IPv4 and v6 after CDNs (and before A.I.
16 views
Skip to first unread message
Sivasubramanian Muthusamy
Jul 28, 2025, 8:49:29 PM7/28/25
to coreinter...@googlegroups.com, Olivier MJ Crepin-Leblond
( This link is NOT shared in a manner of critiquing CDNs, but to see if we can examine the changing perception about the importance to end user's for IP addresses, especially static IPs. )
Joly MacFie
Jul 29, 2025, 2:45:45 AM7/29/25
to coreinter...@googlegroups.com, Olivier MJ Crepin-Leblond
I actually asked about precisely this issue at the CIV's 3rd meeting in Baku, 2012. I got a complete non-answer from the panelists.
I have dug up the video https://youtu.be/Je36zEtQSvk?t=2806 - and here is the transcript FWIW.
>> REMOTE MODERATOR: Thank you very much. As a follow-up to previous questions, we got several questions from our remote participants. First, a question was from Joly MacFie. As entertainment is increasingly delivered via content distribution networks, how does this affect peering arrangements and the end-to-end principle, as users access content rather than hosts?
The next question was from the United States, from Marcus Ledbetter. Do we all agree that this is just one Internet?
And the last one, was to Mr. Vint Cerf, balancing sovereignty, openeness, regulation, and national laws, seems to me a very tricky job to do. So my question to Vint Cerf, which body do you think would have the task to manage this complex task?
>> VINT CERF: OK. Shall I try to answer the last one? Maybe this Dynamic Coalition is where that solution starts. Maybe this is a group that can begin examining what's possible and what isn't. It's pretty clear, though, if you are going to have international agreements that create a kind of homologized legal framework, that ultimately you will have to go to bodies like the World Trade Organization, or the Worls Intellectual Property Organization, or other parts of the UN -- or you are going to have to go to a collection of multilateral treaties in order to establish agreement. I think we will probably end up starting with the lowest common denominator, simple things. For example, what does a notarization mean, and what's a digital signature mean, and does it have common weight in all countries? We're going to have to build this up a little bit at a time. I don't think there's one body that solves all the problems.
>> LYNN ST. AMOUR: There were two other questions that were posed. One was do we all agree there is one Internet and the other has to do with content and peer to peer and whether the impact on the end-to-end.
So I'm sure Vint's ready to jump in and respond to that. But is there anyone else who wants to -- Nick, and then Paul.
>> NICK ASHTON-HART This is Nick Ashton-Hart. So on the content question, I'll take that one. I will be cursed for the rest of my life in dealing with -- with copyrighted material and what happens to it, given that I was a music manager for over 20 years, off and on. This is the great -- this is a perfect example of the clash between sovereignty law and the real world of the Internet and how it's really used. The copyright system is a national system and it's implemented different in different countries and yet cloud computing by its nature means that you access the same resource, two different times in the same day and you are accessing multiple different servers in multiple different countries on each of those occasions.
And the application -- how to deal with the legal issues there. There has been a treaty negotiation going on in Europe for 50 years to try to determine how international law and private law, the law of individual countries, works together? And they have been unable to agree this. This is an enormously thorny question. I think it's certainly true that the desire for enforcement has an impact on what people can access. We can see that the iTunes store has different material at different times. And I do think we're going to have to come up with some way to internationalize the way in which rights -- national rights work in an international environment. There's going to have to be some way around that. It's not just for entertainment contentment but simply for the efficient functioning for services upon which increasingly large amounts of the economy rely. Pfizer, one of the world's largest drug companies, recently transferred its entire supply chain and directed all of the vendors to a cloud-based system so that they can see in realtime absolutely everything about their product. Where they are being made, where they are being shipped, where they are running out of them? This is going to become increasingly the case, and the more of the world that is integrated in that way, the more of which the conflicts of laws become very difficult. There is going to have to be some conception of how laws work on the Internet. And I think the 50-year conversation will end much sooner -- it won't take another 50 years because the commercial realities of dealing with this will require a pragmatic result that wasn't required by the situation over the last 50 years. It was an academic discussion for 50 years because it could be. Now it's not academic anymore.
>> LYNN ST. AMOUR: Thank you, Nick. Paul?
>> PAUL WILSON: I wanted to answer the question about one Internet in a slightly different way, and that's -- but it's a way that depends on how you define the Internet in asking the question. Because I used the term loosely before in terms of how, what would the Internet be like in ten years down the track, and would it become a different Internet. In that case, the Internet's kind of everything, it's the universe that we are talking about. There's only just one of those. But if you start to drill down through that either through the level of users, or content, or applications, then it's really -- the Internet is in the eye of the beholder, and I think it's in all of those layers that we start to get confused in Internet Governance. What are we really talking about? There's the broad definition, there's the narrow definition. But actually speaking technically the Internet is the transport layer of the network that we are talking about. It's the thing that I was referring to before that is the single global neutral network that allows any point to connect to any other point, and actually that thing is in its ideal form that we are all working to preserve. It is one network. And that is the beauty of it. So let's not sort of mix up ourselves too much about saying which Internet we are talking about, and yes there are many, or yes there are none, because if you want to be quite specific about the Internet layer of the network that we all enjoy, the Internet layer is the transport layer. There has to be just one of those and it's really not a matter of perspective, it really is -- is simply the technical infrastructure and that is something that, as within this discussion about values, we should really identify, as I say, which Internet we are talking about and be quite precise about that. Thanks.
>> VINT CERF: So it's Vint again. I would like to make a small nuance here. We all understand that the Internet protocols don't necessarily have to be used in the global interconnected sytem. People have used these same prorocols to build private networks. But I don't consider those to be capital-I Internet, those are lowercase-I, clones that don't have the same scope and probably have different intent. I wanted to come back to this question of rights management and dealing with intellectual property in adigital environment. It occurs to me that, if we treat content as digital objects for just a moment, not differentiating what they are, whether they are books, novels or some game or some other thing, piece of software, just imagine them as bags full of bits.
And if we thought that it was possible to build mechanisms for access control to those bags of bits so there was some form of enforcement for access and use, if we thought it was possible to achieve that, then we might actually come to a general purpose solution to the problem of --that you were talking about, Nick.
And so, I think there may be technical mechanisms that might be implemented to make access to digital content, and digital objects of all kinds, manageable. And here, if we were able to demonstrate that you could establish whatever terms and conditions you wished and these are for access and use, and if those terms and conditions could really be enforced, technically enforced, then many of the problems that have arisen in the national context of copyright, for instance, would evaporate, and be assimilated into this more general system.
>> REMOTE MODERATOR: Thank you very much. As a follow-up to previous questions, we got several questions from our remote participants. First, a question was from Joly MacFie. As entertainment is increasingly delivered via content distribution networks, how does this affect peering arrangements and the end-to-end principle, as users access content rather than hosts?
The next question was from the United States, from Marcus Ledbetter. Do we all agree that this is just one Internet?
And the last one, was to Mr. Vint Cerf, balancing sovereignty, openeness, regulation, and national laws, seems to me a very tricky job to do. So my question to Vint Cerf, which body do you think would have the task to manage this complex task?
>> VINT CERF: OK. Shall I try to answer the last one? Maybe this Dynamic Coalition is where that solution starts. Maybe this is a group that can begin examining what's possible and what isn't. It's pretty clear, though, if you are going to have international agreements that create a kind of homologized legal framework, that ultimately you will have to go to bodies like the World Trade Organization, or the Worls Intellectual Property Organization, or other parts of the UN -- or you are going to have to go to a collection of multilateral treaties in order to establish agreement. I think we will probably end up starting with the lowest common denominator, simple things. For example, what does a notarization mean, and what's a digital signature mean, and does it have common weight in all countries? We're going to have to build this up a little bit at a time. I don't think there's one body that solves all the problems.
>> LYNN ST. AMOUR: There were two other questions that were posed. One was do we all agree there is one Internet and the other has to do with content and peer to peer and whether the impact on the end-to-end.
So I'm sure Vint's ready to jump in and respond to that. But is there anyone else who wants to -- Nick, and then Paul.
>> NICK ASHTON-HART This is Nick Ashton-Hart. So on the content question, I'll take that one. I will be cursed for the rest of my life in dealing with -- with copyrighted material and what happens to it, given that I was a music manager for over 20 years, off and on. This is the great -- this is a perfect example of the clash between sovereignty law and the real world of the Internet and how it's really used. The copyright system is a national system and it's implemented different in different countries and yet cloud computing by its nature means that you access the same resource, two different times in the same day and you are accessing multiple different servers in multiple different countries on each of those occasions.
And the application -- how to deal with the legal issues there. There has been a treaty negotiation going on in Europe for 50 years to try to determine how international law and private law, the law of individual countries, works together? And they have been unable to agree this. This is an enormously thorny question. I think it's certainly true that the desire for enforcement has an impact on what people can access. We can see that the iTunes store has different material at different times. And I do think we're going to have to come up with some way to internationalize the way in which rights -- national rights work in an international environment. There's going to have to be some way around that. It's not just for entertainment contentment but simply for the efficient functioning for services upon which increasingly large amounts of the economy rely. Pfizer, one of the world's largest drug companies, recently transferred its entire supply chain and directed all of the vendors to a cloud-based system so that they can see in realtime absolutely everything about their product. Where they are being made, where they are being shipped, where they are running out of them? This is going to become increasingly the case, and the more of the world that is integrated in that way, the more of which the conflicts of laws become very difficult. There is going to have to be some conception of how laws work on the Internet. And I think the 50-year conversation will end much sooner -- it won't take another 50 years because the commercial realities of dealing with this will require a pragmatic result that wasn't required by the situation over the last 50 years. It was an academic discussion for 50 years because it could be. Now it's not academic anymore.
>> LYNN ST. AMOUR: Thank you, Nick. Paul?
>> PAUL WILSON: I wanted to answer the question about one Internet in a slightly different way, and that's -- but it's a way that depends on how you define the Internet in asking the question. Because I used the term loosely before in terms of how, what would the Internet be like in ten years down the track, and would it become a different Internet. In that case, the Internet's kind of everything, it's the universe that we are talking about. There's only just one of those. But if you start to drill down through that either through the level of users, or content, or applications, then it's really -- the Internet is in the eye of the beholder, and I think it's in all of those layers that we start to get confused in Internet Governance. What are we really talking about? There's the broad definition, there's the narrow definition. But actually speaking technically the Internet is the transport layer of the network that we are talking about. It's the thing that I was referring to before that is the single global neutral network that allows any point to connect to any other point, and actually that thing is in its ideal form that we are all working to preserve. It is one network. And that is the beauty of it. So let's not sort of mix up ourselves too much about saying which Internet we are talking about, and yes there are many, or yes there are none, because if you want to be quite specific about the Internet layer of the network that we all enjoy, the Internet layer is the transport layer. There has to be just one of those and it's really not a matter of perspective, it really is -- is simply the technical infrastructure and that is something that, as within this discussion about values, we should really identify, as I say, which Internet we are talking about and be quite precise about that. Thanks.
>> VINT CERF: So it's Vint again. I would like to make a small nuance here. We all understand that the Internet protocols don't necessarily have to be used in the global interconnected sytem. People have used these same prorocols to build private networks. But I don't consider those to be capital-I Internet, those are lowercase-I, clones that don't have the same scope and probably have different intent. I wanted to come back to this question of rights management and dealing with intellectual property in adigital environment. It occurs to me that, if we treat content as digital objects for just a moment, not differentiating what they are, whether they are books, novels or some game or some other thing, piece of software, just imagine them as bags full of bits.
And if we thought that it was possible to build mechanisms for access control to those bags of bits so there was some form of enforcement for access and use, if we thought it was possible to achieve that, then we might actually come to a general purpose solution to the problem of --that you were talking about, Nick.
And so, I think there may be technical mechanisms that might be implemented to make access to digital content, and digital objects of all kinds, manageable. And here, if we were able to demonstrate that you could establish whatever terms and conditions you wished and these are for access and use, and if those terms and conditions could really be enforced, technically enforced, then many of the problems that have arisen in the national context of copyright, for instance, would evaporate, and be assimilated into this more general system.
--
You received this message because you are subscribed to the Google Groups "coreinternetvalues" group.
To unsubscribe from this group and stop receiving emails from it, send an email to coreinternetval...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/coreinternetvalues/CAB%2Bxu1FpZmbJTE%2BEkCXHWWhduTh%3DwQ5cmyvOWQW%2B--hQWvtSoQ%40mail.gmail.com.
--------------------------------------
Joly MacFie +12185659365
Joly MacFie +12185659365
--------------------------------------
-
Vint Cerf
Jul 29, 2025, 3:38:50 AM7/29/25
to coreinter...@googlegroups.com, Olivier MJ Crepin-Leblond
1. the article is wrong. You still need to translate a domain name into an IP address to use a CDN. the fact that the IP address is "computed" using information from the source IP address of the DNS query is distinct from the fact that you still need either QUIC/UDP/IP or TCP/IP to carry the traffic to/from the CDN server.
2. Joly, exactly what question did you ask that you thought you got a non-answer to?
vint
To view this discussion visit https://groups.google.com/d/msgid/coreinternetvalues/CAM9VJk3qexaHnzL%3DK3x06Qjx4iMFMgKv%2BjC6vL%3D42kYv2Zbcxw%40mail.gmail.com.
Please send any postal/overnight deliveries to:
until further notice
Dr. Alejandro Pisanty Baruch
Jul 29, 2025, 3:42:25 AM7/29/25
to coreinter...@googlegroups.com, Olivier MJ Crepin-Leblond
So Mr. Muthusamy,
what is your question? The article is seriously criticized in the comments of the original publication for several, important errors (though some of the comments may contain questionable elements, in turn.)
Vint has already explained it here so, I repeat, what was the original question, and what remains of it after Vint's explanation?
Alejandro Pisanty
De: coreinter...@googlegroups.com <coreinter...@googlegroups.com> en nombre de Sivasubramanian Muthusamy <Inte...@nameshop.in>
Enviado: lunes, 28 de julio de 2025 06:49 p. m.
Para: coreinter...@googlegroups.com
CC: Olivier MJ Crepin-Leblond
Asunto: [CoreInternetValues] IPv4 and v6 after CDNs (and before A.I.
Enviado: lunes, 28 de julio de 2025 06:49 p. m.
Para: coreinter...@googlegroups.com
CC: Olivier MJ Crepin-Leblond
Asunto: [CoreInternetValues] IPv4 and v6 after CDNs (and before A.I.
--
Olivier MJ Crépin-Leblond
Jul 29, 2025, 5:14:13 AM7/29/25
to Vint Cerf, coreinter...@googlegroups.com
On 29/07/2025 08:38, Vint Cerf wrote:
1. the article is wrong. You still need to translate a domain name into an IP address to use a CDN. the fact that the IP address is "computed" using information from the source IP address of the DNS query is distinct from the fact that you still need either QUIC/UDP/IP or TCP/IP to carry the traffic to/from the CDN server.
For more information about the Google CDN, there's a page that
explains it:
https://cloud.google.com/cdn/docs/overview
There's also a page that explains how to set-up a managed instance group back end. I do not know if it is publicly available: https://cloud.google.com/cdn/docs/setting-up-cdn-with-mig
And of course, everything uses TCP-IP. The Google global Network is native on IPv6.
The difference that CDNs have made to the Internet is to introduce a geographical element to DNS queries directing the user to a different cache depending on geography and also including other elements of its CDN.
I have concerns that articles like the one Siva referred to can then become sources for AI to hallucinate.
Kindest regards,
Olivier
https://cloud.google.com/cdn/docs/overview
There's also a page that explains how to set-up a managed instance group back end. I do not know if it is publicly available: https://cloud.google.com/cdn/docs/setting-up-cdn-with-mig
And of course, everything uses TCP-IP. The Google global Network is native on IPv6.
The difference that CDNs have made to the Internet is to introduce a geographical element to DNS queries directing the user to a different cache depending on geography and also including other elements of its CDN.
I have concerns that articles like the one Siva referred to can then become sources for AI to hallucinate.
Kindest regards,
Olivier
Christian de Larrinaga
Jul 29, 2025, 7:48:16 AM7/29/25
to 'Vint Cerf' via coreinternetvalues, Olivier MJ Crepin-Leblond
I wonder where this idea is coming from?
Even content networks such as IPFS - IPNS that use distributed hash
tables (DHTs) to refer to content addresses route over the Internet
layer via IP and optionally DNS.
PS disclosure I am not keen on so called CDNs which are introducing
layers of disintermediation and dependencies to run stuff through DNS
that really should be standardised and controlled at and by the user
edge.
</rant>
"'Vint Cerf' via coreinternetvalues" <coreinter...@googlegroups.com> writes:
> 1. the article is wrong. You still need to translate a domain name into an
> IP address to use a CDN. the fact that the IP address is "computed" using
> information from the source IP address of the DNS query is distinct from
> the fact that you still need either QUIC/UDP/IP or TCP/IP to carry the
> traffic to/from the CDN server.
>
> 2. Joly, exactly what question did you ask that you thought you got a
> non-answer to?
>
> vint
>
>
> On Tue, Jul 29, 2025 at 2:45 AM Joly MacFie <jo...@punkcast.com> wrote:
>
>> I actually asked about precisely this issue at the CIV's 3rd meeting in
>> Baku, 2012. I got a complete non-answer from the panelists.
>>
>> I have dug up the video https://youtu.be/Je36zEtQSvk?t=2806 - and here is
>> the transcript FWIW.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> questions, we got several questions from our remote participants. First, a
>> question was from Joly MacFie. As entertainment is increasingly delivered
>> via content distribution networks, how does this affect peering
>> arrangements and the end-to-end principle, as users access content rather
>> than hosts?The next question was from the United States, from Marcus
>> question was from Joly MacFie. As entertainment is increasingly delivered
>> via content distribution networks, how does this affect peering
>> arrangements and the end-to-end principle, as users access content rather
>> Ledbetter. Do we all agree that this is just one Internet?And the last
>> one, was to Mr. Vint Cerf, balancing sovereignty, openeness, regulation,
>> and national laws, seems to me a very tricky job to do. So my question to
>> Vint Cerf, which body do you think would have the task to manage this
>> complex task?>> VINT CERF: OK. Shall I try to answer the last one? Maybe
>> this Dynamic Coalition is where that solution starts. Maybe this is a
>> group that can begin examining what's possible and what isn't. It's pretty
>> clear, though, if you are going to have international agreements that
>> create a kind of homologized legal framework, that ultimately you will have
>> to go to bodies like the World Trade Organization, or the Worls
>> Intellectual Property Organization, or other parts of the UN -- or you are
>> going to have to go to a collection of multilateral treaties in order to
>> establish agreement. I think we will probably end up starting with the
>> lowest common denominator, simple things. For example, what does a
>> notarization mean, and what's a digital signature mean, and does it have
>> common weight in all countries? We're going to have to build this up a
>> little bit at a time. I don't think there's one body that solves all the
>> problems.>> LYNN ST. AMOUR: There were two other questions that were
>> posed. One was do we all agree there is one Internet and the other has to
>> do with content and peer to peer and whether the impact on the
>> end-to-end.So I'm sure Vint's ready to jump in and respond to that. But is
>> and national laws, seems to me a very tricky job to do. So my question to
>> Vint Cerf, which body do you think would have the task to manage this
>> complex task?>> VINT CERF: OK. Shall I try to answer the last one? Maybe
>> this Dynamic Coalition is where that solution starts. Maybe this is a
>> group that can begin examining what's possible and what isn't. It's pretty
>> clear, though, if you are going to have international agreements that
>> create a kind of homologized legal framework, that ultimately you will have
>> to go to bodies like the World Trade Organization, or the Worls
>> Intellectual Property Organization, or other parts of the UN -- or you are
>> going to have to go to a collection of multilateral treaties in order to
>> establish agreement. I think we will probably end up starting with the
>> lowest common denominator, simple things. For example, what does a
>> notarization mean, and what's a digital signature mean, and does it have
>> common weight in all countries? We're going to have to build this up a
>> little bit at a time. I don't think there's one body that solves all the
>> problems.>> LYNN ST. AMOUR: There were two other questions that were
>> posed. One was do we all agree there is one Internet and the other has to
>> do with content and peer to peer and whether the impact on the
>> there anyone else who wants to -- Nick, and then Paul.>> NICK ASHTON-HART
>> This is Nick Ashton-Hart. So on the content question, I'll take that one.
>> I will be cursed for the rest of my life in dealing with -- with
>> copyrighted material and what happens to it, given that I was a music
>> manager for over 20 years, off and on. This is the great -- this is a
>> perfect example of the clash between sovereignty law and the real world of
>> the Internet and how it's really used. The copyright system is a national
>> system and it's implemented different in different countries and yet cloud
>> computing by its nature means that you access the same resource, two
>> different times in the same day and you are accessing multiple different
>> servers in multiple different countries on each of those occasions.And the
>> This is Nick Ashton-Hart. So on the content question, I'll take that one.
>> I will be cursed for the rest of my life in dealing with -- with
>> copyrighted material and what happens to it, given that I was a music
>> manager for over 20 years, off and on. This is the great -- this is a
>> perfect example of the clash between sovereignty law and the real world of
>> the Internet and how it's really used. The copyright system is a national
>> system and it's implemented different in different countries and yet cloud
>> computing by its nature means that you access the same resource, two
>> different times in the same day and you are accessing multiple different
>> it was possible to build mechanisms for access control to those bags of
>> bits so there was some form of enforcement for access and use, if we
>> thought it was possible to achieve that, then we might actually come to a
>> general purpose solution to the problem of --that you were talking about,
>> Nick.And so, I think there may be technical mechanisms that might be
>> bits so there was some form of enforcement for access and use, if we
>> thought it was possible to achieve that, then we might actually come to a
>> general purpose solution to the problem of --that you were talking about,
>> implemented to make access to digital content, and digital objects of all
>> kinds, manageable. And here, if we were able to demonstrate that you could
>> establish whatever terms and conditions you wished and these are for access
>> and use, and if those terms and conditions could really be enforced,
>> technically enforced, then many of the problems that have arisen in the
>> national context of copyright, for instance, would evaporate, and be
>> assimilated into this more general system.*
>> kinds, manageable. And here, if we were able to demonstrate that you could
>> establish whatever terms and conditions you wished and these are for access
>> and use, and if those terms and conditions could really be enforced,
>> technically enforced, then many of the problems that have arisen in the
>> national context of copyright, for instance, would evaporate, and be
>>
>> On Mon, Jul 28, 2025 at 8:49 PM Sivasubramanian Muthusamy <
>> Inte...@nameshop.in> wrote:
>>
>>>
>>> https://www.xda-developers.com/who-cares-about-ipv4-or-ipv6-when-cdns-run-the-internet/
>>>
>>> ( This link is NOT shared in a manner of critiquing CDNs, but to see if
>>> we can examine the changing perception about the importance to end user's
>>> for IP addresses, especially static IPs. )
>>>
>>> Sivasubramanian M
>>> Nameshop
>>> India.
>>> Inte...@nameshop.in
>>>
>>> --
>>> You received this message because you are subscribed to the Google Groups
>>> "coreinternetvalues" group.
>>> To unsubscribe from this group and stop receiving emails from it, send an
>>> email to coreinternetval...@googlegroups.com.
>>> To view this discussion visit
>>> https://groups.google.com/d/msgid/coreinternetvalues/CAB%2Bxu1FpZmbJTE%2BEkCXHWWhduTh%3DwQ5cmyvOWQW%2B--hQWvtSoQ%40mail.gmail.com
>>>
> <https://groups.google.com/d/msgid/coreinternetvalues/CAB%2Bxu1FpZmbJTE%2BEkCXHWWhduTh%3DwQ5cmyvOWQW%2B--hQWvtSoQ%40mail.gmail.com?utm_medium=email&utm_source=footer>
>> On Mon, Jul 28, 2025 at 8:49 PM Sivasubramanian Muthusamy <
>> Inte...@nameshop.in> wrote:
>>
>>>
>>> https://www.xda-developers.com/who-cares-about-ipv4-or-ipv6-when-cdns-run-the-internet/
>>>
>>> ( This link is NOT shared in a manner of critiquing CDNs, but to see if
>>> we can examine the changing perception about the importance to end user's
>>> for IP addresses, especially static IPs. )
>>>
>>> Sivasubramanian M
>>> Nameshop
>>> India.
>>> Inte...@nameshop.in
>>>
>>> --
>>> You received this message because you are subscribed to the Google Groups
>>> "coreinternetvalues" group.
>>> To unsubscribe from this group and stop receiving emails from it, send an
>>> email to coreinternetval...@googlegroups.com.
>>> To view this discussion visit
>>> https://groups.google.com/d/msgid/coreinternetvalues/CAB%2Bxu1FpZmbJTE%2BEkCXHWWhduTh%3DwQ5cmyvOWQW%2B--hQWvtSoQ%40mail.gmail.com
>>>
>>> .
>>>
>>
>>
>> --
>> --------------------------------------
>> Joly MacFie +12185659365 <(218)%20565-9365>
>> --------------------------------------
>> -
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "coreinternetvalues" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to coreinternetval...@googlegroups.com.
>> To view this discussion visit
>> https://groups.google.com/d/msgid/coreinternetvalues/CAM9VJk3qexaHnzL%3DK3x06Qjx4iMFMgKv%2BjC6vL%3D42kYv2Zbcxw%40mail.gmail.com
>>
> <https://groups.google.com/d/msgid/coreinternetvalues/CAM9VJk3qexaHnzL%3DK3x06Qjx4iMFMgKv%2BjC6vL%3D42kYv2Zbcxw%40mail.gmail.com?utm_medium=email&utm_source=footer>
>> -
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "coreinternetvalues" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to coreinternetval...@googlegroups.com.
>> To view this discussion visit
>> https://groups.google.com/d/msgid/coreinternetvalues/CAM9VJk3qexaHnzL%3DK3x06Qjx4iMFMgKv%2BjC6vL%3D42kYv2Zbcxw%40mail.gmail.com
>>
>> .
>>
>
>
> --
> Please send any postal/overnight deliveries to:
> Vint Cerf
> Google, LLC
> 1900 Reston Metro Plaza, 16th Floor
> Reston, VA 20190
> +1 (571) 213 1346
>
>
> until further notice
--
>
>
> --
> Please send any postal/overnight deliveries to:
> Vint Cerf
> Google, LLC
> 1900 Reston Metro Plaza, 16th Floor
> Reston, VA 20190
> +1 (571) 213 1346
>
>
> until further notice
Christian de Larrinaga
Joly MacFie
Jul 29, 2025, 10:14:38 AM7/29/25
to coreinter...@googlegroups.com, Olivier MJ Crepin-Leblond
Hi Vint,
My question was there:
My question was there:
>> First, a
>> question was from Joly MacFie. As entertainment is increasingly delivered
>> via content distribution networks, how does this affect peering
>> arrangements and the end-to-end principle, as users access content rather
>> than hosts
The answers addressed copyright rather than this.
Your response here was kind of what i was looking for
Your response here was kind of what i was looking for
> You still need to translate a domain name into an
> IP address to use a CDN. the fact that the IP address is "computed" using
> information from the source IP address of the DNS query is distinct from
> the fact that you still need either QUIC/UDP/IP or TCP/IP to carry the
> traffic to/from the CDN server.
> IP address to use a CDN. the fact that the IP address is "computed" using
> information from the source IP address of the DNS query is distinct from
> the fact that you still need either QUIC/UDP/IP or TCP/IP to carry the
> traffic to/from the CDN server.
It just seemed to me that this of violated the usercentricity principle, somewhat.
I have since seen some pretty eloquent stuff from Geoff Huston on why this matters, wholisticly.
I have since seen some pretty eloquent stuff from Geoff Huston on why this matters, wholisticly.
Joly
To view this discussion visit https://groups.google.com/d/msgid/coreinternetvalues/8734afdxqc.fsf%40firsthand.net.
--------------------------------------
Joly MacFie +12185659365
Joly MacFie +12185659365
--------------------------------------
-Dr. Alejandro Pisanty Baruch
Jul 29, 2025, 1:32:28 PM7/29/25
to 'Vint Cerf' via coreinternetvalues, Olivier MJ Crepin-Leblond
Christian,
the point in your second paragraph is indeed a cause of concern cfr. Core Values - potential interference with global reach, openness, etc.
One of the many flaws in the article is the statement that CDNs were started, or expanded, due to the scarcity of IPv4 addresses. CDNs - beginning maybe almost 30 years ago with Akamai - were introduced in order to rationalize traffic, reducing redundant traffic (many copies of the same data traveling through the same links mostly at the same time) and to reduce latency by placing copies of the files closer to where (and when) users would access them. They started in regions where IPv4 addresses were not in short supply at the time.
Another point the article misses is the many massive redistributions of IPv4 addresses that took place over the decades, with some organizations returning to the common pool huge chunks of IP addresses they had been allocated (like, a /8 i.e. one eighth of all IP addresses assigned to a single institution early in Internet history.)
I concur with Olivier in the concern that this kind of poorly informed opinion becomes part of the AI slop that should be a major concern for all.
Yours,
Alejandro Pisanty
De: coreinter...@googlegroups.com <coreinter...@googlegroups.com> en nombre de Christian de Larrinaga <cd...@firsthand.net>
Enviado: martes, 29 de julio de 2025 05:48 a. m.
Para: 'Vint Cerf' via coreinternetvalues
CC: Olivier MJ Crepin-Leblond
Asunto: Re: [CoreInternetValues] IPv4 and v6 after CDNs (and before A.I.
Enviado: martes, 29 de julio de 2025 05:48 a. m.
Para: 'Vint Cerf' via coreinternetvalues
CC: Olivier MJ Crepin-Leblond
Asunto: Re: [CoreInternetValues] IPv4 and v6 after CDNs (and before A.I.
To view this discussion visit
https://groups.google.com/d/msgid/coreinternetvalues/8734afdxqc.fsf%40firsthand.net.
Joly MacFie
Jul 30, 2025, 2:14:11 AM7/30/25
to coreinter...@googlegroups.com, Olivier MJ Crepin-Leblond
Dr. Alejandro Pisanty Baruch
Jul 30, 2025, 3:17:02 AM7/30/25
to coreinter...@googlegroups.com, Olivier MJ Crepin-Leblond
Joly,
thanks! really, immensely valuable.
I think we never got a team for the wiki and it didn't get started. Over the years contributions to our documents have come from precious few volunteers and we have handled them through word processors offline and document editors online. I wish we had a more robust participation and operation!
Alejandro Pisanty
De: coreinter...@googlegroups.com <coreinter...@googlegroups.com> en nombre de Joly MacFie <jo...@punkcast.com>
Enviado: miércoles, 30 de julio de 2025 12:13 a. m.
Para: coreinter...@googlegroups.com
Enviado: miércoles, 30 de julio de 2025 12:13 a. m.
Para: coreinter...@googlegroups.com
CC: Olivier MJ Crepin-Leblond
Asunto: Re: [CoreInternetValues] IPv4 and v6 after CDNs (and before A.I.
Christian de Larrinaga
Jul 30, 2025, 6:55:26 AM7/30/25
to Dr. Alejandro Pisanty Baruch, 'Vint Cerf' via coreinternetvalues, Olivier MJ Crepin-Leblond
"Dr.Alejandro Pisanty Baruch" <api...@unam.mx> writes:
I experienced much the same motivations. Early CDNS that I recall were
trying to cache closer to the user edge to solve two issues. Save long
distance transit cost and capacity used for their content by being able
to cache popular static content to improve performance for the majority
of users who were at the end of slow dial up / aDSL circuits.
Today with a growing number of people on 100Mb + circuits and
concentration of hosting / content into fewer and fewer behemoths that
deploy their own capacity into the user edge directly the motivation
seems to be shifting to abstracting away hosting infrastructure for app
and web developers.
It turns out that despite the design to enable "IP over anything and
Anything over IP". That today is an insufficient abstraction for content
and services which deploy via CDNs to manage all the complexities that
implementation practices and filters around the IP layer.
Steve Deering used to talk about the fattening of the IP layer breaking
the abstraction necessary for end to end Internet communications. He saw
IPv6 goal perhaps primary goal to keep the IP layer uncomplicated. The
way CDNs are used today is proof that fattening at and around the IP
layer is making self hosting and end to end deployment increasingly
hard. So CDNs such as Cloudflare and the large hosting infrastructure
providers offer services to integrate all this fatness to web and app
developers and publishers. One stop obesity as a service if you like.
This type of solution is a signal of fragmentation that is at least as
protracted as complexities of whether we prioritise a single DNS root or
not.
Indeed for CDNs which run their own DNS infrastructure which allows them
to off services running through their CDNs their own root data and
domains should they so choose.
C
--
Christian de Larrinaga
I experienced much the same motivations. Early CDNS that I recall were
trying to cache closer to the user edge to solve two issues. Save long
distance transit cost and capacity used for their content by being able
to cache popular static content to improve performance for the majority
of users who were at the end of slow dial up / aDSL circuits.
Today with a growing number of people on 100Mb + circuits and
concentration of hosting / content into fewer and fewer behemoths that
deploy their own capacity into the user edge directly the motivation
seems to be shifting to abstracting away hosting infrastructure for app
and web developers.
It turns out that despite the design to enable "IP over anything and
Anything over IP". That today is an insufficient abstraction for content
and services which deploy via CDNs to manage all the complexities that
implementation practices and filters around the IP layer.
Steve Deering used to talk about the fattening of the IP layer breaking
the abstraction necessary for end to end Internet communications. He saw
IPv6 goal perhaps primary goal to keep the IP layer uncomplicated. The
way CDNs are used today is proof that fattening at and around the IP
layer is making self hosting and end to end deployment increasingly
hard. So CDNs such as Cloudflare and the large hosting infrastructure
providers offer services to integrate all this fatness to web and app
developers and publishers. One stop obesity as a service if you like.
This type of solution is a signal of fragmentation that is at least as
protracted as complexities of whether we prioritise a single DNS root or
not.
Indeed for CDNs which run their own DNS infrastructure which allows them
to off services running through their CDNs their own root data and
domains should they so choose.
C
Christian de Larrinaga
Reply all
Reply to author
Forward
0 new messages