Most academic journals are well-known within the fields they serve. They don't need additional "reputation points" by being cited in some student's footnotes.
Knowing whether a system that shuts itself off every few hours all by itself and has a 3G modem that just stops working after an hour or less is broken or not, and getting Dell to deal with the problem.
I've had a similar problem with my Linksys USB wifi adapter that uses the Ralink 2500 chipset. This device worked flawlessly though Ubuntu 9.04, but has not worked reliably since then. In WPA2 networks like mine, it won't establish a connection at full speed and will randomly disconnect for no apparent reason. I'm now running Maverick whose kernel seems to include a fix for this problem, but it's taken almost two full years to get this issue resolved.
While I'm a long-time Linux user and supporter of open source, this example is precisely why Linux fails to establish itself with ordinary users over the long run. Even given my dedication to Linux and open source, having a device that suddenly stops working for no good reason is exasperating to say the least. Unfortunately the Ralink chipset is very common, so it's hard to find competing devices. I always choose Intel wireless if given the opportunity because its support in Linux has been rock-solid. I've done numerous searches to find an equivalent PCI or USB device using the Intel chipsets, or the Atheros ones which are highly rated, and they're very hard to find. Mainstream manufacturers like Linksys use Ralink devices which work under Windows but are more flaky under Linux.
Until every mainsteam hardware device works flawlessly out of the box, and continues to work flawlessly regardless of changes to the kernel, even the most adventurous Linux users will ultimately fall by the wayside. Personally I think it's unconscionable to rewrite drivers and interfaces between kernel releases without sufficient testing to insure that the installed base of hardware continues to work correctly. Microsoft gets a lot of derision for its commitment to backwards-compatibility, but I'd like to see a bit more emphasis on maintaining compatibility with the installed base among Linux developers.
Is your response a sad acknowledgement that most people are incapable of learning multiple variations of the same skill?
Yes and no. One problem with computer training is that it's largely fairly rote. "Click X to do Y" types of training. Unfortunately American schools seem to think that this is the best way to train people looking for clerical and similarly skilled positions to deal with the computer demands that they will face on the job. It's still a Microsoft world, and most "training" consists of telling people how to do something with Word or Excel, not how to think about the actual "skills" involved as you suggest.
On the other hand, I also think many people fear their computers and worry that doing anything out of the ordinary will "break" them. I'm always impressed by how many fall for scams like the phony virus scan. Just the other day I was speaking with a friend whose wife fell for this. They're both attorneys, so it's not a question of intelligence. It has much more to do with trust. The profusion of Windows malware over the years has led people not to trust their computers. When told that they need to take some step to further protect themselves, they don't trust Windows or the software they've already installed to protect them.
So, in answer to your broader question, I don't think a lot of people think about computers in terms of generic "functions" or "skills" as you put it. Perhaps this is a generational thing (I'm 60, and most of my friends are 50+.), but I've had many conversations with people in support positions who encounter the reactions I mentioned in my earlier posting every day. The people they support are usually not as old as me.
I'd also say your Ford/Chevy analogy somewhat misses the point. It's not a question of where the speedometer is located. Differences between Outlook and Evolution concern things like address books. What if the person is tied into a company-wide address book on an Exchange server? What if they're used to using Sharepoint? What about workflow applications? There's a lot of "stuff" out there that relies on a Windows infrastructure; moving to Linux is not easy for people in situations like these.
Eh? What exactly is harder in Ubuntu than it is on Windows?
Opening a document someone sent you from work in Word 2007 and editing it in Word 2007. Where's the ribbon?
The reply to this is not to say, "well, then, just open it in Open Office." It may look similar to its version in Word, it may not. When Open Office looks exactly like Word 2007+, and works exactly the same way as well, then you can ask your question again.
How about people who have used Outlook for years in an office environment? Nothing, not even Evolution, will look exactly the same and work identically.
You've obviously not dealt with people for whom moving a icon from one place on the desktop to another results in complaints that "my Windows is broken," usually followed by, "I can't do my work until it's fixed."
I don't know what the situation is today, but in the past this functionality was definitely not available to Verizon Wireless users. VZ disabled direct connections to the USB port so the only way to transfer pictures to and from the phone was via a data connection over its network or via an SD card if the phone has that capability. In general you had to email or text the photos to yourself and pay any consequent data charges.
I'd venture to guess that it's still the case that most cellphone users never use their USB cables to transfer anything. They take pictures on their phones and use them as local wallpapers or hold them up so their friends can see them on the screen. The market for ringtones and sales of music tracks for phones is estimated to exceed $10B in the next few years. I have a technologically sophisticated teen-aged daughter who could move stuff to any from her phone over USB the same way she does with her Sansa player. Most of her friends wouldn't have the slightest idea how to accomplish any of that.
I only buy unlocked phones to avoid these problems.
I'd agree that your computer owned by you on an Internet connection you purchased and maintain is analogous to a printing press. Your blog on a computer hosted by someone else who has entered into a specific contract with yet another entity (the hosting company) doesn't qualify in my mind. If you decide that your information is sufficiently important, and perhaps sufficiently controversial, that it might raise someone's hackles, host it yourself and deal with the political and legal consequences that might follow. Placing your content on a server owned by some random guy with 72,000 other people doesn't suggest to me that you think your speech is worth protecting.
Hosting blogs is probably conceptually closer to publishing a magazine. If you write an incisive article on the misadventures of some government agency and publish it in a magazine where someone else has posted a picture on page 37 of some eight-year-old sucking a guy's dick, do you really think the authorities are going to be worrying about your First Amendment rights when they show up at the magazine publisher's offices with a warrant? Hosting providers are treated rather akin to common carriers when it comes to copyright infringement (via the DMCA take-down procedure). But I don't think that should absolve the provider of all editorial responsibilities any more than it would if he or she were publishing a print magazine with multiple independent contributors.
Mr. Blog-Hoster seemed to think he could just sit back and collect his AdSense revenues without taking any responsibility for the material he was hosting. I think that's nuts.
As a former university professor, I find the limits on access to scientific papers appalling. Companies like Wiley, Elsevier and the like have had nice little fiefdoms that depended on the captive audience of academic libraries. I don't begrudge them charging for articles written in the past year or two, but I find it offensive that I cannot read a decade-old piece of academic research today for free. I might pay a dollar or two for an article of that age, but $20 or more? Nope, not gonna happen.
Let's not forget the huge indirect subsidies these publishers get from places like the National Institutes of Health; research proposals usually include a line-item to cover the costs of publication.
So what happens when a student with little or no funds wants to read an article? Is he or she going to pony up the $20+ fee? No, they're going take the library's copy to the photocopier. How does the publisher earn anything on that transaction?
I once let my Economist subscription lapse for a couple of weeks. When I renewed they sent me the missing issues automatically.
There are just too many free alternatives at the moment to make a paywall work for a newspaper. Someday the New York Times might make me start paying, and there's some chance I might if it's cheap enough. If it's anything like Murdoch's $3/week, I'll just send some more money to NPR and keep my Economist subscription.
After reading that thread at webhostingtalk, I'm going to guess the Patriot Act was invoked. It has that air about it. If it was child porn, would Burst be facing an obvious gag order from the government?
The server operator sounds incredibly clueless. Did it never cross his mind that offering free hosting to 70,000+ accounts might put him in the position of hosting material that might violate some laws?
This doesn't sound like the usual copyright-infringement case; it looks much more serious than that.
I'd expect the next thing will be to find ways to compromise E-mail accounts en masse (hacking a server at a free E-mail provider and using accounts, or compromising a backbone SMTP server.)
Just this week I've seen two spams that appear to have come from real accounts at AOL and Hotmail. I know for a fact that the first was a real account since it belonged to someone subscribed to a limited-membership listserver I manage. The second was from an account I knew nothing about, but it was essentially identical in content to the first. Both came from the mail providers' own SMTP servers. They also both have a X- header identifying the originator of the message, and both came from an IP in 115.48/12 described as "China Unicom Henan province network."
I don't really care about OpenSolaris, but I have been a happy user of VirtualBox since before it was acquired by Sun. Sun developed some nice, but proprietary, tweaks to VirtualBox in areas like graphics drivers. I do see development continuing as I get prompted to upgrade fairly regularly, but I've been nervous that VirtualBox will also eventually be treated as roadkill by Oracle. Obviously there will always be a free implementation since the "open-source edition" is GPL-licensed.
I can understand Oracle's lack of interest in OpenSolaris since they've supported Linux for a long time now. (Hell, they even compete directly against RedHat with their Oracle Enterprise Linux distribution.) I do wonder, though, whether they'll stay committed to VirtualBox down the road.
Obviously Deval Patrick supports this obvious attempt at pandering to the Scott Brown brigade since he signed the bill into law. What about our other three candidates for Governor? There's no mention of this law on any of their websites.
I'm amused this appears on the same page as the discussion about liability for breaches. We all know that enforcing large, public, and expensive fines is the only solution that corporations will pay any attention to. In fact, why not make CIOs (and CEOs?) personally liable.
I was talking about foreign press coverage in the mainstream media, not whether the issue is being discussed on Australian advocacy sites. My point was simply that the Internet censorship issue never appeared in foreign reports about the transition from Rudd to Gillard.
You may be surprised to hear that I saw nothing about the Internet filtering issues when Gillard took office. At the time I was in Europe and reading the International Herald-Tribune and the Financial Times. The conflicts over carbon trading and the mining tax were the ones described in the foreign media as the keys to Rudd's departure. Here's a representative sample from another, well-respected paper: http://www.csmonitor.com/World/Asia-Pacific/2010/0624/Julia-Gillard-takes-helm-in-Australia-after-Kevin-Rudd-ouster
I read some of the follow-on stories in the days after; Internet filtering was never mentioned.
I don't know how he can write, "None of the companies involved would disclose details of the suits...," with a straight face. These are lawsuits, after all, so presumably they've been filed with a court and become public information. Unfortunately in the world of tech journalism, "research" consists of asking consultants like The Yankee Group what they think rather than actually, you know, reading documents.
Right now, the U.S. is lagging far behind many other countries in communications infrastructure. Why? Statistically, the answer is equally clear: in nearly every case where another country had, on average, faster and cheaper network communication than the United States, it was correlated with 2 things: (1) "net neutrality". To be clear about that, it meant the telecom industry acts as a simple carrier of bits, like a telephone company, and does not get entangled in content, policing, or tiered pricing structures. (2) Shared backbone: other companies were allowed to lease infrastructure at competitive rates. By law, that is.
Got any citations? Back in the 1980's I wrote some papers on the socio-economic and political factors affecting cross-national variations in television and radio penetration rates. I've considered returning to this question with regard to the Internet. So I'd love to see some decent academic research on this question if it exists.
But I don't see the meagre profits from the app store, or from video rentals, or similar low-margin operations (possibly, in fact probably, including their music sales) as being reasonable; they strike me as ignorant of how businesses think. Yes, it's all about money with Apple (and every other successful corporation), but for Apple, the money is in the device sales.
Might that be a bit short-sighted over the longer term? Apple might primarily a hardware company today with a smaller content marketing arm, but ten years from now, the profitability of the two might well be reversed. Apple won't be able to dominate the mobile devices market in the future as it has in the past, but it could still be a major player in the ever-expanding business of content distribution. I also wouldn't be surprised to see the often-mooted Apple/Disney deal happen in the years ahead. Disney has already signed on for AppleTV. I bet those programs won't be available via h.264 over Flash.
I thought it was "branding eye candy" when I first used Office 2007 myself. I looked and looked for the equivalent of the File menu, then clicked the button when I couldn't think of any other solution.
If you host your own mail server so that the mail is never stored on the Internet, then by definition that server has to have presence on the Internet, and again, the risk is still there.
You made a lot of good points, but I have to take issue with this one. There's no reason why the server that stores the mail needs to be visible on the Internet. Usually I have three machines involved -- one running a secure store-and-forward SMTP listener, one dedicated to scanning for spam and viruses, and a third where the mailboxes actually reside. Only the first of these has any exposure whatsoever, and depending on volumes the scanning and storage functions might be collapsed onto a single, internal server.
Some of my clients use third-parties like Postini, but the clients' firewalls and servers are configured to accept inbound SMTP transfers only from Postini's IP block.
Webmail creates the potential for intrusions, of course, though again we separate the server with the external client software from the machines where the mail is actually stored.
I did a bit of digging, and all the data on host counts appear to be compiled from the ISC Domain Survey. According to the summary on that page, "The Domain Survey attempts to discover every host on the Internet by doing a complete search of the allocated address space and following links to domain names." This would seem to exclude hosts without reverse-DNS records, but I'd need to read the complete study methodology before I could comment intelligently.
I also looked to see if there were easily-available figures on the number of IP addresses allocated by country but couldn't find any.
Regardless of the method for counting hosts, it still seems quite likely that US hosts make up considerably more than 13% of all hosts worldwide.
Slashdot Mirror
Somehow, I don't think you work in marketing.
Thanks be to God!
MailScanner has had the option of "disarming" scripts in email for years now.
Allowing scripts in email messages is as bad as allowing them in advertisements on web sites.
Most academic journals are well-known within the fields they serve. They don't need additional "reputation points" by being cited in some student's footnotes.
Knowing whether a system that shuts itself off every few hours all by itself and has a 3G modem that just stops working after an hour or less is broken or not, and getting Dell to deal with the problem.
I've had a similar problem with my Linksys USB wifi adapter that uses the Ralink 2500 chipset. This device worked flawlessly though Ubuntu 9.04, but has not worked reliably since then. In WPA2 networks like mine, it won't establish a connection at full speed and will randomly disconnect for no apparent reason. I'm now running Maverick whose kernel seems to include a fix for this problem, but it's taken almost two full years to get this issue resolved.
While I'm a long-time Linux user and supporter of open source, this example is precisely why Linux fails to establish itself with ordinary users over the long run. Even given my dedication to Linux and open source, having a device that suddenly stops working for no good reason is exasperating to say the least. Unfortunately the Ralink chipset is very common, so it's hard to find competing devices. I always choose Intel wireless if given the opportunity because its support in Linux has been rock-solid. I've done numerous searches to find an equivalent PCI or USB device using the Intel chipsets, or the Atheros ones which are highly rated, and they're very hard to find. Mainstream manufacturers like Linksys use Ralink devices which work under Windows but are more flaky under Linux.
Until every mainsteam hardware device works flawlessly out of the box, and continues to work flawlessly regardless of changes to the kernel, even the most adventurous Linux users will ultimately fall by the wayside. Personally I think it's unconscionable to rewrite drivers and interfaces between kernel releases without sufficient testing to insure that the installed base of hardware continues to work correctly. Microsoft gets a lot of derision for its commitment to backwards-compatibility, but I'd like to see a bit more emphasis on maintaining compatibility with the installed base among Linux developers.
Is your response a sad acknowledgement that most people are incapable of learning multiple variations of the same skill?
Yes and no. One problem with computer training is that it's largely fairly rote. "Click X to do Y" types of training. Unfortunately American schools seem to think that this is the best way to train people looking for clerical and similarly skilled positions to deal with the computer demands that they will face on the job. It's still a Microsoft world, and most "training" consists of telling people how to do something with Word or Excel, not how to think about the actual "skills" involved as you suggest.
On the other hand, I also think many people fear their computers and worry that doing anything out of the ordinary will "break" them. I'm always impressed by how many fall for scams like the phony virus scan. Just the other day I was speaking with a friend whose wife fell for this. They're both attorneys, so it's not a question of intelligence. It has much more to do with trust. The profusion of Windows malware over the years has led people not to trust their computers. When told that they need to take some step to further protect themselves, they don't trust Windows or the software they've already installed to protect them.
So, in answer to your broader question, I don't think a lot of people think about computers in terms of generic "functions" or "skills" as you put it. Perhaps this is a generational thing (I'm 60, and most of my friends are 50+.), but I've had many conversations with people in support positions who encounter the reactions I mentioned in my earlier posting every day. The people they support are usually not as old as me.
I'd also say your Ford/Chevy analogy somewhat misses the point. It's not a question of where the speedometer is located. Differences between Outlook and Evolution concern things like address books. What if the person is tied into a company-wide address book on an Exchange server? What if they're used to using Sharepoint? What about workflow applications? There's a lot of "stuff" out there that relies on a Windows infrastructure; moving to Linux is not easy for people in situations like these.
Turned out it was some bug using NVIDEA cards with dual monitors.
A common setup found on, oh maybe, 0.05% of all Windows machines.
Eh? What exactly is harder in Ubuntu than it is on Windows?
Opening a document someone sent you from work in Word 2007 and editing it in Word 2007. Where's the ribbon?
The reply to this is not to say, "well, then, just open it in Open Office." It may look similar to its version in Word, it may not. When Open Office looks exactly like Word 2007+, and works exactly the same way as well, then you can ask your question again.
How about people who have used Outlook for years in an office environment? Nothing, not even Evolution, will look exactly the same and work identically.
You've obviously not dealt with people for whom moving a icon from one place on the desktop to another results in complaints that "my Windows is broken," usually followed by, "I can't do my work until it's fixed."
I don't know what the situation is today, but in the past this functionality was definitely not available to Verizon Wireless users. VZ disabled direct connections to the USB port so the only way to transfer pictures to and from the phone was via a data connection over its network or via an SD card if the phone has that capability. In general you had to email or text the photos to yourself and pay any consequent data charges.
I'd venture to guess that it's still the case that most cellphone users never use their USB cables to transfer anything. They take pictures on their phones and use them as local wallpapers or hold them up so their friends can see them on the screen. The market for ringtones and sales of music tracks for phones is estimated to exceed $10B in the next few years. I have a technologically sophisticated teen-aged daughter who could move stuff to any from her phone over USB the same way she does with her Sansa player. Most of her friends wouldn't have the slightest idea how to accomplish any of that.
I only buy unlocked phones to avoid these problems.
I'd agree that your computer owned by you on an Internet connection you purchased and maintain is analogous to a printing press. Your blog on a computer hosted by someone else who has entered into a specific contract with yet another entity (the hosting company) doesn't qualify in my mind. If you decide that your information is sufficiently important, and perhaps sufficiently controversial, that it might raise someone's hackles, host it yourself and deal with the political and legal consequences that might follow. Placing your content on a server owned by some random guy with 72,000 other people doesn't suggest to me that you think your speech is worth protecting.
Hosting blogs is probably conceptually closer to publishing a magazine. If you write an incisive article on the misadventures of some government agency and publish it in a magazine where someone else has posted a picture on page 37 of some eight-year-old sucking a guy's dick, do you really think the authorities are going to be worrying about your First Amendment rights when they show up at the magazine publisher's offices with a warrant? Hosting providers are treated rather akin to common carriers when it comes to copyright infringement (via the DMCA take-down procedure). But I don't think that should absolve the provider of all editorial responsibilities any more than it would if he or she were publishing a print magazine with multiple independent contributors.
Mr. Blog-Hoster seemed to think he could just sit back and collect his AdSense revenues without taking any responsibility for the material he was hosting. I think that's nuts.
As a former university professor, I find the limits on access to scientific papers appalling. Companies like Wiley, Elsevier and the like have had nice little fiefdoms that depended on the captive audience of academic libraries. I don't begrudge them charging for articles written in the past year or two, but I find it offensive that I cannot read a decade-old piece of academic research today for free. I might pay a dollar or two for an article of that age, but $20 or more? Nope, not gonna happen.
Let's not forget the huge indirect subsidies these publishers get from places like the National Institutes of Health; research proposals usually include a line-item to cover the costs of publication.
So what happens when a student with little or no funds wants to read an article? Is he or she going to pony up the $20+ fee? No, they're going take the library's copy to the photocopier. How does the publisher earn anything on that transaction?
I once let my Economist subscription lapse for a couple of weeks. When I renewed they sent me the missing issues automatically.
There are just too many free alternatives at the moment to make a paywall work for a newspaper. Someday the New York Times might make me start paying, and there's some chance I might if it's cheap enough. If it's anything like Murdoch's $3/week, I'll just send some more money to NPR and keep my Economist subscription.
After reading that thread at webhostingtalk, I'm going to guess the Patriot Act was invoked. It has that air about it. If it was child porn, would Burst be facing an obvious gag order from the government?
The server operator sounds incredibly clueless. Did it never cross his mind that offering free hosting to 70,000+ accounts might put him in the position of hosting material that might violate some laws?
This doesn't sound like the usual copyright-infringement case; it looks much more serious than that.
Just this week I've seen two spams that appear to have come from real accounts at AOL and Hotmail. I know for a fact that the first was a real account since it belonged to someone subscribed to a limited-membership listserver I manage. The second was from an account I knew nothing about, but it was essentially identical in content to the first. Both came from the mail providers' own SMTP servers. They also both have a X- header identifying the originator of the message, and both came from an IP in 115.48/12 described as "China Unicom Henan province network."
I don't really care about OpenSolaris, but I have been a happy user of VirtualBox since before it was acquired by Sun. Sun developed some nice, but proprietary, tweaks to VirtualBox in areas like graphics drivers. I do see development continuing as I get prompted to upgrade fairly regularly, but I've been nervous that VirtualBox will also eventually be treated as roadkill by Oracle. Obviously there will always be a free implementation since the "open-source edition" is GPL-licensed.
I can understand Oracle's lack of interest in OpenSolaris since they've supported Linux for a long time now. (Hell, they even compete directly against RedHat with their Oracle Enterprise Linux distribution.) I do wonder, though, whether they'll stay committed to VirtualBox down the road.
Obviously Deval Patrick supports this obvious attempt at pandering to the Scott Brown brigade since he signed the bill into law. What about our other three candidates for Governor? There's no mention of this law on any of their websites.
I'm amused this appears on the same page as the discussion about liability for breaches. We all know that enforcing large, public, and expensive fines is the only solution that corporations will pay any attention to. In fact, why not make CIOs (and CEOs?) personally liable.
I was talking about foreign press coverage in the mainstream media, not whether the issue is being discussed on Australian advocacy sites. My point was simply that the Internet censorship issue never appeared in foreign reports about the transition from Rudd to Gillard.
You may be surprised to hear that I saw nothing about the Internet filtering issues when Gillard took office. At the time I was in Europe and reading the International Herald-Tribune and the Financial Times. The conflicts over carbon trading and the mining tax were the ones described in the foreign media as the keys to Rudd's departure. Here's a representative sample from another, well-respected paper: http://www.csmonitor.com/World/Asia-Pacific/2010/0624/Julia-Gillard-takes-helm-in-Australia-after-Kevin-Rudd-ouster
I read some of the follow-on stories in the days after; Internet filtering was never mentioned.
I hate to break it to you, but Immigration and Customs Enforcement is part of the Department of Homeland Security.
I don't know how he can write, "None of the companies involved would disclose details of the suits...," with a straight face. These are lawsuits, after all, so presumably they've been filed with a court and become public information. Unfortunately in the world of tech journalism, "research" consists of asking consultants like The Yankee Group what they think rather than actually, you know, reading documents.
Got any citations? Back in the 1980's I wrote some papers on the socio-economic and political factors affecting cross-national variations in television and radio penetration rates. I've considered returning to this question with regard to the Internet. So I'd love to see some decent academic research on this question if it exists.
But I don't see the meagre profits from the app store, or from video rentals, or similar low-margin operations (possibly, in fact probably, including their music sales) as being reasonable; they strike me as ignorant of how businesses think. Yes, it's all about money with Apple (and every other successful corporation), but for Apple, the money is in the device sales.
Might that be a bit short-sighted over the longer term? Apple might primarily a hardware company today with a smaller content marketing arm, but ten years from now, the profitability of the two might well be reversed. Apple won't be able to dominate the mobile devices market in the future as it has in the past, but it could still be a major player in the ever-expanding business of content distribution. I also wouldn't be surprised to see the often-mooted Apple/Disney deal happen in the years ahead. Disney has already signed on for AppleTV. I bet those programs won't be available via h.264 over Flash.
I thought it was "branding eye candy" when I first used Office 2007 myself. I looked and looked for the equivalent of the File menu, then clicked the button when I couldn't think of any other solution.
If you host your own mail server so that the mail is never stored on the Internet, then by definition that server has to have presence on the Internet, and again, the risk is still there.
You made a lot of good points, but I have to take issue with this one. There's no reason why the server that stores the mail needs to be visible on the Internet. Usually I have three machines involved -- one running a secure store-and-forward SMTP listener, one dedicated to scanning for spam and viruses, and a third where the mailboxes actually reside. Only the first of these has any exposure whatsoever, and depending on volumes the scanning and storage functions might be collapsed onto a single, internal server.
Some of my clients use third-parties like Postini, but the clients' firewalls and servers are configured to accept inbound SMTP transfers only from Postini's IP block.
Webmail creates the potential for intrusions, of course, though again we separate the server with the external client software from the machines where the mail is actually stored.
I did a bit of digging, and all the data on host counts appear to be compiled from the ISC Domain Survey. According to the summary on that page, "The Domain Survey attempts to discover every host on the Internet by doing a complete search of the allocated address space and following links to domain names." This would seem to exclude hosts without reverse-DNS records, but I'd need to read the complete study methodology before I could comment intelligently.
I also looked to see if there were easily-available figures on the number of IP addresses allocated by country but couldn't find any.
Regardless of the method for counting hosts, it still seems quite likely that US hosts make up considerably more than 13% of all hosts worldwide.