RE: The root nameservers will be replaced August 1st
On Wednesday, July 17, 1996 4:51 AM, Tim Salo[SMTP:salo@msc.edu] wrote: <snip> @ @ More interestingly, if someone wants to create an alternative set of @ root servers, there is no particularly good reason for them to be located @ at exchange points, (unless I am confused about what networks are all @ about...). True, servers at exchange points should exhibit greater @ availability, but that is probably not the largest challenge faced by @ alternative root servers. @ These are very good points. I think that the exchange points are just convienant co-location sites. Also, in the future, there may be other services on those boxes that have not been announced yet. It is useful to have the boxes in strategic locations in advance of additional changes. -- Jim Fleming UNETY Systems, Inc. Naperville, IL e-mail: JimFleming@unety.net
Convenient for some, a headache for others. Any responsible NAP operator would have to either approve the application running on his network or shutdown its unauthorized operation. You know, like disable its connection to the exchange point! steve On Wed, 17 Jul 1996, Jim Fleming wrote:
On Wednesday, July 17, 1996 4:51 AM, Tim Salo[SMTP:salo@msc.edu] wrote: <snip> @ @ More interestingly, if someone wants to create an alternative set of @ root servers, there is no particularly good reason for them to be located @ at exchange points, (unless I am confused about what networks are all @ about...). True, servers at exchange points should exhibit greater @ availability, but that is probably not the largest challenge faced by @ alternative root servers. @
These are very good points. I think that the exchange points are just convienant co-location sites. Also, in the future, there may be other services on those boxes that have not been announced yet. It is useful to have the boxes in strategic locations in advance of additional changes.
-- Jim Fleming UNETY Systems, Inc. Naperville, IL
e-mail: JimFleming@unety.net
Hmm..... NAP operators making judgement about content... Hmm... I wonder if there are any applicable precedents here... -alan ......... Steven Schnell is rumored to have said: ] ] Convenient for some, a headache for others. Any responsible NAP operator ] would have to either approve the application running on his network or ] shutdown its unauthorized operation. You know, like disable its ] connection to the exchange point! ] ] steve ] ] ] On Wed, 17 Jul 1996, Jim Fleming wrote: ] ] > On Wednesday, July 17, 1996 4:51 AM, Tim Salo[SMTP:salo@msc.edu] wrote: ] > <snip> ] > @ ] > @ More interestingly, if someone wants to create an alternative set of ] > @ root servers, there is no particularly good reason for them to be located ] > @ at exchange points, (unless I am confused about what networks are all ] > @ about...). True, servers at exchange points should exhibit greater ] > @ availability, but that is probably not the largest challenge faced by ] > @ alternative root servers. ] > @ ] > ] > These are very good points. I think that the exchange points ] > are just convienant co-location sites. Also, in the future, there ] > may be other services on those boxes that have not been ] > announced yet. It is useful to have the boxes in strategic ] > locations in advance of additional changes. ] > ] > -- ] > Jim Fleming ] > UNETY Systems, Inc. ] > Naperville, IL ] > ] > e-mail: JimFleming@unety.net ] > ] > ] ]
Hmm.....
NAP operators making judgement about content... Hmm... I wonder if there are any applicable precedents here...
-alan
Sure... "It's ours and we set the rules." As far as I know, rules aren't even available for the MAEs from MFS, and the Pennsauken agreement speaks only in generalities about what one's not supposed to do routing-wise... Avi
The two perspectives I didn't bring up (hoping someone else would/will) are: o How does the NSF support/activity impact content and connections to the NAPs (and to a far lesser extent the MAEs, and the random XPs popping up everywhere [which I feel are a geat idea {but I won't keep rambling}]). o In the freedom-of-speech world, it is my understanding that there are significant precedents set that say once a body exhibits ANY editorial control, they are then responsible for all content. [of course, not exhibiting control doesn't insure that one isn't responsible]. Of course, in a purely objective, idealistic world, I'd agree and support Avi's assertions. -alan ......... Avi Freedman is rumored to have said: ] ] > Hmm..... ] > ] > NAP operators making judgement about content... Hmm... I wonder ] > if there are any applicable precedents here... ] > ] > -alan ] ] Sure... "It's ours and we set the rules." ] As far as I know, rules aren't even available for the MAEs from MFS, ] and the Pennsauken agreement speaks only in generalities about what ] one's not supposed to do routing-wise... ] ] Avi ] ] ]
You'll kindly notice that I mentioned application, not content, in my previous comments. As a Layer 2 network provider I have no interest in content (as I cannot survey what traffic goes on past above me). However, I am keenly interested in the collocation of certain host-based applications that may cause unnecessary traffic to be directed to the NAP. There are a host (no pun intended) of other issues related to supporting hosts at the NAP, such as the excessive space and power they require, coordinating visits for maintenance and repair, security (physical and network), etc. I suspect that not all NAP operators are set up to take on the responsibilities of supporting these types of applications on their sites. Steve On Thu, 18 Jul 1996, Alan Hannan wrote:
Hmm.....
NAP operators making judgement about content... Hmm... I wonder if there are any applicable precedents here...
-alan
......... Steven Schnell is rumored to have said: ] ] Convenient for some, a headache for others. Any responsible NAP operator ] would have to either approve the application running on his network or ] shutdown its unauthorized operation. You know, like disable its ] connection to the exchange point! ] ] steve ] ] ] On Wed, 17 Jul 1996, Jim Fleming wrote: ] ] > On Wednesday, July 17, 1996 4:51 AM, Tim Salo[SMTP:salo@msc.edu] wrote: ] > <snip> ] > @ ] > @ More interestingly, if someone wants to create an alternative set of ] > @ root servers, there is no particularly good reason for them to be located ] > @ at exchange points, (unless I am confused about what networks are all ] > @ about...). True, servers at exchange points should exhibit greater ] > @ availability, but that is probably not the largest challenge faced by ] > @ alternative root servers. ] > @ ] > ] > These are very good points. I think that the exchange points ] > are just convienant co-location sites. Also, in the future, there ] > may be other services on those boxes that have not been ] > announced yet. It is useful to have the boxes in strategic ] > locations in advance of additional changes. ] > ] > -- ] > Jim Fleming ] > UNETY Systems, Inc. ] > Naperville, IL ] > ] > e-mail: JimFleming@unety.net ] > ] > ] ]
Steve, I appreciate your response. Thanks. ] You'll kindly notice that I mentioned application, not content, in my ] previous comments. As a Layer 2 network provider I have no interest in ] content (as I cannot survey what traffic goes on past above me). ] However, I am keenly interested in the collocation of certain host-based ] applications that may cause unnecessary traffic to be directed to the ] NAP. I would assert that given that your concern centers on applications, you are indeed looking higher up the stack than layer 2, and that you are making judgements on content, and purpose of use. ] There are a host (no pun intended) of other issues related to ] supporting hosts at the NAP, such as the excessive space and power they ] require, coordinating visits for maintenance and repair, security ] (physical and network), etc. I suspect that not all NAP operators are ] set up to take on the responsibilities of supporting these types of ] applications on their sites. I'm not sure that I accept the assertion that hosts take more care and feeding than routers... Perhaps it's true. Regardless, thanks for your insight. I feel this issue will someday soon become rather important. -alan ] On Thu, 18 Jul 1996, Alan Hannan wrote: ] ] > ] > Hmm..... ] > ] > NAP operators making judgement about content... Hmm... I wonder ] > if there are any applicable precedents here... ] > ] > -alan ] > ] > ......... Steven Schnell is rumored to have said: ] > ] ] > ] Convenient for some, a headache for others. Any responsible NAP operator ] > ] would have to either approve the application running on his network or ] > ] shutdown its unauthorized operation. You know, like disable its ] > ] connection to the exchange point! ] > ] ] > ] steve ] > ] ] > ] ] > ] On Wed, 17 Jul 1996, Jim Fleming wrote: ] > ] ] > ] > On Wednesday, July 17, 1996 4:51 AM, Tim Salo[SMTP:salo@msc.edu] wrote: ] > ] > <snip> ] > ] > @ ] > ] > @ More interestingly, if someone wants to create an alternative set of ] > ] > @ root servers, there is no particularly good reason for them to be located ] > ] > @ at exchange points, (unless I am confused about what networks are all ] > ] > @ about...). True, servers at exchange points should exhibit greater ] > ] > @ availability, but that is probably not the largest challenge faced by ] > ] > @ alternative root servers. ] > ] > @ ] > ] > ] > ] > These are very good points. I think that the exchange points ] > ] > are just convienant co-location sites. Also, in the future, there ] > ] > may be other services on those boxes that have not been ] > ] > announced yet. It is useful to have the boxes in strategic ] > ] > locations in advance of additional changes. ] > ] > ] > ] > -- ] > ] > Jim Fleming ] > ] > UNETY Systems, Inc. ] > ] > Naperville, IL ] > ] > ] > ] > e-mail: JimFleming@unety.net ] > ] > ] > ] > ] > ] ] > ] ] > ] > ] ]
participants (4)
-
Alan Hannan
-
Avi Freedman
-
Jim Fleming
-
Steven Schnell