Natch. IP address range for spammers is not enough. Insist that spammers turn on the Evil bit - RFC 3514.
(More seriously - RFC 3675, ".sex considered dangerous", gives a little more thinking about why this is a Bad Idea. This is NOT an Internet standard, but it's a well written document.)
Microsoft is pushing a solution called "Caller ID", which involves putting (wince) XML documents into the DNS telling you how to check the (argh) From: header. A lot of other people are pushing a solution called SPF, which involves putting text "code snippets" into the DNS telling you how to check the MAIL FROM: envelope return address. This topic will be discussed at the IETF next week in Seoul, Korea. Hot topic!
There's a slight problem.... in that until everyone signs their email, you'll have to be willing to handle unsigned email as well as signed. That leaves the signing people worse off than the non-signing people (more pain, no gain). Difficult deployment problem.
-----BEGIN PGP SIGNATURE----- Version: PGP 7.0.1
iQA/AwUBQCOn5jjI/tvlmNBeEQLIdwCfTzU3AFyy3vAyqJ1T re ICmreO16YAoJ3J Yl8AGPs6HHxEEGJfkmV857m1 =XHyf - ----END PGP SIGNATURE-----
Since real numbers are hard to come by, we get by on the very weak ones we have. According to the version monitoring page at the Linux Counter, 35 out of the 4862 monitored machines run the 2.0 kernel - 0.7% of the total number of monitored machines. If that holds true for the (who knows???) 20 million Linux machines out there in the wild, there should be something like 142000 2.0-kernel Linux boxes out there. Perhaps more - the "enthusiasts" who register with the Linux Counter may be more prone to upgrading than others....
the first IETF IM standard to make it through the process was the CPIM package (draft-ietf-impp-cpim-msgfmt). It's a specification on how to interconnect IM systems rather than a complete IM protocol specification. The other major player in IETF standards-space is SIMPLE - the presence specification documents for that (draft-ietf-simple-presence) are in the RFC Editor's queue. The nice thing about standards is that there are so many of them.....
this has actually happened - not with ezpass but with mobile phones. in a certain (rather spectacular) murder trial in Norway, one suspect's mobile phone was on an extended trip very far away from the murder site at the time, tracked by your ever-friendly telco's "cell tower association records". We do not know if the suspect went along.....
the linux counter was "slightly" misconfigured in how many apache servers it allowed to run simultaneously - it went into trashing. I've re-tuned it (to MaxClients = 32, and KeepAliveTimeout = 2), and it seems to work slightly better. You can still expect it to be SLOW, though..... (anyone got a gig of RAM for a Dec Alpha lying around...? stock price for that is more than a new PC....)
1) this is just the same idea as URN (provide identification rather than protocol:hostport). That's a basically good idea, IMHO. But we don't need a multitude of slightly different variants. 2) the DDDS (name resolution that can be based on DNS) is already an Internet (proposed) standard that can be used to resolve arbitrary URIs with DNS support - if the authors so desire. References at an RFC library near you.
The characteristic of P2P is that it allows communication across the network without specific single points of control. The first application that exploited this on an Internet-wide scale was email. Other applications built to this principle have the potential to be as world-changing as email was. P2P is NOT just file copying.
If you know some heavyweight OpenSource people within IBM, ask them to check whatthehell is going on. IBM's big enough, the left hand needs a telescope to see what the right hand's doing - and that's before you try to figure out what others are doing on their behalf.
Since this is about statistics, I thought I'd add some more.... from http://counter.li.org/ Interesting pieces: - Of Linux users, 47% use it at work; a whooping 90% use it at home. Lesson: All the developers use it at home, too. - 99.5%!!! of Linux boxes have Ethernet. This is likely to mean that just about all the home Linux boxes are in multi-machine homes (or DSL). - Red Hat leads the pack, but has only 30% of the market - there are 5 other distributions above the 10% mark - All statistics are biased.
The Linux Counter count has gone down by 7.000 (from 140.000 to 133.000) since Jan 1, 2003 - this is mostly Slashdot users who registered in November 2001, and whose emails have gone bad since then. You're welcome back!
The reason Japan's so hot for IPv6 is that it got rather shortchanged in the IPv4 handout - the ratio of IPv4 addresses to users is much worse than in the US.
The reasons why new IPv4 addresses are harder to come by in Japan than in the US are mainly internal JP politics - the policies of ARIN and APNIC are almost the same, and there's still new IP addresses to be had, if you're able to navigate the paperwork. (For a few years more).
is there any information on whether the DDOS attack on UltraDNS actually affected service? The UltraDNS infrastructure has 16 or so machines on the same IP number. So it's harder to hit all of them. And it's not BIND, so it may be harder to bring down. (not sure it matters - the root DDOS didn't crash BIND either). And of course UltraDNS is typically not serving all of the secondaries for a zone. If anyone has real info....
the IETF is a funny organization. it is 99% "volunteer" (ie non-hired) personnel, and the hired staff just does the clerical stuff, not the technology. one reason for that is that if there is no more purpose for the IETF, we WANT it to be able to go away in reasonable style. I just hope that when the time comes, we have the courage to recognize it.
please don't confuse Verisign the REGISTRY (holder of.net and.com) with Verisign the REGISTRAR (the people responsible for the inaccurate registrations). Revoking Verisign's registrar business would be ironic indeed - it would get them out of the dual role that they had promised to give up, but gave up.org in order to be allowed to continue doing when the contract was renegotiated.
You've identified one of the best points of Jabber: It's a protocol, not a service. AOL and MSN are services that run their own proprietary protocols; if you buy into them, you have no choice but to accept their terms of service. With Jabber, as with any open protocol, if you don't like this provider's service, try another.
(The reason there is so much noise around the DNS is that it's the known example of an open protocol that implies a single service....)
The exchange cited shows that Linux gods are no different from other humans.... The DNS is good at looking up strings. It's a lousy search engine. The idea that one should try to "control" a name in all domains is silly - but happened BECAUSE people tried DNS as a search engine. Personally, I type names into Google when I want to look them up, not my browser bar. There are other angles of attack - see draft-klensin-dns-search, for instance - but currently that works. AND Google gives me enough context to show me WHAT kind of "good vibrations" I'm headed for....
Slashdot Mirror
Natch. IP address range for spammers is not enough.
Insist that spammers turn on the Evil bit - RFC 3514.
(More seriously - RFC 3675, ".sex considered dangerous", gives a little more thinking about why this is a Bad Idea. This is NOT an Internet standard, but it's a well written document.)
if Verisign doesn't try to roll out Sitefinder until the lawsuit is resolved.... that sounds good...
Microsoft is pushing a solution called "Caller ID", which involves putting (wince) XML documents into the DNS telling you how to check the (argh) From: header.
A lot of other people are pushing a solution called SPF, which involves putting text "code snippets" into the DNS telling you how to check the MAIL FROM: envelope return address.
This topic will be discussed at the IETF next week in Seoul, Korea. Hot topic!
Correction: It's .net and .com. .org is handled by PIR, which is a subsidiary of ISOC. ISOC is a not-for-profit.
-----BEGIN PGP SIGNED MESSAGE-----
.... in that until everyone signs their
T re ICmreO16YAoJ3J
- ----END PGP SIGNATURE-----
Hash: SHA1
There's a slight problem
email, you'll have to be willing to handle unsigned email as well as
signed. That leaves the signing people worse off than the non-signing
people (more pain, no gain).
Difficult deployment problem.
-----BEGIN PGP SIGNATURE-----
Version: PGP 7.0.1
iQA/AwUBQCOn5jjI/tvlmNBeEQLIdwCfTzU3AFyy3vAyqJ1
Yl8AGPs6HHxEEGJfkmV857m1
=XHyf
Since real numbers are hard to come by, we get by on the very weak ones we have.
According to the version monitoring page at the Linux Counter, 35 out of the 4862 monitored machines run the 2.0 kernel - 0.7% of the total number of monitored machines.
If that holds true for the (who knows???) 20 million Linux machines out there in the wild, there should be something like 142000 2.0-kernel Linux boxes out there. Perhaps more - the "enthusiasts" who register with the Linux Counter may be more prone to upgrading than others....
Stand up and be counted! The Linux Counter wants you!
the first IETF IM standard to make it through the process was the CPIM package (draft-ietf-impp-cpim-msgfmt). It's a specification on how to interconnect IM systems rather than a complete IM protocol specification.
The other major player in IETF standards-space is SIMPLE - the presence specification documents for that (draft-ietf-simple-presence) are in the RFC Editor's queue.
The nice thing about standards is that there are so many of them.....
Follow the numbers at The Linux Counter!
this has actually happened - not with ezpass but with mobile phones.
in a certain (rather spectacular) murder trial in Norway, one suspect's mobile phone was on an extended trip very far away from the murder site at the time, tracked by your ever-friendly telco's "cell tower association records". We do not know if the suspect went along.....
- 0.8% using 2.0
- 8.9% running 2.2
- 86.5% running 2.4
- 3.5% running 2.6.
There's every reason to believe many people will continue running 2.4 for a LONG time still.(Statistics based on 4503 machines that choose to send in updates. The method is obviously biased.You have been warned.)
the linux counter was "slightly" misconfigured in how many apache servers it allowed to run simultaneously - it went into trashing.
I've re-tuned it (to MaxClients = 32, and KeepAliveTimeout = 2), and it seems to work slightly better.
You can still expect it to be SLOW, though.....
(anyone got a gig of RAM for a Dec Alpha lying around...? stock price for that is more than a new PC....)
The largest uptime ever recorded at the counter was a little more than a thousand days - this was Linux 2.0 running on an Alpha.
It has apparently been turned off now.
1) this is just the same idea as URN (provide identification rather than protocol:hostport). That's a basically good idea, IMHO. But we don't need a multitude of slightly different variants.
2) the DDDS (name resolution that can be based on DNS) is already an Internet (proposed) standard that can be used to resolve arbitrary URIs with DNS support - if the authors so desire.
References at an RFC library near you.
The characteristic of P2P is that it allows communication across the network without specific single points of control.
The first application that exploited this on an Internet-wide scale was email.
Other applications built to this principle have the potential to be as world-changing as email was.
P2P is NOT just file copying.
Installing the patch?
I'll bet AT&T knew this before today.
If you know some heavyweight OpenSource people within IBM, ask them to check whatthehell is going on. IBM's big enough, the left hand needs a telescope to see what the right hand's doing - and that's before you try to figure out what others are doing on their behalf.
Since this is about statistics, I thought I'd add some more.... from http://counter.li.org/
Interesting pieces:
- Of Linux users, 47% use it at work; a whooping 90% use it at home.
Lesson: All the developers use it at home, too.
- 99.5%!!! of Linux boxes have Ethernet. This is likely to mean that just about all the home Linux boxes are in multi-machine homes (or DSL).
- Red Hat leads the pack, but has only 30% of the market - there are 5 other distributions above the 10% mark
- All statistics are biased.
The Linux Counter count has gone down by 7.000 (from 140.000 to 133.000) since Jan 1, 2003 - this is mostly Slashdot users who registered in November 2001, and whose emails have gone bad since then. You're welcome back!
Log in and check if you're registered!
From the Linux Counter:
Kernel Count Percentage
2.0 34 1.0%
2.2 549 15.6%
2.4 2927 82.9%
2.5 16 0.5%
Others 0.1%
No, it's not dead yet.
The reason Japan's so hot for IPv6 is that it got rather shortchanged in the IPv4 handout - the ratio of IPv4 addresses to users is much worse than in the US.
The reasons why new IPv4 addresses are harder to come by in Japan than in the US are mainly internal JP politics - the policies of ARIN and APNIC are almost the same, and there's still new IP addresses to be had, if you're able to navigate the paperwork. (For a few years more).
is there any information on whether the DDOS attack on UltraDNS actually affected service?
The UltraDNS infrastructure has 16 or so machines on the same IP number. So it's harder to hit all of them. And it's not BIND, so it may be harder to bring down. (not sure it matters - the root DDOS didn't crash BIND either).
And of course UltraDNS is typically not serving all of the secondaries for a zone.
If anyone has real info....
According to ftp://ftp.apnic.net/pub/apnic/stats/apnic/apnic-20 02-10-01, China has 25.425.152 addresses.
This is more than MIT has.
the IETF is a funny organization.
it is 99% "volunteer" (ie non-hired) personnel, and the hired staff just does the clerical stuff, not the technology.
one reason for that is that if there is no more purpose for the IETF, we WANT it to be able to go away in reasonable style.
I just hope that when the time comes, we have the courage to recognize it.
please don't confuse Verisign the REGISTRY (holder of .net and .com) with Verisign the REGISTRAR (the people responsible for the inaccurate registrations). .org in order to be allowed to continue doing when the contract was renegotiated.
Revoking Verisign's registrar business would be ironic indeed - it would get them out of the dual role that they had promised to give up, but gave up
You've identified one of the best points of Jabber: It's a protocol, not a service.
AOL and MSN are services that run their own proprietary protocols; if you buy into them, you have no choice but to accept their terms of service.
With Jabber, as with any open protocol, if you don't like this provider's service, try another.
(The reason there is so much noise around the DNS is that it's the known example of an open protocol that implies a single service....)
The exchange cited shows that Linux gods are no different from other humans....
The DNS is good at looking up strings. It's a lousy search engine.
The idea that one should try to "control" a name in all domains is silly - but happened BECAUSE people tried DNS as a search engine.
Personally, I type names into Google when I want to look them up, not my browser bar.
There are other angles of attack - see draft-klensin-dns-search, for instance - but currently that works.
AND Google gives me enough context to show me WHAT kind of "good vibrations" I'm headed for....