Clever new tools for kernel config
on
Linux 2.6.26 Out
·
· Score: 5, Insightful
What I would like to see more emphasis on in future kernels is a discussion of possible clever new tools and methods for configuring the thousands of kernel config options. None of the existing in-kernel-tree or out-of-tree config tools seems ideal.
Oil is a fundamental commodity in economic terms because changes in its price cause changes, with different time lags, in the prices of all other goods and services. An increasing oil price causes inflation. A severe increase causes severe inflation. It's the price increase, not the absolute price, that's important.
The oil price has more than doubled in the last year, and quadrupled in the last three years. There has never before been such an extreme, sustained increase in the oil price. This will cause severe inflation, and the economic consequences will be severe.
This is what's causing all the fuss. The economies of the world are in the early stages of heading into a very severe inflationary recession. Some people go further and anticipate economic collapse, others fear something similar to The Great Depression. The technical term for it is stagflation. Investors look for ways out of trouble, but the consensus is that there is no easy way out of this one. Some investors have therefore panicked. Panic is dangerous because it fuels itself, making the panic worse.
You are not going to see the same impact in Finland because Finland has much higher fuel taxes than in the USA, so the price increase of retail fuels has been much smaller in Finland than in the USA. But recession in the USA, which is the world's largest economy, will be felt in other countries, including Finland.
Why bother? It's not difficult to understand. Monsanto is spending truck loads of cash, lobbying for GM crops everywhere, both inside and outside the US, from NGOs to politicians, all the way to WHO. All in search of maximizing profit. Perfectly legal. Many people criticize Monsanto, many don't. Perhaps, don't bite the hand that feeds...
No conspiracy required. Profit is the driver. When Monsanto sells more GM crops, it makes more profit. US and EU politicians both look after their own, they push laws that support their own. Monsanto is US based, so US and US-sponsored countries get GM crops, EU doesn't. Food is in surplus in both US and in EU, retails at similar price levels. What's the difference? Monsanto. No politics required. Just profit.
The thing is GM agriculture does not provide cheap and plentiful fruits and vegetables.
First of all, there are no GM fruits grown on a commercial scale in the US. There are only a few different GM vegetables grown commercially in the US.
Secondly, US and EU farming enterprises both get huge subsidies from their respective governments. US and EU farming enterprises both generate huge surpluses of food and wine. US farming enterprises pay large amounts to Monsanto for rights to use GM seeds. EU farming enterprises pay nothing to Monsanto. Who is upset? Monsanto.
In the developing world, farmers grow their own extremely cheap non-GM food. Why is it so expensive to import in US and EU? Trade tariffs, imposed because US and EU farmers have lobbied their respective governments to tax imports of cheap food that would compete with their own produce.
I foresee this major oil bubble serving an extremely useful purpose if Iran really were attacked very soon.
Let's assume the expected oil price minus all the speculation, is per IEA calculations, around $70-80/barrel. If Iran gets attacked, it is likely they will attempt to block the Straits of Hormuz by destroying passing oil tankers. If they succeed, it will at a stroke remove 60% of world oil supply from the market. The shock of that happening would cause the oil price to spike to well over $200/barrel, with devastating consequences to world stock markets.
Speculation has already pushed the price up close to $150/barrel. However, the speculative bubble can be quickly and deliberately popped with the help of a trillion dollar hedge fund around the same time that war breaks out and the Straits get blocked. There would be much less of a surge in the oil price.
Who's to say the big speculators haven't been playing the market for this very purpose, in clever anticipation of upcoming attacks on Iran in September?
The good news for us is that after it is all over, the oil price should fall back around $70-80.
I guess you meant "But other admins can still see it [any edit deleted by an administrator]"?
Is ordinary admin (non-oversight) deletion used frequently compared to oversight deletion? I've seen articles where the entire edit history before a certain date containing several years' worth of edits was erased.
What could be causing some edit histories to get out of chronological order as mentioned in this post.
I agree. I use dechunker to re-assemble discussions that have been split into multiple pages. It works well, though there can be a significant delay when reading a multi-page discussion because it tries to be nice to/. bandwidth by waiting a few seconds between loading each successive page. There is also D22D1 for converting D2 to D1. It could be an option for people who prefer D1 if/. ever decides to drop support for D1, though it is still a bit buggy. Sorry, I can't seem to find any links for dechunker and D22D1 right now.
Here is an example. So called "short code" text messages are always chargeable on most UK networks including Vodafone, O2, and Virgin. The maximum charge is UKP 1.50 per incoming message. Any business (usually spammers sending invitations to enter prize draws) can ask a network provider to allocate them a short-code which is a 5-digit unreturnable number (it cannot be used to receive replies). The business can then send subscriber-to-pay text messages and the business collects the profit generated by each text message. There is no way for subscribers to block these short-code messages. You can, however, register your telephone number with the Telephone Preference Service; short-code users should consult the register to block short-code messages being sent to registered numbers, but many don't bother and send them anyway even to registered numbers. If you receive a short code message without any contact details (as often happens with spam), the only way to find out who sent it is to contact your network provider who can decode the short code number giving you the name of the company using it and the contact details.
A spam could be as short as in your example, but as I said, I am discussing my example, so I will point out the actual lengths of the spam emails were just under 1000 bytes.
My calculation was based on a fact, not an assumption, that there was 100MBps of inward bandwidth at my server. No matter many bots the spammers may use to create a huge aggregate output bandwidth, they simply cannot send spam to my server faster than its inward bandwidth of 100Mbps allows. I would also notice if spammers started using any of that inward bandwidth for address guessing; they have not.
I believe that spammers do try to guess Hotmail addresses, but as you have noticed in your other comment, my example concerns the spamming of private addresses at my server, not Hotmail's.
Cross-site scripted request forgery would work if the victims had browsed other websites, but the victims in my example whom I know personally, apparently did not browse anywhere except Hotmail and Googlemail on the 7th and 8th.
No real alternative explanations are left, except a security breach c/o Hotmail.
Sorry, my intention in this thread has been to discuss my example of a small number of private email addresses being spammed at the same domain. In my example there were virgin addresses, meaning that no emails were ever sent or received by those addresses until the spam arrived, so the possibilities you suggest of bots guessing the addresses and multiple recipients are not relevant to explaining my example.
In your reply to my example, you appeared to imply the explanation for how the spammers guessed the addresses was that they could "cover the full address space fairly easily and quickly". That is false for my example because the addresses at the domain are unguessable without doing a brute-force search of the address space at the domain. The spammers would have to test O(k^N) addresses at the domain before they could be expected, on average, to hit the spammable private addresses (the only thing that varies between the different private addresses is the local part [see precise definition in RFC2822], not the domain). In my example, it simply is not physically possible for spammers to do that economically, i.e. in a sensible amount of time less than 16 years, and to gain more than just a few spammable addresses at one domain. There is no evidence that spammers in my example have even attempted such a foolish and practically impossible task.
"Actually, they don't go *sequentially* anymore, because that's too easy to detect, but they do cover the full address space fairly easily and quickly."
No, they don't. No spammer or group of spammers can "cover the full address space fairly easily and quickly". It's simply not possible because the address space grows exponentially large with address length. The number of different possible addresses is k^N (k raised to the power of N), where k is the number of different possible characters that are available to use anywhere in the initial part of any email address which is N-characters long. RFC 2822 defines k=82 characters (usually k=56 due to case-insensitivity) that can be used in the initial part of any email address and the maximum length of the initial part as 64 characters. If k=36 and N=8 for example, the address space contains over 2821 giga addresses (2,821,109,907,456 to be exact). Even if the spammers used a minimum of 100 bytes (message length+SMTP) to send a tiny spam email to a recipient with an abundant 100 MBps inward bandwidth, it would take them over 16 years to cover the address space for just that one recipient.
Because all of the addresses were deliberately chosen to be obscure and hard to guess. For example, none of the addresses contain words that appear in any dictionaries.
Anyway, thank you for your question. It is good to ask oneself these sorts of questions to see if there is a possible alternative explanation for why these particular addresses were spammed, i.e. one that does not involve spammers accessing Hotmail's internal systems via a security breach, but after careful review of the facts there really does not seem to be anygoodalternative explanation. The only obvious explanation that fits the facts (for details, see my four earlier replies further down in this thread) is that spammersdidget these addresses by accessing Hotmail's internal systems via a security breach.
1) The addresses stored at Hotmail were private disposable addresses at domains under my control There is no evidence of a security breach at any of the systems relating to these domains.
2) No email had ever been sent or received by those addresses. They were virgin addresses.
It seems there is no obvious explanation other than a security breach at Hotmail's systems. The only way for people outside Hotmail to detect it is by using private disposable addresses, so most people who use ordinary non-disposable or public addresses will have no way of knowing or proving the spam problem is caused by Hotmail itself.
1) I'm not sure what you mean. As I said, these people all use Hotmail with the unique addresses stored there in their personal contacts lists. Can you clarify your question?
2) Yes, if emails were ever transmitted, but several of the spammed addresses had never received any emails prior to these spams
A1) Because the addresses that were spammed were not addresses "at" hotmail. They were unique private and unpublished addresses these people stored in their personal contacts lists at hotmail.
A2) Because they accessed hotmail only via https which is not easily sniffable.
I think it is much more likely that Hotmail's IT systems have been compromised following a security breach by the spammers. I have indirect evidence that this has happened.
I and some other people I know give out unique disposable email addresses to our contacts. There is a different unique address for each of our friends and family.
Yesterday I and they received spam emails sent to several of the disposable email addresses. This points us to several of our friends and family as having had their email address lists stolen by spammers.
The common factors are:
They all accessed Hotmail on 7th or 8th.
Their email contacts are stored on Hotmail.
They all use Apple Macs and browse using Safari. There is no evidence that any of these Macs have been compromised.
There is therefore no obvious way for the spammers to have obtained these unique email addresses, except by the spammers accessing Hotmail's internal systems via a security breach. The security breach could be technical (an unpatched vulnerability in one of Hotmail's systems) or human (one of their members of Hotmail's (outsourced?) staff copied the contents of some/all of their servers and sold them to the spammers)
My point was that the quote you misleadingly supplied does not prove that international laws exist to limit broadcasting of TV transmissions to intended audience.
"I apologise if I have pasted the wrong version - I found 3"
Then I challenge you to post the other two and to explain how they relate to international laws which specifically limit broadcasting of TV transmissions to intended audience.
"the point of the post is to show that there are rules applying to broadcasts."
The issue is not whether there are "rules applying to broadcasts" (of course there are), but rather, whether there are international laws limiting the broadcasting of TV transmissions to intended audience as you appeared to imply.
"I do not believe I have made a false claim."
You wrote:
"There ARE international laws regarding the broadcasting of TV transmissions. To limit the transmission to its intended audience they [BBC] feel that they must use DRM."
You appeared to imply there are "international laws" to limit the broadcasting of TV transmissions "to intended audience", which is false.
"A broadcast license is a specific type of spectrum licence that grants the licensee the right to use a portion of the radio frequency spectrum in a given geographical area for broadcasting purposes.
The use of particular frequency is limited to geographic area. This fact, when combined with the licensed transmission power, limits the region covered by that transmission.
Both of these technicalities are controlled and enforced by international law, through national bodies which work with the ITU and other agencies, to prevent adjacent channel interference and to ensure effective use worldwide of the electro-magnetic spectrum. The intended audience for the BBC terrestrial service is the UK and the power of individual transmissions is partly decided by the footprint required by a specific transmitter."
The fact that broadcasters use different frequencies in neighbouring regions has nothing to do with limiting audience per se. It is to avoid harmful interference between different transmissions in neighbouring areas. You are confusing various mainly technical issues relating to physical transmission coverage with the non-technical issue of who is the intended audience.
In fact, the BBC Charter does not give a definition of who is the intended audience for any type of BBC broadcasting. The only references to "audience" are for the BBC Audience Councils which are to advise the BBC Trust and represent licence payers. For most of the BBC's TV and radio broadcasts the intended audience is presumably licence payers at least (but noting that "overspill" of BBC analog broadcasts into foreign countries has not been entirely due to technical limitations and has often been very helpful, some would say coincidentally, to UK interests). Anyway, the intended audience is different from "the UK". Indeed, for BBC World Service broadcasting, the intended audience is clearly the residents of foreign countries.
"Now, take the example of something like the Olympic Games or any other event which is of interest globally."
I accept that sports and entertainment events can be licensed to a broadcaster for transmission only within specified geographical areas, and that these agreements are often international in scope. My understanding, however, is that these agreements are private contracts between two parties - the rights-holder for the event and the broadcaster - and are therefore subject to the national law of contract in the particular jurisdiction specified in the agreement.
"I do not like DRM or the Windows OS but I can understand why the BBC are planning to use it and why they are concentrating initially on compatibility with the Windows OS."
"I would suggest that the BBC does have a moral obligation to make profit from its sales of programs overseas although I will agree that it is open to some flexibility of interpretation."
Then, perhaps because English is your second language, you failed to understand the part of the BBC Charter that you quoted. The language of the BBC Charter is clear and unambiguous. It has no implication whatsoever that the BBC has a duty to maximise profit from the programs it makes, which was your original claim. I note you appear to have conceded your original claim is false, and are now making a different claim that the BBC has a "moral obligation" to "make profit" from its sales of programs "overseas". There is absolutely nothing in the BBC Charter to suggest it has any "moral obligation" to do as you suggest. And do you understand the meaning of "overseas"? Consider, for example, that the Republic of Ireland is a sovereign nation that is not "overseas" relative to the UK. I sincerely hope you do not sit on the BBC Trust.
"The Licencee will implement all necessary work on maintenance and adjustments of its radio equipment, in order to ensure that the broadcast of the Licencee is in accordance with the technical Rules of the International Telecommunications Union (ITU)."
That quote with its obviously incorrect English spelling is in fact from the website of the Communications Regulatory Agency of Bosnia and Herzegovina(found by Google search).. It is misleading of you to quote it (without even an honest citation) as if it were relevant to this discussion about the BBC.
In addition, there are various international protocols, which are covered and enforced by international law, to which the UK is a signatory. Included in such protocols are those which dictate frequency allocations and deconfliction, output powers, transmission of callsigns and/or identifying features, etc.
Irrelevant. You implied there are international laws limiting the broadcasting of TV transmissions "to intended audience". There are no such laws. Please stop trying to support your false claim.
There are no "international laws" limiting the broadcasting of TV transmissions "to intended audience" as you seem to be implying. It is UK law that anybody in the UK possessing television receiving equipment must pay for a television licence.
"secondly, the BBC is are duty bound to the licence payers (i.e. you) to try to make as much profit from the programs that your money has helped fund."
That is not true. The BBC does not have any duty to maximise profit from the programs it makes. If you don't believe me, read the BBC Charter.
You are asking us to believe that Linux market share in January was 0.35%, and then more than doubled three months later to 0.80% in April? Sorry, but those figures are absurd and unbelievable. I'd call them "crude estimates" at best because the website which produces them admits "We collect data from the browsers of site visitors to our exclusive on demand network of small to medium enterprise live stats customers." They are using the user-agent string to determine OS type of the browser. In other words, the figures are based on a flawed methodology because the user-agent string is not reliable data, even more so for estimating Linux usage because many Linux users change it for use with web ad cutters, etc. Also, the figures do not include the large number of Linux servers, although the website misleadingly calls the figures "Linux market share", instead of "Linux desktop market share". I think those figures are no more accurate than ones from the other sources quoted.
While you're being sarcastic, you might like to know that desktop Linux's share is comparable to or even exceeding Apple's share and as also reported desktop Linux's share will reach 7.5 percent by 2008. I'm not sure it is a positive step in a democracy for a 7.5% minority to be ignored by a quasi-state-sponsored broadcaster (anyone who owns television receiving equipment is required by law to buy a licence every year, even if they subscribe to pay-tv channels and never watch BBC).
Slashdot Mirror
What I would like to see more emphasis on in future kernels is a discussion of possible clever new tools and methods for configuring the thousands of kernel config options. None of the existing in-kernel-tree or out-of-tree config tools seems ideal.
The oil price has more than doubled in the last year, and quadrupled in the last three years. There has never before been such an extreme, sustained increase in the oil price. This will cause severe inflation, and the economic consequences will be severe.
This is what's causing all the fuss. The economies of the world are in the early stages of heading into a very severe inflationary recession. Some people go further and anticipate economic collapse, others fear something similar to The Great Depression. The technical term for it is stagflation. Investors look for ways out of trouble, but the consensus is that there is no easy way out of this one. Some investors have therefore panicked. Panic is dangerous because it fuels itself, making the panic worse.
You are not going to see the same impact in Finland because Finland has much higher fuel taxes than in the USA, so the price increase of retail fuels has been much smaller in Finland than in the USA. But recession in the USA, which is the world's largest economy, will be felt in other countries, including Finland.
Why bother? It's not difficult to understand. Monsanto is spending truck loads of cash, lobbying for GM crops everywhere, both inside and outside the US, from NGOs to politicians, all the way to WHO. All in search of maximizing profit. Perfectly legal. Many people criticize Monsanto, many don't. Perhaps, don't bite the hand that feeds...
No conspiracy required. Profit is the driver. When Monsanto sells more GM crops, it makes more profit. US and EU politicians both look after their own, they push laws that support their own. Monsanto is US based, so US and US-sponsored countries get GM crops, EU doesn't. Food is in surplus in both US and in EU, retails at similar price levels. What's the difference? Monsanto. No politics required. Just profit.
First of all, there are no GM fruits grown on a commercial scale in the US. There are only a few different GM vegetables grown commercially in the US.
Secondly, US and EU farming enterprises both get huge subsidies from their respective governments. US and EU farming enterprises both generate huge surpluses of food and wine. US farming enterprises pay large amounts to Monsanto for rights to use GM seeds. EU farming enterprises pay nothing to Monsanto. Who is upset? Monsanto.
In the developing world, farmers grow their own extremely cheap non-GM food. Why is it so expensive to import in US and EU? Trade tariffs, imposed because US and EU farmers have lobbied their respective governments to tax imports of cheap food that would compete with their own produce.
Let's assume the expected oil price minus all the speculation, is per IEA calculations, around $70-80/barrel. If Iran gets attacked, it is likely they will attempt to block the Straits of Hormuz by destroying passing oil tankers. If they succeed, it will at a stroke remove 60% of world oil supply from the market. The shock of that happening would cause the oil price to spike to well over $200/barrel, with devastating consequences to world stock markets.
Speculation has already pushed the price up close to $150/barrel. However, the speculative bubble can be quickly and deliberately popped with the help of a trillion dollar hedge fund around the same time that war breaks out and the Straits get blocked. There would be much less of a surge in the oil price.
Who's to say the big speculators haven't been playing the market for this very purpose, in clever anticipation of upcoming attacks on Iran in September?
The good news for us is that after it is all over, the oil price should fall back around $70-80.
Is ordinary admin (non-oversight) deletion used frequently compared to oversight deletion? I've seen articles where the entire edit history before a certain date containing several years' worth of edits was erased.
What could be causing some edit histories to get out of chronological order as mentioned in this post.
I agree. I use dechunker to re-assemble discussions that have been split into multiple pages. It works well, though there can be a significant delay when reading a multi-page discussion because it tries to be nice to /. bandwidth by waiting a few seconds between loading each successive page. There is also D22D1 for converting D2 to D1. It could be an option for people who prefer D1 if /. ever decides to drop support for D1, though it is still a bit buggy. Sorry, I can't seem to find any links for dechunker and D22D1 right now.
Here is an example. So called "short code" text messages are always chargeable on most UK networks including Vodafone, O2, and Virgin. The maximum charge is UKP 1.50 per incoming message. Any business (usually spammers sending invitations to enter prize draws) can ask a network provider to allocate them a short-code which is a 5-digit unreturnable number (it cannot be used to receive replies). The business can then send subscriber-to-pay text messages and the business collects the profit generated by each text message. There is no way for subscribers to block these short-code messages. You can, however, register your telephone number with the Telephone Preference Service; short-code users should consult the register to block short-code messages being sent to registered numbers, but many don't bother and send them anyway even to registered numbers. If you receive a short code message without any contact details (as often happens with spam), the only way to find out who sent it is to contact your network provider who can decode the short code number giving you the name of the company using it and the contact details.
Following up in the original thread
My calculation was based on a fact, not an assumption, that there was 100MBps of inward bandwidth at my server. No matter many bots the spammers may use to create a huge aggregate output bandwidth, they simply cannot send spam to my server faster than its inward bandwidth of 100Mbps allows. I would also notice if spammers started using any of that inward bandwidth for address guessing; they have not.
I believe that spammers do try to guess Hotmail addresses, but as you have noticed in your other comment, my example concerns the spamming of private addresses at my server, not Hotmail's.
Cross-site scripted request forgery would work if the victims had browsed other websites, but the victims in my example whom I know personally, apparently did not browse anywhere except Hotmail and Googlemail on the 7th and 8th.
No real alternative explanations are left, except a security breach c/o Hotmail.
In your reply to my example, you appeared to imply the explanation for how the spammers guessed the addresses was that they could "cover the full address space fairly easily and quickly". That is false for my example because the addresses at the domain are unguessable without doing a brute-force search of the address space at the domain. The spammers would have to test O(k^N) addresses at the domain before they could be expected, on average, to hit the spammable private addresses (the only thing that varies between the different private addresses is the local part [see precise definition in RFC2822], not the domain). In my example, it simply is not physically possible for spammers to do that economically, i.e. in a sensible amount of time less than 16 years, and to gain more than just a few spammable addresses at one domain. There is no evidence that spammers in my example have even attempted such a foolish and practically impossible task.
Anyway, thank you for your question. It is good to ask oneself these sorts of questions to see if there is a possible alternative explanation for why these particular addresses were spammed, i.e. one that does not involve spammers accessing Hotmail's internal systems via a security breach, but after careful review of the facts there really does not seem to be any good alternative explanation. The only obvious explanation that fits the facts (for details, see my four earlier replies further down in this thread) is that spammers did get these addresses by accessing Hotmail's internal systems via a security breach.
1) Just to clarify: The addresses were stored only at Hotmail.
2) No email had ever been sent or received by those addresses. They were virgin addresses.
It seems there is no obvious explanation other than a security breach at Hotmail's systems. The only way for people outside Hotmail to detect it is by using private disposable addresses, so most people who use ordinary non-disposable or public addresses will have no way of knowing or proving the spam problem is caused by Hotmail itself.
2) Yes, if emails were ever transmitted, but several of the spammed addresses had never received any emails prior to these spams
A2) Because they accessed hotmail only via https which is not easily sniffable.
I and some other people I know give out unique disposable email addresses to our contacts. There is a different unique address for each of our friends and family.
Yesterday I and they received spam emails sent to several of the disposable email addresses. This points us to several of our friends and family as having had their email address lists stolen by spammers.
The common factors are:
There is therefore no obvious way for the spammers to have obtained these unique email addresses, except by the spammers accessing Hotmail's internal systems via a security breach. The security breach could be technical (an unpatched vulnerability in one of Hotmail's systems) or human (one of their members of Hotmail's (outsourced?) staff copied the contents of some/all of their servers and sold them to the spammers)
Then I challenge you to post the other two and to explain how they relate to international laws which specifically limit broadcasting of TV transmissions to intended audience.
The issue is not whether there are "rules applying to broadcasts" (of course there are), but rather, whether there are international laws limiting the broadcasting of TV transmissions to intended audience as you appeared to imply.
You wrote:
You appeared to imply there are "international laws" to limit the broadcasting of TV transmissions "to intended audience", which is false.
The fact that broadcasters use different frequencies in neighbouring regions has nothing to do with limiting audience per se. It is to avoid harmful interference between different transmissions in neighbouring areas. You are confusing various mainly technical issues relating to physical transmission coverage with the non-technical issue of who is the intended audience.
In fact, the BBC Charter does not give a definition of who is the intended audience for any type of BBC broadcasting. The only references to "audience" are for the BBC Audience Councils which are to advise the BBC Trust and represent licence payers. For most of the BBC's TV and radio broadcasts the intended audience is presumably licence payers at least (but noting that "overspill" of BBC analog broadcasts into foreign countries has not been entirely due to technical limitations and has often been very helpful, some would say coincidentally, to UK interests). Anyway, the intended audience is different from "the UK". Indeed, for BBC World Service broadcasting, the intended audience is clearly the residents of foreign countries.
I accept that sports and entertainment events can be licensed to a broadcaster for transmission only within specified geographical areas, and that these agreements are often international in scope. My understanding, however, is that these agreements are private contracts between two parties - the rights-holder for the event and the broadcaster - and are therefore subject to the national law of contract in the particular jurisdiction specified in the agreement.
I agree with you.
You are asking us to believe that Linux market share in January was 0.35%, and then more than doubled three months later to 0.80% in April? Sorry, but those figures are absurd and unbelievable. I'd call them "crude estimates" at best because the website which produces them admits "We collect data from the browsers of site visitors to our exclusive on demand network of small to medium enterprise live stats customers." They are using the user-agent string to determine OS type of the browser. In other words, the figures are based on a flawed methodology because the user-agent string is not reliable data, even more so for estimating Linux usage because many Linux users change it for use with web ad cutters, etc. Also, the figures do not include the large number of Linux servers, although the website misleadingly calls the figures "Linux market share", instead of "Linux desktop market share". I think those figures are no more accurate than ones from the other sources quoted.
While you're being sarcastic, you might like to know that desktop Linux's share is comparable to or even exceeding Apple's share and as also reported desktop Linux's share will reach 7.5 percent by 2008. I'm not sure it is a positive step in a democracy for a 7.5% minority to be ignored by a quasi-state-sponsored broadcaster (anyone who owns television receiving equipment is required by law to buy a licence every year, even if they subscribe to pay-tv channels and never watch BBC).