Windows is an especially bad culprit in these cases. It caches client-side DNS lookups unless you reboot or run "ipconfig/flushdns" from the terminal. It always drives me crazy when I'm mucking with a DNS server in a client's office. The Windows machines refuse to acknowledge the changes I've made unless forced to do so.
Wow, so Rick Boucher thinks net neutrality is a good thing. This is hardly news.
Then we discover Boucher criticized Comcast in an interview initiated at the behest of a grad student who freelances for CNET. It's not like Boucher introduced legislation or even gave a floor speech in the House. As for the article's author, his journalistic credentials came immediately into question for me when he wrote "Until last month, the opponents of Net neutrality were doing just great. The issue, which had become one of national importance in 2006, had shrunk to a mere footnote in the annals of tech policy history."
Anyone who thinks that net neutrality became an issue of "national importance" in 2006 isn't living in the same America I do. Anyone who thinks the recent events mentioned in this article will elevate net neutrality into an issue of national importance in 2007 or 2008 isn't living in the same America I do.
I'm not saying it's not important, but there's no way arcane issues like these will ever be front-burner issues in the political world. They're just not enough voters who care about these issues to make them politically relevant.
I thought the 386 and successor architectures were fundamentally different from the original 8086 and 80286 architectures, particularly with regard to support for multitasking and memory protection. The transition to the 386 was pretty apparent to me. I used Desqview/386 to run more than one DOS program at a time in parallel, a feat not possible on any of the earlier x86 chips.
I admit to being a total novice in these areas, but intuitively I imagined that a hypervisor was nothing more than a stripped-down multitasking OS which runs the VMs as separate processes, presents virtual hardware interfaces to the VMs, and manages hardware contention among them. Linux seems to do a pretty good job of preventing processes from overstepping their bounds and stepping on other processes. Isn't a hypervisor doing essentially the same thing Linux is doing, just at a higher level of abstraction?
Regardless of Theo's language or his comments about the quality of OS engineers, I thought this argument was more compelling:
x86 virtualization is about basically placing another nearly full kernel, full of new bugs, on top of a nasty x86 architecture which barely has correct page protection.
Just how nasty is the x86 architecture as a platform for virtualization? Certainly the x86 wasn't originally designed to be an S/370 in a smaller box. I read Theo's post as raising the question of how exposed the hypervisor is to attack given the allegedly rather dicey hardware platform on which these VM programs run. He seems to suggest that virtualization might make denial-of-service attacks against the host more possible, with the consequence that a successful attack could take down all the VMs at once. But I'm not well-versed in any of these matters, so I'm relying on the collective "wisdom" of Slashdot to learn more.
Cable operators do not supply broadcast over the air (OTA) programming.
No, but they redistribute it. Connecting the coax lead directly into my HDTV brought me the ATSC signals from all the local broadcast stations (Comcast, Boston suburbs). If I could only get On-Demand programming this way, I'd dump the STB tomorrow. I also see better-quality SD reception over the coax connection than I do taking composite out of the cable box into one of the set's composite inputs.
I continue to be puzzled as to why none of the major distros have adopted KDE as the standard desktop. I hate how GNOME tries to hide things from the user; in some ways, it's worse than Microsoft Windows in that department. As you say, SuSE was the only major distro to support KDE, but the Novell acquisition put an end to that. What is it about GNOME that makes it so compelling to distro manufacturers?
I just bought an HDTV about a week ago and experimented with HD reception via an antenna and via direct cable input (from Comcast). My house happens to have an obsolete UHF yagi on the chimney so the quality of reception over the air using the antenna was quite good. When I connected the cable directly to the TV's coaxial connector, I got the same program quality but more digital channels since there were a couple of distant PBS stations that are unavailable via broadcast. I was actually quite surprised how simple it was to add the channels using ATSC; I just connected the cable and told the TV to scan the available bandwidth.
There's no mystery to finding out which patents Microsoft holds. Just go here You'll need some spare time to do this, though, as there are over 7,000 patents on the list.
Sometime in the past decade if I recall. I came across it while browsing the 7,000+ patents Microsoft has received. I skimmed the first couple hundred again but didn't see it. If you're really interested, you can start here.
Let me guess that most of those other buyers you know are fairly hardcore gamers.
For the rest of us, backwards-compatibility is a must. We still play PS/1 games on our PS/2 and will continue to play both sets of games on a PS/3 if we ever buy one.
If backwards-compatibility is implemented in the PS/3 the same way it is in the PS/2 (by including the PS/1's chipset on the PS/2 motherboard), how much can this cost? I'd guess a few dollars per device at most in quantity. And, if backwards-compatibility is implemented in software, then it costs essentially nothing.
If anything, this seems to be a way of making a lot more profit for Sony. Many people would prefer having just one PS device (especially in space-challenged Japanese apartments), so I'm guessing a lot of people will pay the premium for backwards-compatibility.
Most people who play videogames are not "gamers." For them, a single device with backwards-compatibility is a clear winner.
The various posters here who say "show us the patents, we'll code around them," obviously don't understand the situation as well as you do.
Let's take a hypothetical (and false) case as an example. Suppose the original coders of the NCSA Mosaic browser (the predecessor of Netscape and essentially all graphical web browsers) decided to patent the back and forward navigation arrows as a "device to permit navigation of network-delivered content" or some such patentese. The only way to avoid infringing this patent is not to use navigation arrows, not to write a different block of code to implement navigation arrows. It's like the trash can issue mentioned in another posting in this thread. Patents don't cover the code itself but the process the code implements. Even if Microsoft showed us the patents, it could be quite difficult to avoid infringing some of these.
As a real example, take the case of the patent Microsoft holds on a procedure to update Powerpoint shows online. (I read this a while ago and don't want to spend the time to look it up at the USPTO again.) The patent covers a procedure whereby an application checks online to see whether there's a newer version of the file it is about to open. If so, it automatically downloads the update and opens it instead. Obviously this was targeted at people making presentations so that the staff back in the home office could make last-minute corrections, and the presenter would never need to download the updated file herself. The patents for processes like these often contain sweeping declarations of what's covered in the patent, so even updating systems that don't precisely match the specific one described might be deemed infringing. Suppose, for instance, that Azureus checked to see if a new Java version were available when launched and downloaded and installed the new version automatically. Infringement?
The French court required Acer to pay the plaintiff at the retail cost of each of the items bundled, not the prices that Acer itself paid. See http://www.heise.de/english/newsticker/news/96581. Acer originally offered a refund of 30 euros.
For instance, the plaintiff was reimbursed 135.20 euros for Windows XP Home. Every study of OEM pricing I've seen indicates that no OEM pays anything close to that amount.
I stopped reading TFA as soon as the author quoted these figures in defense of the notion that the bundled software constituted nearly half the retail price of a computer. If that's the level of sophistication of the remainder of the article, it's not worth my time.
A lot of spam identifies itself with The Bat in the (optional) X-Mailer header. So few real people use The Bat, and so many spammers do, that I routinely add SpamAssassin points to any messages listing The Bat in X-Mailer.
Not to mention that, on Bittorrent, most clients tell you how many copies of a file have been seeded. (Azureus logs other things that might have applied to her defense, too, like total traffic volumes.) I think she was using Kazaa, though. I have no idea what information the Kazaa client collects, but apparently neither the defendant nor her attorney thought to look there for evidence in support of her defense.
Try VirtualBox which lets you run most any Windows flavor, plus a number of Unix/Linux OS's like FreeBSD, on top of Linux, Windows, or (beta) OS X. VirtualBox has rpm and deb packages so installation on the mainstream Linux desktops is a breeze.
Just the thing for those times I need to look at a web site with Internet Explorer.
2) Why are you using MSSQL where there are perfectly fine, open-source alternatives like PostgreSQL? And, if you really must use a commercial database, I've found PHP's support for Oracle is pretty easy to set up as well. Why would you expect it to be easy to work with MSSQL, of all things, from an open-source LAMP platform? Do you think Microsoft has any interest in supporting this configuration?
So why doesn't Dell pre-load Open Office or Star Office on their computers instead of or with MS Office trials?
Maybe because Dell collects a percentage of the revenue from every MS Office installation they sell? What would they collect if they installed OpenOffice?
I've advocated in past postings here that OEMs include Firefox, Thunderbird, OpenOffice, and maybe even the GIMP on all new Windows machines. But the financial incentives all run the other way, especially in the case of OO.
Next week's class will cover why we don't see installations of CCCP on every Windows machine.
I've got a number of high-quality 720p anime fansubs in H.264 with AAC audio. A 25-minute episode fits in about 350 MB, so a two-hour movie could easily fit on a standard DVD with loads of room to spare. I think the bitrates are pretty high, too. I've converted some of them to XviD+MP3 in AVI, and they easily deliver 1200 b/s rates in the second XviD pass. They look quite pristine even when I watch them on my 1280x1024 monitor at close range.
This is the second time the FCC has undertaken rule-makings to expand access to what was once called the "Instructional Television Fixed Service" and is now called the "Educational Broadband Service."
In a nutshell, 32, 6-MHz channels (the same size as OTA television channels) were set aside in the 2.5 GHz band for nonprofit, instructional organizations. Many of these channels are used by colleges and universities and some public and private school districts, to distribute programming from a central location to receivers scattered across the reception area. Catholic school programmers have been big users of this spectrum to distribute programming from diocesan centers to parochial and private Catholic schools. In some markets many of these channels lie fallow; in other markets, mostly the large citieis, all or most of them are in use.
In the mid-80's the FCC was pressured by pay-TV operators who distributed services like HBO over microwave to enable them to expand their operations into the ITFS bands. Unused spectrum on eight of these channels was licensed for commercial use as the "Multichannel Microwave Distribution Service," and potential pay-TV providers were encouraged to work with any existing instructional licensees to share their bandwidth. I worked on license applications for this service for a number of clients as well as submitting applications myself in partnership with some private investors. The MMDS licenses were awarded by lottery, and our partnership actually won a "construction permit" for Fort Collins, Colorado. Nevertheless we never built anything with this permit, nor did most of the other entrepreneurs looking to develop services in these bands. All the financial models were premised on rolling out multichannel pay-TV services in markets where cable had yet to be built. The rapid expansion of cable into the major markets (from which they had previously been banned by FCC regulation) destroyed the market for microwave-based pay-TV services. The regulatory process took so long that by the time the FCC had changed the rules, the economic rationale for the service had begun to wither away.
Considering that the ITFS dates back to the 1970's, I'd say Sprint and Clearwire would have had to be incredibly blind not to know that these channels were licensed to instructional television services.
My guess is that the FCC again wants to see some kind of negotiation between commercial and educational users of this block of spectrum in hopes of utilizing it more fully and subsidizing educational services.
A commentator on the Diane Rehm show talked about China's strategic interest in Burma as well as China's desire for Burma's energy and other resources. Having Burma as a client gives landlocked areas in southern China direct access to the Indian Ocean. Apparently the Chinese have been investing substantially in the development of Burma's infrastructure to facilitate this access. A Burmese client also gives China more leverage in Southeast Asia where it hopes to expand its sphere of influence.
Even worse is their hijacking of the ghost of Paul Douglas, one of the most famous and influential progressive politicians of the postwar period. The "about us" statement says the purpose of the group is to be a watchdog for the Code of Ethics authored by Douglas.
I bet he rolls over in his grave every time someone reads that web page.
If you read their history page, you'll see that their primary targets are Democratic politicians, labor unions, and progressive organizations.
Anyone claiming that the NLPC is "a US-based copyright watchdog," as stated on the Slashdot front page, hasn't done his or her homework.
Slashdot Mirror
Windows is an especially bad culprit in these cases. It caches client-side DNS lookups unless you reboot or run "ipconfig /flushdns" from the terminal. It always drives me crazy when I'm mucking with a DNS server in a client's office. The Windows machines refuse to acknowledge the changes I've made unless forced to do so.
Wow, so Rick Boucher thinks net neutrality is a good thing. This is hardly news.
Then we discover Boucher criticized Comcast in an interview initiated at the behest of a grad student who freelances for CNET. It's not like Boucher introduced legislation or even gave a floor speech in the House. As for the article's author, his journalistic credentials came immediately into question for me when he wrote "Until last month, the opponents of Net neutrality were doing just great. The issue, which had become one of national importance in 2006, had shrunk to a mere footnote in the annals of tech policy history."
Anyone who thinks that net neutrality became an issue of "national importance" in 2006 isn't living in the same America I do. Anyone who thinks the recent events mentioned in this article will elevate net neutrality into an issue of national importance in 2007 or 2008 isn't living in the same America I do.
I'm not saying it's not important, but there's no way arcane issues like these will ever be front-burner issues in the political world. They're just not enough voters who care about these issues to make them politically relevant.
I thought the 386 and successor architectures were fundamentally different from the original 8086 and 80286 architectures, particularly with regard to support for multitasking and memory protection. The transition to the 386 was pretty apparent to me. I used Desqview/386 to run more than one DOS program at a time in parallel, a feat not possible on any of the earlier x86 chips.
I admit to being a total novice in these areas, but intuitively I imagined that a hypervisor was nothing more than a stripped-down multitasking OS which runs the VMs as separate processes, presents virtual hardware interfaces to the VMs, and manages hardware contention among them. Linux seems to do a pretty good job of preventing processes from overstepping their bounds and stepping on other processes. Isn't a hypervisor doing essentially the same thing Linux is doing, just at a higher level of abstraction?
Regardless of Theo's language or his comments about the quality of OS engineers, I thought this argument was more compelling:
x86 virtualization is about basically placing another nearly full kernel, full of new bugs, on top of a nasty x86 architecture which barely has correct page protection.
Just how nasty is the x86 architecture as a platform for virtualization? Certainly the x86 wasn't originally designed to be an S/370 in a smaller box. I read Theo's post as raising the question of how exposed the hypervisor is to attack given the allegedly rather dicey hardware platform on which these VM programs run. He seems to suggest that virtualization might make denial-of-service attacks against the host more possible, with the consequence that a successful attack could take down all the VMs at once. But I'm not well-versed in any of these matters, so I'm relying on the collective "wisdom" of Slashdot to learn more.
Cable operators do not supply broadcast over the air (OTA) programming.
No, but they redistribute it. Connecting the coax lead directly into my HDTV brought me the ATSC signals from all the local broadcast stations (Comcast, Boston suburbs). If I could only get On-Demand programming this way, I'd dump the STB tomorrow. I also see better-quality SD reception over the coax connection than I do taking composite out of the cable box into one of the set's composite inputs.
I continue to be puzzled as to why none of the major distros have adopted KDE as the standard desktop. I hate how GNOME tries to hide things from the user; in some ways, it's worse than Microsoft Windows in that department. As you say, SuSE was the only major distro to support KDE, but the Novell acquisition put an end to that. What is it about GNOME that makes it so compelling to distro manufacturers?
I just bought an HDTV about a week ago and experimented with HD reception via an antenna and via direct cable input (from Comcast). My house happens to have an obsolete UHF yagi on the chimney so the quality of reception over the air using the antenna was quite good. When I connected the cable directly to the TV's coaxial connector, I got the same program quality but more digital channels since there were a couple of distant PBS stations that are unavailable via broadcast. I was actually quite surprised how simple it was to add the channels using ATSC; I just connected the cable and told the TV to scan the available bandwidth.
There's no mystery to finding out which patents Microsoft holds. Just go here You'll need some spare time to do this, though, as there are over 7,000 patents on the list.
Sometime in the past decade if I recall. I came across it while browsing the 7,000+ patents Microsoft has received. I skimmed the first couple hundred again but didn't see it. If you're really interested, you can start here.
Let me guess that most of those other buyers you know are fairly hardcore gamers.
For the rest of us, backwards-compatibility is a must. We still play PS/1 games on our PS/2 and will continue to play both sets of games on a PS/3 if we ever buy one.
If backwards-compatibility is implemented in the PS/3 the same way it is in the PS/2 (by including the PS/1's chipset on the PS/2 motherboard), how much can this cost? I'd guess a few dollars per device at most in quantity. And, if backwards-compatibility is implemented in software, then it costs essentially nothing.
If anything, this seems to be a way of making a lot more profit for Sony. Many people would prefer having just one PS device (especially in space-challenged Japanese apartments), so I'm guessing a lot of people will pay the premium for backwards-compatibility.
Most people who play videogames are not "gamers." For them, a single device with backwards-compatibility is a clear winner.
The various posters here who say "show us the patents, we'll code around them," obviously don't understand the situation as well as you do.
Let's take a hypothetical (and false) case as an example. Suppose the original coders of the NCSA Mosaic browser (the predecessor of Netscape and essentially all graphical web browsers) decided to patent the back and forward navigation arrows as a "device to permit navigation of network-delivered content" or some such patentese. The only way to avoid infringing this patent is not to use navigation arrows, not to write a different block of code to implement navigation arrows. It's like the trash can issue mentioned in another posting in this thread. Patents don't cover the code itself but the process the code implements. Even if Microsoft showed us the patents, it could be quite difficult to avoid infringing some of these.
As a real example, take the case of the patent Microsoft holds on a procedure to update Powerpoint shows online. (I read this a while ago and don't want to spend the time to look it up at the USPTO again.) The patent covers a procedure whereby an application checks online to see whether there's a newer version of the file it is about to open. If so, it automatically downloads the update and opens it instead. Obviously this was targeted at people making presentations so that the staff back in the home office could make last-minute corrections, and the presenter would never need to download the updated file herself. The patents for processes like these often contain sweeping declarations of what's covered in the patent, so even updating systems that don't precisely match the specific one described might be deemed infringing. Suppose, for instance, that Azureus checked to see if a new Java version were available when launched and downloaded and installed the new version automatically. Infringement?
Well, there's always IBM.
The French court required Acer to pay the plaintiff at the retail cost of each of the items bundled, not the prices that Acer itself paid. See http://www.heise.de/english/newsticker/news/96581. Acer originally offered a refund of 30 euros.
For instance, the plaintiff was reimbursed 135.20 euros for Windows XP Home. Every study of OEM pricing I've seen indicates that no OEM pays anything close to that amount.
I stopped reading TFA as soon as the author quoted these figures in defense of the notion that the bundled software constituted nearly half the retail price of a computer. If that's the level of sophistication of the remainder of the article, it's not worth my time.
A lot of spam identifies itself with The Bat in the (optional) X-Mailer header. So few real people use The Bat, and so many spammers do, that I routinely add SpamAssassin points to any messages listing The Bat in X-Mailer.
How to install custom nVidia drivers on Fedora 7:
1) Install Fedora
2) Install the Livna rpm from rpm.livna.org
3) Run "yum install kmod-nvidia xorg-x11-drv-nvidia"
4) Enjoy!
I'm really tired of how everyone seems to think that there's only one worthwhile desktop Linux distribution these days, and that it's Ubuntu.
VirtualBox
Not to mention that, on Bittorrent, most clients tell you how many copies of a file have been seeded. (Azureus logs other things that might have applied to her defense, too, like total traffic volumes.) I think she was using Kazaa, though. I have no idea what information the Kazaa client collects, but apparently neither the defendant nor her attorney thought to look there for evidence in support of her defense.
Shoot, something went wrong with the link in the only post I haven't previewed in weeks, of course.
Here it is: VirtualBox.
And, yes, this time I previewed.
Try VirtualBox which lets you run most any Windows flavor, plus a number of Unix/Linux OS's like FreeBSD, on top of Linux, Windows, or (beta) OS X. VirtualBox has rpm and deb packages so installation on the mainstream Linux desktops is a breeze.
Just the thing for those times I need to look at a web site with Internet Explorer.
1) You could have used the ODBC functions.
2) Why are you using MSSQL where there are perfectly fine, open-source alternatives like PostgreSQL? And, if you really must use a commercial database, I've found PHP's support for Oracle is pretty easy to set up as well. Why would you expect it to be easy to work with MSSQL, of all things, from an open-source LAMP platform? Do you think Microsoft has any interest in supporting this configuration?
So why doesn't Dell pre-load Open Office or Star Office on their computers instead of or with MS Office trials?
Maybe because Dell collects a percentage of the revenue from every MS Office installation they sell? What would they collect if they installed OpenOffice?
I've advocated in past postings here that OEMs include Firefox, Thunderbird, OpenOffice, and maybe even the GIMP on all new Windows machines. But the financial incentives all run the other way, especially in the case of OO.
Next week's class will cover why we don't see installations of CCCP on every Windows machine.
I've got a number of high-quality 720p anime fansubs in H.264 with AAC audio. A 25-minute episode fits in about 350 MB, so a two-hour movie could easily fit on a standard DVD with loads of room to spare. I think the bitrates are pretty high, too. I've converted some of them to XviD+MP3 in AVI, and they easily deliver 1200 b/s rates in the second XviD pass. They look quite pristine even when I watch them on my 1280x1024 monitor at close range.
This is the second time the FCC has undertaken rule-makings to expand access to what was once called the "Instructional Television Fixed Service" and is now called the "Educational Broadband Service."
In a nutshell, 32, 6-MHz channels (the same size as OTA television channels) were set aside in the 2.5 GHz band for nonprofit, instructional organizations. Many of these channels are used by colleges and universities and some public and private school districts, to distribute programming from a central location to receivers scattered across the reception area. Catholic school programmers have been big users of this spectrum to distribute programming from diocesan centers to parochial and private Catholic schools. In some markets many of these channels lie fallow; in other markets, mostly the large citieis, all or most of them are in use.
In the mid-80's the FCC was pressured by pay-TV operators who distributed services like HBO over microwave to enable them to expand their operations into the ITFS bands. Unused spectrum on eight of these channels was licensed for commercial use as the "Multichannel Microwave Distribution Service," and potential pay-TV providers were encouraged to work with any existing instructional licensees to share their bandwidth. I worked on license applications for this service for a number of clients as well as submitting applications myself in partnership with some private investors. The MMDS licenses were awarded by lottery, and our partnership actually won a "construction permit" for Fort Collins, Colorado. Nevertheless we never built anything with this permit, nor did most of the other entrepreneurs looking to develop services in these bands. All the financial models were premised on rolling out multichannel pay-TV services in markets where cable had yet to be built. The rapid expansion of cable into the major markets (from which they had previously been banned by FCC regulation) destroyed the market for microwave-based pay-TV services. The regulatory process took so long that by the time the FCC had changed the rules, the economic rationale for the service had begun to wither away.
Considering that the ITFS dates back to the 1970's, I'd say Sprint and Clearwire would have had to be incredibly blind not to know that these channels were licensed to instructional television services.
My guess is that the FCC again wants to see some kind of negotiation between commercial and educational users of this block of spectrum in hopes of utilizing it more fully and subsidizing educational services.
A commentator on the Diane Rehm show talked about China's strategic interest in Burma as well as China's desire for Burma's energy and other resources. Having Burma as a client gives landlocked areas in southern China direct access to the Indian Ocean. Apparently the Chinese have been investing substantially in the development of Burma's infrastructure to facilitate this access. A Burmese client also gives China more leverage in Southeast Asia where it hopes to expand its sphere of influence.
Even worse is their hijacking of the ghost of Paul Douglas, one of the most famous and influential progressive politicians of the postwar period. The "about us" statement says the purpose of the group is to be a watchdog for the Code of Ethics authored by Douglas.
I bet he rolls over in his grave every time someone reads that web page.
If you read their history page, you'll see that their primary targets are Democratic politicians, labor unions, and progressive organizations.
Anyone claiming that the NLPC is "a US-based copyright watchdog," as stated on the Slashdot front page, hasn't done his or her homework.