NetworkWorld published an article citing research from Probe Research ranking the top 10 ISPs and their marketshare based on Year-end (2000) revenue. 1. Worldcom (27.9%) 2. AT&T (10%) 3. Sprint (6.5%) 4. Genuity (6.3%) 5. PSINet (4.1%) 6. Cable & Wireless (3.5%) 7. XO Communications (2.8%) 8. Verio (2.6%) 9. Qwest (1.5%) 10. Global Crossing (1.3%) Since PSI is larger than C&W, that C&W should pay for transit :-) I don't know what happened to Level 3, ELI, @HOME, Broadwing, or all the non-US ISPs. http://www.nwfusion.com/newsletters/isp/2001/00846039.html
At 09:47 12/06/01, Sean Donelan wrote:
NetworkWorld published an article citing research from Probe Research ranking the top 10 ISPs and their marketshare based on Year-end (2000) revenue.
Essentially none of these sorts of analyses are meaningful or relevant because one can yield any result or ordering one wants by playing games with the definition of "ISP" and the basis for the ranking. Also very unclear to me how any of this helps folks operate their networks or is applicable to NANOG, unless we've become a marketing organisation recently... Ran
On Tue, 12 Jun 2001, RJ Atkinson wrote:
At 09:47 12/06/01, Sean Donelan wrote:
NetworkWorld published an article citing research from Probe Research ranking the top 10 ISPs and their marketshare based on Year-end (2000) revenue.
Essentially none of these sorts of analyses are meaningful or relevant because one can yield any result or ordering one wants by playing games with the definition of "ISP" and the basis for the ranking.
Well - I think Genuity is the #1 ISP - since they use AS 1 (They got AS1 when they bought(merged in ?) BBN ) and it's quite obvious that the ISP with the #1 Autonomous System is the best - right ? ( ;-) Rafi
Also very unclear to me how any of this helps folks operate their networks or is applicable to NANOG, unless we've become a marketing organisation recently...
Ran
On Tue, 12 Jun 2001 19:05:02 +0300, Rafi Sadowsky said:
Well - I think Genuity is the #1 ISP - since they use AS 1 (They got AS1 when they bought(merged in ?) BBN ) and it's quite obvious that the ISP with the #1 Autonomous System is the best - right ? ( ;-)
Everybody knows that Real Programmers start counting from zero. -- Valdis Kletnieks Operating Systems Analyst Virginia Tech
In message <Pine.GSO.4.31.0106121858460.13144-100000@meron.openu.ac.il>, Rafi S adowsky writes:
Well - I think Genuity is the #1 ISP - since they use AS 1 (They got AS1 when they bought(merged in ?) BBN )
Just for historical accuracy. Genuity used to be a BBN-owned ISP named BBN Planet (thus AS 1). BBN was acquired by GTE. When GTE and Bell Atlantic merged to create Verizon, the ISP (but not the rest of BBN) was spun off as Genuity. Craig Partridge Chief Scientist BBN Technologies (a Verizon company)
All: Does anyone else think this Intellectual Property Claim Service stuff for .BIZ is a scam? We started getting flooded with questions from our customers, who in turn have been flooded with notices from their lawyers on this. A quick read of all the stuff on the ICANN and Nuelevel sites makes me sick. The only benifit I see from this "IP Claim Service" is the registrars getting rich. http://www.neulevel.com/countdown/step1.html http://www.icann.org/announcements/icann-pr15may01.htm Chuck
Chuck, This is a "nanmg" topic, neh? It is a bit late, but as we mostly learn from that frightful intersection of theory and practice, do you have an alternative to propose than the IP Claim process we came up with? That's "NeuLevel", not Nuelevel, and some of the boundary conditions are inflexibie: existance of ICANN, existance of marks, existance of speculators, existance of money, and of course, finite ICANN and operator time to craft a process moderately resiliant to some or all of the above. Please don't sell our product if you think it is a scam. That wouldn't be ethical conduct. Eric Brunner-Williams NeuStar, Inc. Senior Technical Industry Liaison http://www.neustar.com Phone: (Portland, ME) +1.207.xxx.xxxx (Washington, DC) +1.202.533.2600/2975 smail: 1415 Forest Ave. 04103 1120 Vermont Ave. N.W., Ste 400, 20005 email: brunner@nic-naa.net ebw@neustar.com
Eric: Sorry, you are of course correct that this is probably not a nanog appropriate issue. Also, yes, I agree that selling this service to our customers would not be ethical. Chuck
Chuck,
This is a "nanmg" topic, neh?
It is a bit late, but as we mostly learn from that frightful intersection of theory and practice, do you have an alternative to propose than the IP Claim process we came up with?
That's "NeuLevel", not Nuelevel, and some of the boundary conditions are inflexibie: existance of ICANN, existance of marks, existance of speculators, existance of money, and of course, finite ICANN and operator time to craft a process moderately resiliant to some or all of the above.
Please don't sell our product if you think it is a scam. That wouldn't be ethical conduct.
Eric Brunner-Williams NeuStar, Inc. Senior Technical Industry Liaison http://www.neustar.com Phone: (Portland, ME) +1.207.xxx.xxxx (Washington, DC) +1.202.533.2600/2975 smail: 1415 Forest Ave. 04103 1120 Vermont Ave. N.W., Ste 400, 20005 email: brunner@nic-naa.net ebw@neustar.com
--
Chuck, Were you planning on doing anything along the lines of a substantive alternative proposal, or just pissing generally at .biz and/or ICANN and/or marks and/or money and/or gravity? Your "agreement" is with-self, which is about as useful as self-peering. Make it useful, or at least interesting. Eric
Eric: Frankly I see no possible resolution to the problems with new TLD's. There is no reasonable approach short of simply opening those domains for registrations and letting it all shake out in the courts (as do eventually all serious trademark disputes). I have long been, and to many of my business associates and perhaps my customers dissmay, an advocate of dumping everything other than the regional domain structure. Given that there is way too much momentum in .com, .net and .org, it's unlikely that would ever happen. What I can't see is a process that requires a littany of costly IP Claims that are only of value if you happen to have a number of special cases all occur. One, that you correctly predict all creative constructions of domain names that may conflict with your mark. Two, that all parties happen to be friendly to each other and agree to the pannel's decision. And three, that all parties really understand trademark law and know what rights they have and don't have. I really think all three are unlikely and, short of that, expect any final resolutions to be decided, as usual with Trademark cases, by who has the best lawyers and the highest tolerance for Federal court. Therefore I don't see any value in the IP claims process, only cost, delay and grief. So, if you want my proposal, I reccomend that the IP Claims process be dispensed with and we either get on with the chaos of new TLD's, or we forget it and wait for people to give up on dwindling TLD resources and move on the the regional domain structure. I hope that was interesting enough. If you think I missunderstand the situation, please explain how. I will however respect this list (as it has been pointed out that this is somewhat unrelated to network operations) and not post again on this topic unless there's a desire for me to do so. Chuck On Tue, 12 Jun 2001, Eric Brunner-Williams in Portland Maine wrote:
Chuck,
Were you planning on doing anything along the lines of a substantive alternative proposal, or just pissing generally at .biz and/or ICANN and/or marks and/or money and/or gravity?
Your "agreement" is with-self, which is about as useful as self-peering.
Make it useful, or at least interesting.
Eric
--
Chuck, I'll stick to the factual errors.
One, that you correctly predict all creative constructions of domain names that may conflict with your mark.
The IP Claim service we deployed is for exact match, strcmp() returning 0. Your items two and three are speculative, and anyone can speculate, which may be the root problem. Your conclusion (based upon one factual error and two speculative claims) is contradicted by the experience with the URDP, and as the study was done by academics (and fairly interesting, covering the major modes of DRP and the outcome distributions) you may want to fix their methodology, data and conclusions [1].
So, if you want my proposal ...
Only 2 boundary conditions removed: existance of ICANN, existance of marks. Neat. I'd have gone for gravity myself, it is such a bother.
I hope that was interesting enough.
Fairly lame actually, on par with Jim Fleming's v8 cure for what ails the net as a reality-based proposal, and dull-as-ditchwater/common-as-crud as netzine sceanery. Do your "business associates and perhaps [your] customers" give a fig about your irrepressible vision and truth of DNS reform? Why? Are they bored? Feel free to have the last word, its your scam. Follow-ups to the NANFG list. Eric References: [1] Preliminary Report from Max Planck Institute on UDRP study, ICANN DNSO Intellectual Property Constituency Meeting, Stockholm, 1 June 2001.
why aren't you signing your nasty condescending note with your NeuLevel employee status Eric? Eric Brunner-Williams NeuStar, Inc. Senior Technical Industry Liaison http://www.neustar.com Phone: (Portland, ME) +1.207.xxx.xxxx (Washington, DC) +1.202.533.2600/2975 smail: 1415 Forest Ave. 04103 1120 Vermont Ave. N.W., Ste 400, 20005 email: brunner@nic-naa.net ebw@neustar.com from another list yesterday At 04:40 PM 6/12/01 -0400, you wrote:
One and the same.
At 03:59 PM 6/12/2001 -0400, you wrote:
Is this the "native rights" Eric Brunner?
Fasten your seats then this is gonna be a rough ride. I guess it's like what somebody told me about setting up a bar, the first thing you do is get a big bouncer. ======================= I'll leave it to others on Nanog to consider whether your ability to have a reasoned exchange of information fits the above description. consider yourself filtered Eric <plonk>
Chuck,
I'll stick to the factual errors.
One, that you correctly predict all creative constructions of domain names that may conflict with your mark.
The IP Claim service we deployed is for exact match, strcmp() returning 0.
Your items two and three are speculative, and anyone can speculate, which may be the root problem.
Your conclusion (based upon one factual error and two speculative claims) is contradicted by the experience with the URDP, and as the study was done by academics (and fairly interesting, covering the major modes of DRP and the outcome distributions) you may want to fix their methodology, data and conclusions [1].
So, if you want my proposal ...
Only 2 boundary conditions removed: existance of ICANN, existance of marks.
Neat. I'd have gone for gravity myself, it is such a bother.
I hope that was interesting enough.
Fairly lame actually, on par with Jim Fleming's v8 cure for what ails the net as a reality-based proposal, and dull-as-ditchwater/common-as-crud as netzine sceanery.
Do your "business associates and perhaps [your] customers" give a fig about your irrepressible vision and truth of DNS reform? Why? Are they bored?
Feel free to have the last word, its your scam. Follow-ups to the NANFG list.
Eric
References:
[1] Preliminary Report from Max Planck Institute on UDRP study, ICANN DNSO Intellectual Property Constituency Meeting, Stockholm, 1 June 2001.
-- **************************************************************** The COOK Report on Internet, 431 Greenway Ave, Ewing, NJ 08618 USA (609) 882-2572 (phone & fax) cook@cookreport.com Index to 9 years of the COOK Report at http://cookreport.com For info on new Ethernet in the First Mile and 10 Gig E issue go to http://cookreport.com/10.04.shtml Single issues may be purchased for $125 or $250 for single or group use ****************************************************************
On Wed, 13 Jun 2001, Eric Brunner-Williams in Portland Maine wrote:
The IP Claim service we deployed is for exact match, strcmp() returning 0.
Eric: Where is my factual error? Since strcmp() only returs 0 if the whole strings match exactly, it supports my contention that one needs to divine EXACTLY what every possible variation would be and would have to file each separately at a cost of $90 each. If, however, it were a substring match, that would be (somewhat) more reasonable.
Your conclusion (based upon one factual error and two speculative claims) is contradicted by the experience with the URDP, and as the study was done by academics (and fairly interesting, covering the major modes of DRP and the outcome distributions) you may want to fix their methodology, data and conclusions [1].
To invoke UDRP, one must begin by filing a "complaint in a court of proper jurisdiction against the domain-name holder" and then UDRP only deals with the dispensation of the domain once a legal resolution to the claim of infringement is achieved. I gather that all registrars will handle domains in such resolved cases with the .biz TLD as they have other active TLD's--there doesn't seem to be any need for an additional step. I don't see how this does anything but support my concerns.
Neat. I'd have gone for gravity myself, it is such a bother.
I have also been focused on the "gravity" of this situation.
Feel free to have the last word, its your scam. Follow-ups to the NANFG list.
Thanks, for the last word. Chuck
Chuck, It appears that the specific form of the UDRP for .biz, which incorporates what is called "STOP", isn't available to you. Please drop a note to legal, that's legal@neulevel.com or legal@neustar.com, and ask. Factual error #2. Distinct from the IP Claim Service is a substring matching and notice service, you can ask for that also. The larger the Hamming-Distence I have to search, the more I'm going to charge, so try to avoid "watch-on-letter-a" and its raft of equivalents. Factual error #3 (or #1 revisted). Specific knowledge beats general inference, usually. There are nuances in the ICANN contracts, what else are contracts, SLAs, and attornies for, neh? Of course, specifics do get in the way of moral outrage. Is it still a "sham" or one solution to a problem you personally consider intractible, and we have found interesting, challenging, and occasionally quite funny? Eric
Hi We have a small network (5 /24s) and we need to host our web applications internally because they access backend servers that absolutely cannot be collocated. I am thinking about getting 2 T1s or fractional DS3s from InterNAP from different P-NAPS (1 from Philly and 1 from NY) each circuit from a different CO The reason that Im thinking InterNAP is because we dont qualify for a /20 and we would not be able to efficiently multi-home. It seems that InterNAP is perfect in our situation because they buy transit from multiple providers and claim to not have any black holes in their network. I hear many great things about InterNAP and I hear the opposite as well. Any thoughts would GREATLY be appreciated Thanks! -Seth
/24's are sufficient to multihome with most if not all providers out there. Why not conventionally multi-home to 2 large well established providers? -Dave "Seth M. Kusiak" wrote:
Hi
We have a small network (5 /24s) and we need to host our web applications internally because they access backend servers that absolutely cannot be collocated. I am thinking about getting 2 T1s or fractional DS3s from InterNAP from different P-NAPS (1 from Philly and 1 from NY) each circuit from a different CO
The reason that Im thinking InterNAP is because we dont qualify for a /20 and we would not be able to efficiently multi-home. It seems that InterNAP is perfect in our situation because they buy transit from multiple providers and claim to not have any black holes in their network.
I hear many great things about InterNAP and I hear the opposite as well.
Any thoughts would GREATLY be appreciated
Thanks!
-Seth
/24's are sufficient to multihome with most if not all providers out there. Why not conventionally multi-home to 2 large well established providers?
-Dave
People sometimes forget that you can multihome effectively with just a single /24 provided: 1) You pick two (or more) providers that meet with each other directly in several places (or at least are well-connected to whoever is announcing the larger block the /24 comes from). 2) Both (or all) providers agree to accept/announce your /24. 3) Your /24 is inside a larger routable block advertised by a provider that is unlikely to ever withdraw the route entirely. Your two providers are extremely unlikely to lose contact with each other. This provides all the benefits of multihoming with your own block with only four disadvantages: 1) You aren't quite as well protected against certain complex multiple failure scenarios. (For example, if the provider who owns the block your IPs come from withdraws that route, or if you lose your link to that provider at the same time it loses all its links to the provider you still have connectivity to.) 2) You can't as easily add or delete providers (they have to, at least, agree to announce your /24 and you need to make sure they have good connectivity to whoever is announcing the larger block the /24 comes from) and you are to some extent controlled by the ISP who assigned you the addresses you are using (you can't stop getting service from them without renumbering). 3) You may experience finger pointing in the even that you are having connectivity problems. Because of the unusualness of the setup, it tends to be harder to figure out who is really at fault than when you multihome with your own block. 4) Your setup is a bit more fragile with your providers. You run the risk that a configuration change will somehow break your setup, and it's more likely with this type of setup to wind up blackholing traffic then with a convnetional setup where it's more likely to just not carry traffic. Even providers that filter your /24 will still have the route to the larger block it is part of. So they will still get your traffic closer to you, which is good enough to ensure full connectivity. DS
On Wed, 13 Jun 2001, Seth M. Kusiak wrote:
It seems that InterNAP is perfect in our situation because they buy transit from multiple providers and claim to not have any black holes in their network.
I hear many great things about InterNAP and I hear the opposite as well.
Any thoughts would GREATLY be appreciated
If indeed they adhere to a "no black holes" policy, they are to be commended. Who would have thought this will become a strong selling point? Access to the whole uncensored Internet, available only from a few select providers... http://www.peacefire.org/stealth/group-statement.5-17-2001.html --Mitch NetSide
Apparently AboveNet doesn't use the RBL in BGP form anymore. -ajb On Wed, Jun 13, 2001 at 02:35:08PM -0400, Mitch Halmu wrote: -> -> ->On Wed, 13 Jun 2001, Seth M. Kusiak wrote: -> ->> It seems that InterNAP ->> is perfect in our situation because they buy transit from multiple providers ->> and claim to not have any black holes in their network. ->> ->> I hear many great things about InterNAP and I hear the opposite as well. ->> ->> Any thoughts would GREATLY be appreciated -> ->If indeed they adhere to a "no black holes" policy, they are to be ->commended. Who would have thought this will become a strong selling ->point? Access to the whole uncensored Internet, available only from ->a few select providers... -> ->http://www.peacefire.org/stealth/group-statement.5-17-2001.html -> ->--Mitch ->NetSide ---end quoted text--- -- Andrew Barros <abarros@tjhsst.edu> PGP Key Fingerprint: D3B8 0800 C45A 143E 5CF0 E112 0A1B AB36 B655 1FB8
On Tue, 12 Jun 2001, Rafi Sadowsky wrote:
Well - I think Genuity is the #1 ISP - since they use AS 1 (They got AS1 when they bought(merged in ?) BBN ) and it's quite obvious that the ISP with the #1 Autonomous System is the best - right ? ( ;-)
The local Genuity folks gave me a very nice mouse pad, with a huge "AS1" in about 80-pt letters very visible from a distance. Now I have people stopping by to ask who "ASS ONE" is. I tell them it's Genuity, they're proud to be it, and in fact they paid a whole bunch of money to become it, so they want everyone to know. They advertise it to every Internet router they can, especially their competitors'. Just waiting for the ad campaign: "We Kick Genuity's AS" Pete.
participants (14)
-
Andrew Barros -
Charles Scott -
Craig Partridge -
David McGaugh -
David Schwartz -
Eric Brunner-Williams in Portland Maine -
Gordon Cook -
Mitch Halmu -
Pete Kruckenberg -
Rafi Sadowsky -
RJ Atkinson -
Sean Donelan -
Seth M. Kusiak -
Valdis.Kletnieks＠vt.edu