Oops, I got sidetracked a little bit in the last post, I meant to finish by posting an interesting link which is relevant to the story once you find it. Check out this image here:
All of the Linux variants ran OK on 2001-grade hardware (P3-800 + 128MB RAM); most of them ran fine on 1999-grade hardware (P2-450/64MB). About half ran OK on 1997-grade P-233/32MB, and only one ran on a 1995-era box, which is over a decade old.
They didn't show how well Win2000 or WinXP or Vista will do on such machines, oddly enough, although the Ziff-Davis article spent a lot of time talking about the requirements for Office versus OpenOffice instead. I wonder why?
It's also a pity they didn't test the BSD family as well; NetBSD in particular would excel for that task, although OpenBSD and FreeBSD up through v4 would probably also work all the way back to the 1995-era 486/16MB machine. All three ought to work fine on the 1997 box.
Agreed, I'm not sure whether the editors are deliberately letting such things slip through, but it's getting to the point where simply finding TFA to read it requires guessing right. Maybe we need to require people submitting articles to disclose whether they are being compensated directly or indirectly via ad revenue.
At least in some places of the world, that's required by law:
When there exists a connection between the endorser and the seller of the advertised product which might materially affect the weight or credibility of the endorsement (i.e., the connection is not reasonably expected by the audience) such connection must be fully disclosed."
In case the relevance isn't clear, when you recommend that people read an article on a site that you own and are making money from (putting up ad banners where people can buy things obviously qualifies), you need to disclose that fact and specificly whether the subject of the article is paying anything for your referral.
Would any Newsforge editor care to speak to this matter?
Agreed, but it's not just Sony and Microsoft that are persuing odious tactics with regard to closed standards, pretty much every part of the DVD-related industry is playing these sort of games: the "no, you don't own that fancy new DVD you just purchased, you only have a license to use it at home with a HDCP-encumbered Blu-Ray or HD-DVD player connected to an HDCP-encumbered TV".
Fact is, the companies involved here are writing products to suit themselves rather than the customers. I don't mind going with a proprietary solution that suits my needs, but why should I purchase a proprietary solution that doesn't?
Therefore, I don't plan on purchasing either format anytime soon, nor do I plan to get either a Sony PS3 or an Xbox, as a result of this infighting.
Is your job to manage a team of sysadmins, or just manage a bunch of desktop machines which happen to be used by sysadmins? The latter is generally a lot easier, although some admins can get as picky about someone managing their personal box as any huffy user.
If you're just managing the machines, make sure you've got a software baseline (start with, what software do we have and how many legit licenses have we paid for), and make sure that all of the machines which are supposed to have that stuff do, and it is patched and up-to-date. Keep a mental checklist of any machines which have severe problems; that may not be the sysadmin user's fault, but it still ought to be a warning sign.
Sysadmins always seem to get buried in the never-ending stream of building new machines, or running Windows update, or virus-scanning, etc, but try to keep the team aware of longer-term goals beyond the humdrum of such routine daily tasks. Try to give everyone at least two long-term goals or tasks to complete: one that's fun or interesting or cool, and one that sucks or is boring but needs to be done. Make sure everyone knows that everyone has got to deal with some suckatude, publicly praise/reward the first guy who finishes a sucky task. On the other hand, if someone gives you problems or blows off the difficult task, make another public point of awarding another sucky task to the guy.
Do you remember Shalebridge Cradle? Beware! The Cradle *doesn't* forget....
That level alone was worth the price of Thief 3, although the Kurshok (sp?) underworld and the haunting melody of the seaside mansion were also excellent levels, too.
I think you'd end up considering bottled water and tap water seperately if you couldn't interchange them safely. I think "bottled water" itself would still be considered a commodity, though, and not a differentiated product:
Let's say you had a 1L bottle of Poland Springs, and a 1L bottle of Evian to choose from. Which one would you get?
It's pretty likely that the specific brand of bottled water you get doesn't matter much to you, or at least I doubt too many people would pay twice as much for one brand as for the other. Most people would buy whichever is cheapest, right?
If you can't choose between different network providers, that's unfortunate (perhaps WiMax or satellite might be doable?), but is not connected to what the word "commodity" means.
If you had several providers to choose from, getting a 1Mbs connection via cable is pretty much the same as getting a 1Mbs connection from DSL. Network bandwidth isn't quite as homogenous as water is, but then, some people are willing to pay a lot for bottled water in a store, too, so it seems the useage doesn't have to be exact.
You're talking about something which already exists in practice: the ISP's, including the big boys like Verizon or Time-Warner Cable, prioritize repair services and make extra efforts to provide redundancies in the infrastructure connected to emergency services, things like hospitals and fire stations.
When 911 hit, our net connection survived (we were terminated at the Verizon building just north of the WTC site), but was taken over within a day in order for some clinic to regain network access. No arguments from us, it's useful to understand priorities in a real emergency, but we suffered along for about two months with a combination of dialup, spotty wireless, and stringing 100m lengths of ethernet around the building to borrow connectivity.
Secondly, the IP protocol itself has mechanisms for type-of-service and quality, see:
/* * Definitions for IP type of service (ip_tos) */ #define IPTOS_LOWDELAY 0x10 #define IPTOS_THROUGHPUT 0x08 #define IPTOS_RELIABILITY 0x04 #define IPTOS_MINCOST 0x02 /* ECN bits proposed by Sally Floyd */ #define IPTOS_CE 0x01/* congestion experienced */ #define IPTOS_ECT 0x02/* ECN-capable transport */
These can be used and some quality-of-service based routers will heed them, but in practice most routers do just as well by trying to process all traffic quickly rather than storing some packets around to send later at a lower priority. Still, noticing IPTOS_LOWDELAY and considering those packets first is a nice thing for the TCP/IP stack, firewall, or bandwidth shaper to do...
True. I don't see any problem with Internet providers offerring tiered services so long as you're talking about bandwidth as a commodity, which is specificly a product like salt or water which is "homogenous" and freely interchangable.
However, as soon as you start talking about charging more or less based on which web sites you go to, or which emails you get (and from whom), Internet service *isn't* being treated as a commodity where all connections are essentially just another stream of bits passing by the routers.
If the carriers apply editorial control over content, agreed, but I am not sure that automated quality-of-service mechanisms which are not under human volition would run afoul of the common carrier requirements. It's an interesting question to consider.
Right now, the carriers are offering tiered delivery models-- different pricing for different bandwidth levels for DSL and cable such as 256/128, 512/128, 1.5/256, etc. Nobody but phone-based dialups try to do usage-based billing, that or proprietary services that have generally lost many of their users (Prodigy?, Compuserve?, AOL?)....
That was fast. I clicked on the article before there were 16 posts made, and I figure that at least as many people would *not* RTFA as would. Wait...reload...reload...maybe...got it!
The site seems to be limping but not quite dead yet, Jim.:-)
I would say that plenty of OSS projects invent communications protocols, and yes, many are documented in the Internet RFC's. Things like IP, UDP, TCP, SMTP, HTTP, IMAP, OpenPGP, LDAP, XDR, RPC, etc...
That depends on which Open Source project you consider, of course. I happen to think that Apache, MySQL, Samba have pretty good documentation for what they do, as does FreeBSD, and even Sun's documentation for OpenSolaris is extremely good.
David Brin wrote more about that notion in "Earth", as well, which is a great fiction novel. The notion of wearing a camera around to record the cameras recording you is disturbing, but perhaps fitting. If nothing else, catching a cop beating or shooting a compliant suspect during an arrest points out that sometimes the guys wearing the badge exceed the law:
Pay attention to these sorts of things, everybody, people need to make sure that nobody is above the law, starting with the people who make and enforce them. One standard for all is better than an infinite number of double-standards driven from self-interest....
Democratic politicans or advocacy groups (PACs)? Cindy whats-her-name?:-) I know, you probably meant that as a rhetorical question, but nevertheless, it does have an answer.
I'm not particularly interested in criticizing BoingBoing for advertising, but let's do a reality check. The top of the page is a banner ad stating "Contact FM to advertise here", the left column starts with a header stating "Sponsored by" and 3 links all going through:
Ditto for the right-side column, only that not has a bunch of ads going through clicker.cgi but is followed by another set of ads going through adbrite.com. And then the article text itself is wrapped around the ad image from craphound.com, and the end has ads by something called Kanoodle.
Take a look at "Page Info" for the links on that page and judge for yourself, but *I'm* willing call an article surrounded by advertising links on all sides what it is, even if you choose not to.
Seems like BoingBoing.com is trolling for hits with several recent articles. I suppose there's nothing wrong with that, but I'd wish the Slashdot editors to prefer primary content to secondary sources being framed within ad bars on all sides....
The newspapers would like you to become a registered, paying user who has to sit through a full-page interstitial ad or two before getting to see the "real content", whatever that may be. Case in point: Times Select of the NY Times.
As others have pointed out earlier, Google honors robots.txt and the META noindex and nofollow tags, so any news site which wanted to keep some or all of their content private from webspiders, Google's cache, news.google.com, or anything else can do so trivially. Frankly, it's up to them to take minimal responsibility before publishing content to the entire world if they want to impose restrictions.
I've found that disabling image rotation ("animated GIFs") and refusing to install Macromedia's Flash player makes the web a far more pleasant place, with a lot fewer ads. Using ad-blocking tools which work with a proxy server like squid can help block most of the rest of the ads...
GPS does indeed make a wonderful external time reference, and many stratum-1 NTP timeservers are using it.
Of course, most machines locked in a rack in a hosting facility don't have even the slightest chance of seeing enough sky to lock onto GPS, so it's safe to say that NTP's death or obsolesence is premature to announce just yet.:-)
--
-Chuck
PS: O Slashdot wizards, why does Slashdot's posting filter claim ntpq output is lame?
It's a conspiracy, I tell you, to force me to write more text!
Bah, that doesn't work, the lameness filter doesn't like a line filled with "=" signs at all, even if I use an <ecode> tag.
It wasn't Cogent's routers which were returning ICMP host unreachable messages; Level3 decided to blackhole the Cogent netblocks. As I just mentioned elsewere in this thread, there isn't a problem if the two depeered, and decided to route traffic between them via some third party. However, when the Level3 routers claim that anything at Cogent is unreachable, it's Level3's responsibility and fault to correct.
My company has been affected by this pretty signicantly, although our regional ISP (reynwood.com) has done a fair amount to set up alternate routes for us since Level3 started blackholing PSI.net/Cogent netblocks. The problem is not that Level3 and Cogent depeered, but that Level3 routers were blackholing the Cogent netblocks by returning ICMP host unreachable messages. Once Level3 decided to turn off their direct connection with Cogent, then they should have chosen to route the traffic via one of their other peers as an intermediate, not claimed that all of Cogent's network was unreachable.
Here's a bunch of traceroutes to some of my Usenet peers showing exactly what was going on:
% traceroute news-in.newsgroups.com traceroute to news-in.newsgroups.com (38.119.100.108), 64 hops max, 44 byte packets
1 polycom1.codefab.com (199.103.21.254) 0.947 ms 0.787 ms 0.874 ms
2 199.103.21.9 (199.103.21.9) 3.011 ms 3.378 ms 3.225 ms
3 sw1.32a.nyc.reynwood.com (199.103.19.125) 3.560 ms 4.381 ms 4.361 ms
4 ge-8-1-241.core1.NewYork1.Level3.net (4.78.160.45) 3.960 ms !H * 4.476 ms !H
% traceroute news-in.spamkiller.net traceroute to news-in.spamkiller.net (38.119.71.4), 64 hops max, 44 byte packets
1 polycom1.codefab.com (199.103.21.254) 0.946 ms 0.771 ms 0.859 ms
2 199.103.21.9 (199.103.21.9) 3.081 ms 3.457 ms 3.312 ms
3 sw1.32a.nyc.reynwood.com (199.103.19.125) 3.896 ms 5.366 ms 4.283 ms
4 ge-8-1-241.core1.NewYork1.Level3.net (4.78.160.45) 3.986 ms !H * 4.455 ms !H
As an update, according to this status page here at http://status.cogentco.com/, Level3 has blinked and is restoring connectivity for now while both sides negotiate:
10/7/05 (4:20pm edt): Level 3 has restored all peers with Cogent as of 4:00pm edt. We are seeing some latency in traffic across to Level 3 as sessions re-establish, mail servers deliver messages, etc. We hope the above normal traffic volumes will decrease within the next hour.
10/7/05 (3:40pm edt): Level 3 has restored some of their peering sessions with Cogent at this time. We do not know if this is a temporary or a permanent change, and will continue to negotiate with Level 3 to resolve all the issues they have with Cogent.
The numbers you're quoting are fractional composition by *mass*, right? I was trying to quote relative composition, ie 18 kg of water has 16 kg of O and 2 kg of H, which is {O:.89, H:.11} by mass but only {O:.33, H:.67} by relative frequency of composition.
Slashdot Mirror
Oops, I got sidetracked a little bit in the last post, I meant to finish by posting an interesting link which is relevant to the story once you find it. Check out this image here:
e /12/0,1425,i=123899,00.jpg
http://common.ziffdavisinternet.com/util_get_imag
All of the Linux variants ran OK on 2001-grade hardware (P3-800 + 128MB RAM); most of them ran fine on 1999-grade hardware (P2-450/64MB). About half ran OK on 1997-grade P-233/32MB, and only one ran on a 1995-era box, which is over a decade old.
They didn't show how well Win2000 or WinXP or Vista will do on such machines, oddly enough, although the Ziff-Davis article spent a lot of time talking about the requirements for Office versus OpenOffice instead. I wonder why?
It's also a pity they didn't test the BSD family as well; NetBSD in particular would excel for that task, although OpenBSD and FreeBSD up through v4 would probably also work all the way back to the 1995-era 486/16MB machine. All three ought to work fine on the 1997 box.
Agreed, I'm not sure whether the editors are deliberately letting such things slip through, but it's getting to the point where simply finding TFA to read it requires guessing right. Maybe we need to require people submitting articles to disclose whether they are being compensated directly or indirectly via ad revenue.
o ad.htm
At least in some places of the world, that's required by law:
http://www.ftc.gov/bcp/conline/pubs/buspubs/ruler
http://www.ftc.gov/bcp/guides/endorse.htm
"255.5 Disclosure of material connections.
When there exists a connection between the endorser and the seller of the advertised product which might materially affect the weight or credibility of the endorsement (i.e., the connection is not reasonably expected by the audience) such connection must be fully disclosed."
In case the relevance isn't clear, when you recommend that people read an article on a site that you own and are making money from (putting up ad banners where people can buy things obviously qualifies), you need to disclose that fact and specificly whether the subject of the article is paying anything for your referral.
Would any Newsforge editor care to speak to this matter?
Agreed, but it's not just Sony and Microsoft that are persuing odious tactics with regard to closed standards, pretty much every part of the DVD-related industry is playing these sort of games: the "no, you don't own that fancy new DVD you just purchased, you only have a license to use it at home with a HDCP-encumbered Blu-Ray or HD-DVD player connected to an HDCP-encumbered TV".
Fact is, the companies involved here are writing products to suit themselves rather than the customers. I don't mind going with a proprietary solution that suits my needs, but why should I purchase a proprietary solution that doesn't?
Therefore, I don't plan on purchasing either format anytime soon, nor do I plan to get either a Sony PS3 or an Xbox, as a result of this infighting.
Is your job to manage a team of sysadmins, or just manage a bunch of desktop machines which happen to be used by sysadmins? The latter is generally a lot easier, although some admins can get as picky about someone managing their personal box as any huffy user.
If you're just managing the machines, make sure you've got a software baseline (start with, what software do we have and how many legit licenses have we paid for), and make sure that all of the machines which are supposed to have that stuff do, and it is patched and up-to-date. Keep a mental checklist of any machines which have severe problems; that may not be the sysadmin user's fault, but it still ought to be a warning sign.
Sysadmins always seem to get buried in the never-ending stream of building new machines, or running Windows update, or virus-scanning, etc, but try to keep the team aware of longer-term goals beyond the humdrum of such routine daily tasks. Try to give everyone at least two long-term goals or tasks to complete: one that's fun or interesting or cool, and one that sucks or is boring but needs to be done. Make sure everyone knows that everyone has got to deal with some suckatude, publicly praise/reward the first guy who finishes a sucky task. On the other hand, if someone gives you problems or blows off the difficult task, make another public point of awarding another sucky task to the guy.
Sounds more like the auth panel which MacOS X uses than like sudo.
While this this sounds like a fine idea, people have been doing the same thing with modems, for, oh, 30-plus years. And then there's "ping -a"...
Sure, although trying to describe the level risks spoiling it a bit. But read the end of this link here, which is a review of the game:
http://arstechnica.com/reviews/games/tds.ars/4
Do you remember Shalebridge Cradle? Beware! The Cradle *doesn't* forget....
That level alone was worth the price of Thief 3, although the Kurshok (sp?) underworld and the haunting melody of the seaside mansion were also excellent levels, too.
I think you'd end up considering bottled water and tap water seperately if you couldn't interchange them safely. I think "bottled water" itself would still be considered a commodity, though, and not a differentiated product:
Let's say you had a 1L bottle of Poland Springs, and a 1L bottle of Evian to choose from.
Which one would you get?
It's pretty likely that the specific brand of bottled water you get doesn't matter much to you, or at least I doubt too many people would pay twice as much for one brand as for the other. Most people would buy whichever is cheapest, right?
If you can't choose between different network providers, that's unfortunate (perhaps WiMax or satellite might be doable?), but is not connected to what the word "commodity" means.
If you had several providers to choose from, getting a 1Mbs connection via cable is pretty much the same as getting a 1Mbs connection from DSL. Network bandwidth isn't quite as homogenous as water is, but then, some people are willing to pay a lot for bottled water in a store, too, so it seems the useage doesn't have to be exact.
When 911 hit, our net connection survived (we were terminated at the Verizon building just north of the WTC site), but was taken over within a day in order for some clinic to regain network access. No arguments from us, it's useful to understand priorities in a real emergency, but we suffered along for about two months with a combination of dialup, spotty wireless, and stringing 100m lengths of ethernet around the building to borrow connectivity.
Secondly, the IP protocol itself has mechanisms for type-of-service and quality, see :
These can be used and some quality-of-service based routers will heed them, but in practice most routers do just as well by trying to process all traffic quickly rather than storing some packets around to send later at a lower priority. Still, noticing IPTOS_LOWDELAY and considering those packets first is a nice thing for the TCP/IP stack, firewall, or bandwidth shaper to do...True. I don't see any problem with Internet providers offerring tiered services so long as you're talking about bandwidth as a commodity, which is specificly a product like salt or water which is "homogenous" and freely interchangable.
However, as soon as you start talking about charging more or less based on which web sites you go to, or which emails you get (and from whom), Internet service *isn't* being treated as a commodity where all connections are essentially just another stream of bits passing by the routers.
That's probably because you have platinum and I only have silver !
I suspect there's a common metal called "irony" which is involved here...
If the carriers apply editorial control over content, agreed, but I am not sure that automated quality-of-service mechanisms which are not under human volition would run afoul of the common carrier requirements. It's an interesting question to consider.
Right now, the carriers are offering tiered delivery models-- different pricing for different bandwidth levels for DSL and cable such as 256/128, 512/128, 1.5/256, etc. Nobody but phone-based dialups try to do usage-based billing, that or proprietary services that have generally lost many of their users (Prodigy?, Compuserve?, AOL?)....
That was fast. I clicked on the article before there were 16 posts made, and I figure that at least as many people would *not* RTFA as would. Wait...reload...reload...maybe...got it!
:-)
The site seems to be limping but not quite dead yet, Jim.
I would say that plenty of OSS projects invent communications protocols, and yes, many are documented in the Internet RFC's. Things like IP, UDP, TCP, SMTP, HTTP, IMAP, OpenPGP, LDAP, XDR, RPC, etc...
That depends on which Open Source project you consider, of course. I happen to think that Apache, MySQL, Samba have pretty good documentation for what they do, as does FreeBSD, and even Sun's documentation for OpenSolaris is extremely good.
David Brin wrote more about that notion in "Earth", as well, which is a great fiction novel. The notion of wearing a camera around to record the cameras recording you is disturbing, but perhaps fitting. If nothing else, catching a cop beating or shooting a compliant suspect during an arrest points out that sometimes the guys wearing the badge exceed the law:
1 feb01,0,7570035.storyp /index.html
http://www.latimes.com/news/local/la-me-highspeed
http://www.cnn.com/2006/LAW/01/31/chase.beating.a
Pay attention to these sorts of things, everybody, people need to make sure that nobody is above the law, starting with the people who make and enforce them. One standard for all is better than an infinite number of double-standards driven from self-interest....
Democratic politicans or advocacy groups (PACs)? Cindy whats-her-name? :-)
I know, you probably meant that as a rhetorical question, but nevertheless, it does have an answer.
I'm not particularly interested in criticizing BoingBoing for advertising, but let's do a reality check. The top of the page is a banner ad stating "Contact FM to advertise here", the left column starts with a header stating "Sponsored by" and 3 links all going through:
http://boingboing.net/cgi-bin/clicker.cgi?somethin g
Ditto for the right-side column, only that not has a bunch of ads going through clicker.cgi but is followed by another set of ads going through adbrite.com. And then the article text itself is wrapped around the ad image from craphound.com, and the end has ads by something called Kanoodle.
Take a look at "Page Info" for the links on that page and judge for yourself, but *I'm* willing call an article surrounded by advertising links on all sides what it is, even if you choose not to.
How about directly linking to the article, rather than bouncing through a portal full of ads?
http://www.eff.org/legal/cases/att/
Seems like BoingBoing.com is trolling for hits with several recent articles. I suppose there's nothing wrong with that, but I'd wish the Slashdot editors to prefer primary content to secondary sources being framed within ad bars on all sides....
The newspapers would like you to become a registered, paying user who has to sit through a full-page interstitial ad or two before getting to see the "real content", whatever that may be. Case in point: Times Select of the NY Times.
As others have pointed out earlier, Google honors robots.txt and the META noindex and nofollow tags, so any news site which wanted to keep some or all of their content private from webspiders, Google's cache, news.google.com, or anything else can do so trivially. Frankly, it's up to them to take minimal responsibility before publishing content to the entire world if they want to impose restrictions.
I've found that disabling image rotation ("animated GIFs") and refusing to install Macromedia's Flash player makes the web a far more pleasant place, with a lot fewer ads. Using ad-blocking tools which work with a proxy server like squid can help block most of the rest of the ads...
Of course, most machines locked in a rack in a hosting facility don't have even the slightest chance of seeing enough sky to lock onto GPS, so it's safe to say that NTP's death or obsolesence is premature to announce just yet. :-)
--
-Chuck
PS: O Slashdot wizards, why does Slashdot's posting filter claim ntpq output is lame?
It's a conspiracy, I tell you, to force me to write more text!
Bah, that doesn't work, the lameness filter doesn't like a line filled with "=" signs at all, even if I use an <ecode> tag.
It wasn't Cogent's routers which were returning ICMP host unreachable messages; Level3 decided to blackhole the Cogent netblocks. As I just mentioned elsewere in this thread, there isn't a problem if the two depeered, and decided to route traffic between them via some third party. However, when the Level3 routers claim that anything at Cogent is unreachable, it's Level3's responsibility and fault to correct.
My company has been affected by this pretty signicantly, although our regional ISP (reynwood.com) has done a fair amount to set up alternate routes for us since Level3 started blackholing PSI.net/Cogent netblocks. The problem is not that Level3 and Cogent depeered, but that Level3 routers were blackholing the Cogent netblocks by returning ICMP host unreachable messages. Once Level3 decided to turn off their direct connection with Cogent, then they should have chosen to route the traffic via one of their other peers as an intermediate, not claimed that all of Cogent's network was unreachable.
Here's a bunch of traceroutes to some of my Usenet peers showing exactly what was going on:
% traceroute news-in.newsgroups.com
traceroute to news-in.newsgroups.com (38.119.100.108), 64 hops max, 44 byte packets
1 polycom1.codefab.com (199.103.21.254) 0.947 ms 0.787 ms 0.874 ms
2 199.103.21.9 (199.103.21.9) 3.011 ms 3.378 ms 3.225 ms
3 sw1.32a.nyc.reynwood.com (199.103.19.125) 3.560 ms 4.381 ms 4.361 ms
4 ge-8-1-241.core1.NewYork1.Level3.net (4.78.160.45) 3.960 ms !H * 4.476 ms
!H
% traceroute news-in.spamkiller.net
traceroute to news-in.spamkiller.net (38.119.71.4), 64 hops max, 44 byte packets
1 polycom1.codefab.com (199.103.21.254) 0.946 ms 0.771 ms 0.859 ms
2 199.103.21.9 (199.103.21.9) 3.081 ms 3.457 ms 3.312 ms
3 sw1.32a.nyc.reynwood.com (199.103.19.125) 3.896 ms 5.366 ms 4.283 ms
4 ge-8-1-241.core1.NewYork1.Level3.net (4.78.160.45) 3.986 ms !H * 4.455 ms
!H
As an update, according to this status page here at http://status.cogentco.com/, Level3 has blinked and is restoring connectivity for now while both sides negotiate:
10/7/05 (4:20pm edt): Level 3 has restored all peers with Cogent as of 4:00pm edt. We are seeing some latency in traffic across to Level 3 as sessions re-establish, mail servers deliver messages, etc. We hope the above normal traffic volumes will decrease within the next hour.
10/7/05 (3:40pm edt): Level 3 has restored some of their peering sessions with Cogent at this time. We do not know if this is a temporary or a permanent change, and will continue to negotiate with Level 3 to resolve all the issues they have with Cogent.
The numbers you're quoting are fractional composition by *mass*, right? I was trying to quote relative composition, ie 18 kg of water has 16 kg of O and 2 kg of H, which is {O: .89, H: .11} by mass but only {O: .33, H: .67} by relative frequency of composition.