantry, since the copyright is automatic and does not need to
be asserted. Ie, that notice is infomational, not a legal requirement.
It simply informs the reader as to who actually owns the copyright. What is
not mere pedantry is where or not this claim is actually true. Many people
have contributed to ntpd, and unless they all transfered copyright to David
Mills, then the claim that he owns the copyright is really false. They all
do (David has copyright interest in all of the works since the other
people's work is derivative of his work, but others have copyright interest
as well.) To keep things perfectly clean, David shoule ask anyone who
contributes to transfer their copyright to him.



>John
>----

>David L. Mills said the following on 06/15/2008 12:05 AM:
>> John,
>>
>> I can tell you haven't read the statement. Please do. Perhaps a lawyer
>> would have phrased it differently and maybe modern times are different
>> than mellow old MIT days, but I am loath to change it on advice, as this
>> might compromise prior use and interpretation of the statement.
>>
>> Dave
>>
>> ************************************************** *********************
>> * *
>> * Copyright (c) David L. Mills 1992-2008 *
>> * *
>> * Permission to use, copy, modify, and distribute this software and *
>> * its documentation for any purpose with or without fee is hereby *
>> * granted, provided that the above copyright notice appears in all *
>> * copies and that both the copyright notice and this permission *
>> * notice appear in supporting documentation, and that the name *
>> * University of Delaware not be used in advertising or publicity *
>> * pertaining to distribution of the software without specific, *
>> * written prior permission. The University of Delaware makes no *
>> * representations about the suitability this software for any *
>> * purpose. It is provided "as is" without express or implied
>> *
>> * warranty. *
>> * *
>> ************************************************** *********************John
>> Ackermann N8UR wrote:
>>
>>> David Woolley said the following on 06/14/2008 04:34 PM:
>>>
>>>
>>>> To the extent that copyright agreements are optional, their purpose is
>>>> normally to give you a copy only on the condition that you do not
>>>> exercise some of the rights that you would have if given the copy
>>>> without making the agreement.
>>>>
>>>> I suspect Open BSD's problem is that the statement of permission doesn't
>>>> actually give all the rights that are restricted by copyright. In
>>>> particular, if it says "use", that has a technical meaning that doesn't
>>>> include copying to give to others.
>>>
>>> Good point, David. The US copyright law lists five exclusive rights
>>> that the copyright holder possesses. "Use" isn't one of them, and that
>>> does make software copyright issues more interesting. The five rights
>>> are the right to: copy, create derivative works, distribute, perform
>>> publicly (applies only to certain types of works), and to display
>>> publicly (also applies only to certain types of works).
>>>
>>> The reason that "use" normally works in the software context is that to
>>> make software function you almost always need to copy it (for example,
>>> from hard disk to RAM). In some situations, "use" would also trigger
>>> the display right. But it's also possible to conceive of a computing
>>> device with nonvolatile memory such that there is no copying involved in
>>> running the the software. So, a right to "use" is inherently ambiguous
>>> from a copyright perspective.
>>>
>>> I believe (pretty strongly) that "use" would not normally be interpreted
>>> to include the right to distribute copies to third parties. However,
>>> that right could be implied by the circumstances (if I publish my
>>> software on a web site and encourage downloads with no restrictions, it
>>> might well be that I've granted an implied license to those who download
>>> to further distribute.
>>>
>>> By the way -- relying on copyright notices drafted in the early '80s
>>> isn't necessarily a good idea. At least in the US, there was still
>>> significant debate on whether software could even be copyrighted up
>>> until 1983, when Apple v. Franklin pretty much laid that to rest.
>>> Rights notices of the time were experimental, to say the least...
>>>
>>> John
>>
>> _______________________________________________
>> questions mailing list
>> questions@lists.ntp.org
>> https://lists.ntp.org/mailman/listinfo/questions
>>
>>

Unruh said the following on 06/15/2008 11:35 AM:

> Yes, it is pedantry, since the copyright is automatic and does not need to
> be asserted. Ie, that notice is infomational, not a legal requirement.
> It simply informs the reader as to who actually owns the copyright. What is
> not mere pedantry is where or not this claim is actually true. Many people
> have contributed to ntpd, and unless they all transfered copyright to David
> Mills, then the claim that he owns the copyright is really false. They all
> do (David has copyright interest in all of the works since the other
> people's work is derivative of his work, but others have copyright interest
> as well.) To keep things perfectly clean, David shoule ask anyone who
> contributes to transfer their copyright to him.

While the notice is not mandatory, it does have legal value -- in
particular, it negates a defense of innocent infringement. And, to
claim the international protection of the UCC treaty I referred to in my
other message, you do need to include the notice.

Your point about whether the copyright notice is accurate is
interesting. I wonder if it would even be possible at this point to
determine who all the contributors are!

I thought that Linux used the CREDITS file as a list of copyright
holders, but I just looked at some source and that doesn't appear to be
the case -- instead, each source file has its own copyright notice in
the name of one or more people, with the COPYING file not listing any
copyright owners at all. I guess that has the effect of putting the
whole kernel tree under the GPL, but the resulting object file(s) as a
collective work of all the contributors.

I just looked at the ntp 4.2.0 source (the only tree I have handy at the
moment) and the random source files I looked at do not have any
copyright notice at all, so the Linux model doesn't appear to be in place.

John

David,

I have resisted to the point of anal retention any changes to the
copyright/license statement on the basis that it might compromise prior
use and interpretation over the last twenty years. A few years ago, and
after much flame and flame retardant, I agreed to change one word on the
advice this did not change the interpretation, only resolved a possible
ambiguity. As I said, what you get is what you get. I have referred the
issue to our campus legal department.

Dave

David Woolley wrote:
>> * *
>> * Copyright (c) David L. Mills 1992-2008 *
>
>
> The above is the copyright notice. Everything else is the licence, so
> rather than not having a licence, all but one line is actually the licence.
>
>> * *
>> * Permission to use, copy, modify, and distribute this software and *
>
>
> This uses more than the term "use", so your summary was wrong. It also
> sounds like the licence has heen changed since the OpenBSD objections,
> but anyone trying to implement version 3 (presumably because they have a
> mandate to only implement official standards, although I'm waiting for
> the actual answer on that question), would see the old licence on the
> reference implementation and therefore might not treat it as having new
> style BSD type licence.
>
> I'd have to search the archives to work out what the exact OpenBSD
> objection was.
>
> Note that my statement about use being illegal was rhetorical, based on
> your claim that is didn't have a licence. You were wrong in saying that
> it didn't have a licence, and I tried to explain that common commercial
> use of "licence agreement" for what is actually a contract limiting
> rights, might be the source of your confusion.

David,

I have anticipated a battle over my carcass and hired expensive estaqte
and portfolio managers. They will take care of the issues you raise.
However, I find your comments in extremely poor taste.

Dave

David Woolley wrote:

> David L. Mills wrote:
> [ In relationship to the ntpd code ]
>
>> * *
>> * Copyright (c) David L. Mills 1992-2008 *
>
>
> If this is actually correct (i.e. it isn't a work for hire actually
> owned by the University of Delaware), it would be highly advisable to
> think about transferring ownership to a trust, or similar.
>
> If Wikipedia is correct, David Mills has reached his three score years
> and ten. When he dies, copyright ownership will be part of his estate,
> and unless his will donates that ownership to an appropriate trust,
> etc., or someone who shares his passion for NTP, you may be left with a
> situation where the licence terms are frozen for 70 years and cannot be
> modified to account for subsequent changes in case or statute law. (I
> don't know if Delaware has any form of inheritance tax, but if it does,
> it is conceivable that the intellectual property might be considered to
> have monetary value and be taxable. There is little precedent, as there
> are very few 70 year old authors of major pieces of open source software
> that still own the copyright.)
>
> Even more of a danger is that he might become legally "incapable" before
> he dies. Without some appropriate legal protection, that could leave
> things in limbo for many years.
>
> .

>>> In article <4854d32f$0$755$5a6aecb4@news.aaisp.net.uk>, David Woolley writes:

David> I'd have to search the archives to work out what the exact OpenBSD
David> objection was.

The original wording was "... and without fee..."

If you look hard enough you will see that it *could* be read different ways.

The potential for the problem would be that a) Dave decides that the 'fee'
means people cannot charge for distributing the code as part of a larger
work instead of 'there is no fee for the license rights', and b) when this
got to a judge the judge decided on the more restrictive reading against the
(unempowered) user instead of deciding that the (empowered) license holder
should be allowed the more restrictive interpretation.

From what I have heard, judges go for the least restrictive interpretation
in these cases and find for the user, since the 'onus' of accurate
description is on the shoulders of the writer of the agreement.
--
Harlan Stenn
http://ntpforum.isc.org - be a member!

Hi,

I want to clarify one important issue since this discussion took a
completely different turn that the technical one I was looking for.

We are NOT implementing the NTP protocol, we are using the reference
implementation straight out of distribution>. I thought that was clear from the syntax of
my configuration and the fact that we use ntpq/ntpdc... ;-)

The question I had was, in other words, should I expect the poll
interval reported by ntpq to increase when using ntpd 4.2.4? And,
if it is to be expected, how fast should it increase? (This is
in the case when ntpd is in synchronization with a server and
the communication breaks down, the server is "down".) (The
reason I named NTP v3 is that we have configured the daemon
to use NTP v3 explicitly..)

I just want to say that I don't mind the license discussion, it's
an important subject to discuss. However; I'm biased and I really
prefer the GPL.. ;-)

Best regards, Anton

-----Original Message-----
From: questions-bounces+anton.a.persson=ericsson.com@lists.ntp.org
[mailto:questions-bounces+anton.a.persson=ericsson.com@lists.ntp.org] On
Behalf Of Harlan Stenn
Sent: den 13 juni 2008 20:44
To: questions@lists.ntp.org
Subject: Re: poll interval - RFC compliance question

Anton,

You are familiar with the draft NTPv4 spec, which includes SNTP, right?

More info at http://support.ntp.org/IETF/WebHome .

Several of your questions would be useful on the IETF NTP workgroup
mailing list, and would be useful and interesting topics for the Support
or Dev areas of http://support.ntp.org .

Finally, I encourage you and everybody else who is working for a company
that has an interest in NTP to get your company to join the the NTP
Forum, http://ntpforum.isc.org . I am happy to do whatever I can to
make membership in the NTP Forum a valuable experience.
--
Harlan Stenn
http://ntpforum.isc.org - be a member!

_______________________________________________
questions mailing list
questions@lists.ntp.org
https://lists.ntp.org/mailman/listinfo/questions

David Woolley writes:

>Unruh wrote:
>> David Woolley writes:
>>
>> To quote the license
>> * Permission to use, copy, modify, and distribute this software and *
>> * its documentation for any purpose with or without fee is hereby *
>> * granted
>>
>> I find it hard to develope a legal theory which makes this restrictive.

>It has already been established that the licence has changed since
>OpenBSD made their objections, so they were not objecting to the exact
>licence from which you are quoting.

Mills claimed that the license has been the same for 10 years now.


>>
>>> Licence: David J Woolley grants permission for all acts covered by
>>> copyright law which contributors to a USENET newsroup gatewayed to a
>>> public mailing list would normally expect to be permitted.
>>
>> Yee gads. That would be terrible. It would create a legal morass in trying
>> to define "normally expect to be permitted" Normal for whom, expect by

>But that is exactly what one is doing by posting to USENET without an
>explicit licence. I was simply trying to state the implied licence,
>which is very subjective.

jra@febo.com (John Ackermann N8UR) writes:

>Unruh said the following on 06/15/2008 11:35 AM:

>> Yes, it is pedantry, since the copyright is automatic and does not need to
>> be asserted. Ie, that notice is infomational, not a legal requirement.
>> It simply informs the reader as to who actually owns the copyright. What is
>> not mere pedantry is where or not this claim is actually true. Many people
>> have contributed to ntpd, and unless they all transfered copyright to David
>> Mills, then the claim that he owns the copyright is really false. They all
>> do (David has copyright interest in all of the works since the other
>> people's work is derivative of his work, but others have copyright interest
>> as well.) To keep things perfectly clean, David shoule ask anyone who
>> contributes to transfer their copyright to him.

>While the notice is not mandatory, it does have legal value -- in
>particular, it negates a defense of innocent infringement. And, to
>claim the international protection of the UCC treaty I referred to in my
>other message, you do need to include the notice.

>Your point about whether the copyright notice is accurate is
>interesting. I wonder if it would even be possible at this point to
>determine who all the contributors are!

>I thought that Linux used the CREDITS file as a list of copyright
>holders, but I just looked at some source and that doesn't appear to be
>the case -- instead, each source file has its own copyright notice in
>the name of one or more people, with the COPYING file not listing any
>copyright owners at all. I guess that has the effect of putting the
>whole kernel tree under the GPL, but the resulting object file(s) as a
>collective work of all the contributors.

>I just looked at the ntp 4.2.0 source (the only tree I have handy at the
>moment) and the random source files I looked at do not have any
>copyright notice at all, so the Linux model doesn't appear to be in place.

In the COPURIGHT file containing Mill's claim of copyright holder, is also
a long list of authors which that files asserts as authors of various parts
of ntp. Now if those authors actually tranfered copyright to Mills, then
this is fine, but I tend to doubt it. In that case all those authors are
the actual copyright holders or share the copyright under derived works.



>John

anton.a.persson@ericsson.com (Anton Persson A) writes:

>Hi,

>I want to clarify one important issue since this discussion took a
>completely different turn that the technical one I was looking for.

>We are NOT implementing the NTP protocol, we are using the reference
>implementation straight out of >distribution>. I thought that was clear from the syntax of
>my configuration and the fact that we use ntpq/ntpdc... ;-)

The distros may change ntp. Ie, they are usually NOT the reference
implimentation.


>The question I had was, in other words, should I expect the poll
>interval reported by ntpq to increase when using ntpd 4.2.4? And,

Yes.

>if it is to be expected, how fast should it increase? (This is
>in the case when ntpd is in synchronization with a server and
>the communication breaks down, the server is "down".) (The
>reason I named NTP v3 is that we have configured the daemon
>to use NTP v3 explicitly..)

Sorry, you told us that you are using some unknown version from some
unknown distribution, and then ask about NTPV3. Is the implimentation an
implimentation of the v3 or v4? Tell us the version number and the distro.
Why are you so coy about the details?

I think that on a lost lock, ntp first decreases the poll interval, on the
theory that the loss was temporary, and that the next time it will be
there. If that does not work, then the poll interval is increased.
But I think the min and max poll are still honoured. Ie with a maxpoll of
10 it does not wait for 8 days (poll 16) before trying again.


>I just want to say that I don't mind the license discussion, it's
>an important subject to discuss. However; I'm biased and I really
>prefer the GPL.. ;-)

The license is far less restrictive than GPL.



> Best regards, Anton

>-----Original Message-----
>From: questions-bounces+anton.a.persson=ericsson.com@lists.ntp.org
>[mailto:questions-bounces+anton.a.persson=ericsson.com@lists.ntp.org] On
>Behalf Of Harlan Stenn
>Sent: den 13 juni 2008 20:44
>To: questions@lists.ntp.org
>Subject: Re: poll interval - RFC compliance question

>Anton,

>You are familiar with the draft NTPv4 spec, which includes SNTP, right?

>More info at http://support.ntp.org/IETF/WebHome .

>Several of your questions would be useful on the IETF NTP workgroup
>mailing list, and would be useful and interesting topics for the Support
>or Dev areas of http://support.ntp.org .

>Finally, I encourage you and everybody else who is working for a company
>that has an interest in NTP to get your company to join the the NTP
>Forum, http://ntpforum.isc.org . I am happy to do whatever I can to
>make membership in the NTP Forum a valuable experience.
>--
>Harlan Stenn
>http://ntpforum.isc.org - be a member!

>_______________________________________________
>questions mailing list
>questions@lists.ntp.org
>https://lists.ntp.org/mailman/listinfo/questions

On 2008-06-16, Anton Persson A wrote:

> The question I had was, in other words, should I expect the poll
> interval reported by ntpq to increase when using ntpd 4.2.4?

Yes.

> And, if it is to be expected, how fast should it increase?

Please see http://doc.ntp.org/4.2.4/ntpd.html#poll for the discussion of
Poll Interval Control pertaining to NTP 4.2.4

You may also wish to review the Poll Rate Control section of
http://www.cis.udel.edu/~mills/ntp/html/rate.html (please note that all
documentation found in this tree applies to the current NTP development
snapshot).

You will also find the documentation for your release of NTP in the
../html directory of the appropriate source tarball.

> (This is in the case when ntpd is in synchronization with a server and
> the communication breaks down, the server is "down".)

When time sources become unreachable ntpd will continue to discipline
the clock using the currently calculated frequency correction value.
If that ntpd is configured with Orphan Mode, or with the Undisciplined
Local Clock, it will be able to continue serving time to others.

--
Steve Kostecke
NTP Public Services Project - http://support.ntp.org/

On 2008-06-16, Unruh wrote:

> anton.a.persson@ericsson.com (Anton Persson A) writes:
>
>>We are NOT implementing the NTP protocol, we are using the reference
>>implementation straight out of >>distribution>. I thought that was clear from the syntax of my
>>configuration and the fact that we use ntpq/ntpdc... ;-)
>
> The distros may change ntp. Ie, they are usually NOT the reference
> implimentation.

It is not uncommon for OSes to patch software to resolve build issues
(e.g. fix compilation problems, adjust for non-standard tool-chains,
customize file locations, etc.)

> I think that on a lost lock, ntp first decreases the poll interval, on
> the theory that the loss was temporary, and that the next time it will
> be there. If that does not work, then the poll interval is increased.
> But I think the min and max poll are still honoured. Ie with a maxpoll
> of 10 it does not wait for 8 days (poll 16) before trying again.

You may wish to read the pertinent portion of the documentation for the
OP's version of NTP:

http://doc.ntp.org/4.2.4/ntpd.html#poll

--
Steve Kostecke
NTP Public Services Project - http://support.ntp.org/

Hi,

Since I was still a bit too vague in my last post i want to clarify
further:

What we are using is the Red Hat ntp package, for ntp v 4.2.4. But
we have removed the ntp-4.2.4p0-sleep.patch since it caused a
segmentation fault.
(This has been reported to one of the red hat mailing lists, however I
have
yet to receive any reply from them about it..)

We have, however, in the /etc/ntp.conf file explicitly stated that the
daemon
should communicate using NTP v3. What this means is that ntpd says, to
to the
server it is synching with, that it uses NTP v3 (even though it is
capable of v4.)
I have no idea how this affects the different state machines running
inside the
ntpd, just that the packet it sends out on the network state v3 instead
of the
default v4. The reasons for this is requirements from a legacy project
which we
are communicating with..

The poll interval is 64 to 1024, seconds. ntpd starts up, locks on
target, stays
synchronized for a while, then we pull the plug. The poll interval is at
this time
still 64 seconds, how long should it take to step up the interval to 128
seconds for
a completely failed link?

This is with the redhat version of ntp as said above, and with no other
configuration than:

server 127.127.1.0 # local clock
fudge 127.127.1.0 stratum 10

driftfile /var/lib/ntp/drift

server 192.168.10.207 version 3

ntpd is started with the following options (-g is because we have no RTC
and so the offset we start with is very very large..):

ntpd -u ntp:ntp -p /var/run/ntpd.pid -g

Best regards,
Anton

-----Original Message-----
From: questions-bounces+anton.a.persson=ericsson.com@lists.ntp.org
[mailto:questions-bounces+anton.a.persson=ericsson.com@lists.ntp.org] On
Behalf Of Unruh
Sent: den 16 juni 2008 19:30
To: questions@lists.ntp.org
Subject: Re: poll interval - Clarification! - RFC
compliancequestion

anton.a.persson@ericsson.com (Anton Persson A) writes:

>Hi,

>I want to clarify one important issue since this discussion took a
>completely different turn that the technical one I was looking for.

>We are NOT implementing the NTP protocol, we are using the reference
>implementation straight out of >distribution>. I thought that was clear from the syntax of
>my configuration and the fact that we use ntpq/ntpdc... ;-)

The distros may change ntp. Ie, they are usually NOT the reference
implimentation.


>The question I had was, in other words, should I expect the poll
>interval reported by ntpq to increase when using ntpd 4.2.4? And,

Yes.

>if it is to be expected, how fast should it increase? (This is
>in the case when ntpd is in synchronization with a server and
>the communication breaks down, the server is "down".) (The
>reason I named NTP v3 is that we have configured the daemon
>to use NTP v3 explicitly..)

Sorry, you told us that you are using some unknown version from some
unknown distribution, and then ask about NTPV3. Is the implimentation an
implimentation of the v3 or v4? Tell us the version number and the
distro.
Why are you so coy about the details?

I think that on a lost lock, ntp first decreases the poll interval, on
the
theory that the loss was temporary, and that the next time it will be
there. If that does not work, then the poll interval is increased.
But I think the min and max poll are still honoured. Ie with a maxpoll
of
10 it does not wait for 8 days (poll 16) before trying again.


>I just want to say that I don't mind the license discussion, it's
>an important subject to discuss. However; I'm biased and I really
>prefer the GPL.. ;-)

The license is far less restrictive than GPL.



> Best regards, Anton

>-----Original Message-----
>From: questions-bounces+anton.a.persson=ericsson.com@lists.ntp.org
>[mailto:questions-bounces+anton.a.persson=ericsson.com@lists.ntp.org]
On
>Behalf Of Harlan Stenn
>Sent: den 13 juni 2008 20:44
>To: questions@lists.ntp.org
>Subject: Re: poll interval - RFC compliance question

>Anton,

>You are familiar with the draft NTPv4 spec, which includes SNTP, right?

>More info at http://support.ntp.org/IETF/WebHome .

>Several of your questions would be useful on the IETF NTP workgroup
>mailing list, and would be useful and interesting topics for the
Support
>or Dev areas of http://support.ntp.org .

>Finally, I encourage you and everybody else who is working for a
company
>that has an interest in NTP to get your company to join the the NTP
>Forum, http://ntpforum.isc.org . I am happy to do whatever I can to
>make membership in the NTP Forum a valuable experience.
>--
>Harlan Stenn
>http://ntpforum.isc.org - be a member!

>_______________________________________________
>questions mailing list
>questions@lists.ntp.org
>https://lists.ntp.org/mailman/listinfo/questions

On 2008-06-17, Anton Persson A wrote:

> The poll interval is 64 to 1024, seconds. ntpd starts up, locks on
> target, stays synchronized for a while, then we pull the plug.

What are you trying to prove by "pulling the plug"?

> The poll interval is at this time still 64 seconds, how long should it
> take to step up the interval to 128 seconds for a completely failed
> link?

Why would you expect the poll interval to change if ntpd is not
receiving replies to polls?

--
Steve Kostecke
NTP Public Services Project - http://support.ntp.org/

Steve Kostecke wrote:
> On 2008-06-17, Anton Persson A wrote:
>
>> The poll interval is 64 to 1024, seconds. ntpd starts up, locks on
>> target, stays synchronized for a while, then we pull the plug.
>
> What are you trying to prove by "pulling the plug"?
>
>> The poll interval is at this time still 64 seconds, how long should it
>> take to step up the interval to 128 seconds for a completely failed
>> link?
>
> Why would you expect the poll interval to change if ntpd is not
> receiving replies to polls?
>

I think he may be talking about the "exponential backoff" required by
the RFCs when a server is not responding. Don't know just how that was
implemented but that may happen only when initializing an association.
This was a change made a couple of years ago in response to the Netgear
and DLink debacles.