His main complaint is really an IPv6 conversion problem. Facebook probably works on Vista from network connections that can't get an IPv6 connection at all, because the client presumably tries IPv4. But he was apparently testing from some connection that could pass IPv6 packets to Facebook, and Vista properly tried to use IPv6.
It's a legitimate complaint if this problem isn't properly reported to the user.
We're going to be seeing more of this. Rollout of IPv6 to consumers only started a few weeks ago.
JSON is almost exactly equivalent to LISP S-expressions. Unfortunately, JSON has major security problems due to a Javascript design error. In LISP, there's the "reader", which takes in a string, parses it, and generates a tree structure (or graph; there's a way to express cross-links), and just returns the data structure. Then lISP has "eval", which takes
a data structure created by the reader and runs it.
Javascript combines both functions into one, called "eval", which takes a string, parses it, and runs it. Giant security hole. There are attempts to patch around this, but it's ugly.
Roland is off in bogosity land, as usual.
The wingspan of a barn swallow is about 0.3m. The wingspan of an A-4 Skyhawk is 8.1m,
which is 27x larger. So, scaled for size, an A-4 Skyhawk actually has about 4x the roll rate of a sparrow.
Historically, aircraft that looked or worked like birds have been spectacularly unsuccessful. Little ornithopter UAVs do work, but the ornithopter concept does not
scale up well.
Zombietime's
Mohammed Image Archive has a collection of most of the available images of Mohammed. The oldest dates from 67 years after his death, and is from a coin in the British Museum.
You don't have to make your own key. Diebold will sell you one. "Replacement Access Keys", part number GS-567311-1000, $5.90/set of 2. Order by phone, 1-800-769-3246. Operators are standing by.
It's possible to filter out the bottom-feeders, as we do at SiteTruth. We're looking at this mostly
from the user side. But there are also serious complaints about "domaining" from the advertiser side.
Clicks on "typosquatting" sites don't lead to many sales. Basically, they're targeting users who click on random stuff. That doesn't mean those users actually buy based on their mis-aimed clicks. More likely, some real company that advertised via Google AdWords is getting money sucked out of their ad budget without much return. The analytics people are skeptical of the claims of domainers.
The Direct Marketing Association has a white paper for advertisers which recommends that advertisers filter those sites out of their campaigns. "The traffic produced by sites utilizing the practices described above is almost always
absolutely worthless. To ensure contextual advertising effectiveness, advertisers should
eliminate these sites from their campaigns." Google, however, makes this difficult, because Google doesn't tell the advertiser where their ads are running, and requires excluding each individual domainer site by name, from Google's user interface. There's no "disable all bottom feeders" option. This is a problem.
The DMA's white paper suggests ways an advertiser can defend their ad costs against domainers, automatically accumulating a list of domainers feeding them clicks, discovering which sites generate poor returns, and excluding them. But with clicks coming in randomly from hundreds of thousands (maybe millions) of constantly changing bottom-feeder sites, blacklisting the bogus sites is like spam filtering by source address - it's a losing battle.
The advertiser community is getting wise to this. We may see some pushback from that side.
Child molestation is mostly by friends and family, plus the occasional priest. 80% friends and family, 20% strangers. So, kids, get out of the house, stay away
from churches, and head for the mall.
Re:Nvida and AMD were already working on Physics
on
NVIDIA To Buy AGEIA
·
· Score: 1
Can I then grab back that batch of models I sent off for physics and then send the updated coordinates to the players computers? Or is it just part of the graphics rendering pipeline?
It's mostly part of the graphics rendering pipeline. Ageia's "physics engine" is mostly used for particle effects (smoke, fire, rain, snowflakes, etc.) which don't affect the gameplay at all. There were attempts to use it for actual game physics, but the performance was no better than doing that on the main CPU. For particle physics, massive uncoordinated parallelism (the easy case) is possible; the particles don't affect other objects or even each other.
The Republican candidates left are the dud ones. McCain is too old (72), and Huckabee is too weird. Neither has any management experience. Romney is the best of the lot, and would probably do OK.
Both Edwards and Giuliani looked better than the three remaining candidates.
None of these men came out of the Bush administration. Coming from "Team Bush" is a political kiss of death at this point.
I notice that the parent post reached 5, and then was marked "overrated" hours after posting to bring it down to 3. This happens consistently with articles critical of Apple, and always hours after the original posting. It's almost as if Apple has a system which watches for criticism of Apple and then sends someone in to mute the criticism.
I can't see a Google/Yahoo alliance getting antitrust approval. Even with the out-to-lunch antitrust division at Justice under the Bush administration. If you're #1 with over half the market and want to buy #2, it's pretty clear. A merger between #2 and #3 which leaves the merged entity with less than #1 is likely to be approved, though.
Voting machines should have to meet the Nevada Gaming Commission Standards for Gaming Devices. Nevada has tough tamper-resistance standards (Immune to static shocks, 27KV sparks, 600V on the power input, and rapid turn on/off; must resist forced illegal entry, locked covers over circuit boards and program media), logging standards (counters that cannot be reset, non-erasable logs of program changes), and auditing standards ("Provide, as a minimum, a two-stage mechanism for validating all program components on
demand via a communication port and protocol approved by the chairman.") There's no question those standards can be met; hundreds of thousands of slot machines are running right now in compliance with them. Those standards have been developed during decades of struggles against organized crime, employee theft, tax fraud, and attacks on slot machines, so they have serious real-world credibility.
Use a minimal, published operating system, like Minix. Linux is too big to audit and changes too much.
Use a paper trail within the machine, one that generates a printed copy of the voter's selections behind a window, along with a bar code representing the voter's choices.
For recounts, run the paper log through a bar code scanner for a quick check, and if necessary, manually check votes against bar codes.
Install two printers, and switch between them randomly, so that the paper trail doesn't provide enough information to tell who voted for whom. Use a printer that doesn't need ink or ribbon and makes a permanent record, like the old "silver printers" used in adding machines. Don't use a thermal printer; the print isn't permanent.
"and then the ability to seamlessly activate via iTunes"
"Seamlessly"? You have to have a computer connected to the Internet just to activate your phone? That is so lame. There's a huge population of people, especially outside the US, who have mobile phones but not computers. I wonder what percentage of those un-activated iPhones were bought by people who didn't realize they had to mess with a PC just to turn the phone on.
And you still can't download music over the air link, can you?
It makes sense for Microsoft to acquire parts of Yahoo, but $44 billion in cash for the whole thing? That's way overpriced. Yahoo, based on their declining earnings, is worth $12-$15 billion. If that. Microsoft will have to issue stock or take on debt to buy Yahoo for cash.
I suspect the hostile offer of $44 billion is a bluff. The real idea is to pressure Yahoo's board into a friendly merger structured as a stock swap. Then Microsoft becomes a bigger company and Yahoo shareholders are issued Microsoft stock to replace their Yahoo stock.
Somebody already tried this, as I discovered during a patent search.
Other weird search ideas included adjusting search preferences based on what other applications are running. If you seach for "gold", you get different responses depending on whether you're running Everquest or Excel.
What's more likely to work is ad personalization based on your social network. If your friends bought something, then promoting it to you is a promising idea. That's been proposed as PriceKut.
There's surprisingly little in Python 3.x that's really needed, and much that isn't. The approach to parameter typing (optional and unenforced) is silly. Having it and not enforcing it is just asking for trouble.
It probably would have been more productive to standardize on 2.6 and get a formal standard out the door, instead of using the CPython implementation as the standard. With a formal standard that couldn't be casually changed, the other implementations, all of which are faster than CPython, would have a firm target to implement. Python is twenty years old, and there still aren't multiple, compatible implementations.
Python could be a much higher performance language. Take a look at the Shed Skin implementation. One guy is trying to implement a hard-code compiler for Python that does type inference to determine types at compile time whenever possible. That yields a 10x-30x speedup. If you have rackmount servers running Python, that's a big win - one rack instead of ten.
Python has some optimization gotchas that really should be fixed. The big problem is "undetectable dynamism", or "if the only tool you have is a dictionary, everything looks like a hash". It's rare to store into a local variable of a function, or modify a method of a class from the outside of the class. Most classes don't have attributes added to them after creation. Python allows all these things, which can occasionally be useful. The trouble is that it's really tough to tell at compile time if the hard cases are going to be needed, and thus code has to be pessimized for the worst case.
This could be fixed with a few restrictions:
Classes which can be dynamically modified from outside themselves should be subclasses of "dynamicobject" instead of "object". This makes everything dynamic but reduces performance. For most objects, the compiler can then find all the variables during
compilation, assign them fixed slots, and avoid having a dictionary in each object.
If an object indulges in self-modification or attribute creation, the compiler can see that at compile time and generate the slow code for the hard case. This is only needed for objects which are patched from outside themselves, something the compiler can't now detect and needs to know about.
Variables cannot change major type during execution. If a variable is initialized with an integer or float value, it cannot thereafter be changed to an object type. Shed Skin imposes this restriction, which means it doesn't have to "box" numbers in objects and can hard-compile arithmetic.
This would make it possible to boost Python performance up to the Java level, and get it within striking distance of C/C++, yet not require declarations.
Last week, the "breakups" column in the Wall Street Journal suggested that Yahoo be broken up and sell off or outsource their search operation to Microsoft. The WSJ columnist thought that YHOO shareholder value could be increased 60% that way.
I can see Microsoft wanting Yahoo's search and ad operations. But what will Microsoft do with the rest of Yahoo?
$44 billion is overpaying, too. Yahoo is profitable and made about $750 million in 2006. But they made $1.8 billion in 2005, and the trend is down, as is their stock. The number for 2007 should be around $600 million. The company is probably worth only about $9-$12 billion, figuring that a P/E of 15-20 is reasonable for a company that isn't growing.
Yes. Paragraph 3.7.9 of the registrar agreement: "Registrar shall abide by any ICANN adopted specifications or policies prohibiting or restricting warehousing of or speculation in domain names by registrars."
Actually, it's probably time to get rid of paging out to disk. At best, virtual memory creates the illusion of having twice as much memory, while seriously reducing performance. With DRAM at $64/GB, it's now more trouble than it's worth. Disk rotation isn't getting
any faster.
(QNX actually does have limited support for virtual memory; it's used for big GCC compiles and not much else.)
Most of the "ICANN accredited registrars" are fronts for domain tasting. There are only a few real registrars; the rest are dummies for picking up dropped domains. Enom has a huge number of dummy fronts - "Enom1, Inc" through "Enom469, Inc".
One step needed is for ICANN to enforce the provision of the registrar agreement which allows ICANN to prohibit registrars from owning or speculating in domains. And the provision which requires that a registrar have assurance of payment before activating a domain. With that, the end of the "grace period", and Google refusing to monetize domains for the first five days, we should see this problem decrease. The.org TLD recently got rid of their grace period, and domain transactions dropped 90%.
We're working on this from the browser end. The general idea of our SiteTruth system is to filter out the bottom-feeders.
It's the next step after ad-blocking - make the link pages, directory pages, typosquatters, and similar junk far less visible.
It's not even clear that advertisers benefit from all those junk pages. If you advertise with Google ads, and get clicks from junk pages, do they really result in sales? Or is this just a way to take money from the real advertiser and divert it to some bottom-feeder?
As someone who's done operating system internals work and has written extensively for QNX, I should comment.
Down at the bottom, microkernels are about interprocess communication.
The key problem is getting interprocess communication right. Botch that, from a performance or functionality standpoint, and your system will be terrible.
In a world where most long-running programs now have interprocess communication, it's amazing that most operating systems still do it so badly.
For interprocess communication, the application usually needs a subroutine call, and the operating system usually gives it read and write. Pipes, sockets, and System V IPC are all queues. So clunky subroutine call systems are built on top of them. Many different clunky subroutine call systems: SOAP, JSON, XMLHttpRequest, CORBA, OpenRPC, MySQL protocol, etc. Plus all Microsoft's stuff, from OLE onward. All of this is a workaround for the mess at the bottom.
The performance penalty of those kludges dwarfs that of microkernel-based interprocess communication.
I've recently been writing a web app that involves many long-running processes on a server, and I wish I had QNX messaging. I'm currently using Python, pickle, and pipes, and it is not fun. Most notably, handling all the error cases is much harder than under QNX.
Driver overhead for drivers in user-space isn't that bad. I wrote a FireWire camera driver for QNX, and when sending 640 x 480 x 24 bits x 30 FPS, it used about 3% of a Pentium III, with the uncompressed data going through QNX messaging with one frame per message. So quit worrying about copying cost.
The big problem with microkernels is that the base design is very tough. Mach is generally considered to have been botched (starting from BSD was a mistake). There have been very few good examples anyone could look at. Now that QNX source is open, developers can see how it's done. (The other big success, IBM's VM, is still proprietary.)
Incidentally, there's another key feature a microkernel needs that isn't mentioned much - the ability to load user-space applications and shared libraries during the boot process. This removes the temptation to put stuff in the kernel because it's needed during boot. For example, in QNX, there are no display drivers in the kernel, not even a text mode driver. A driver is usually in the boot image, but it runs in user space. Also, program loading ("exec") is a subroutine in a shared object, not part of the kernel. Networking, disk drivers, and such are all user-level applications but are usually part of the boot image.
Incidentally, the new head of Palm's OS development team comes from QNX, and I think we'll be seeing a more microkernel-oriented system from that direction.
Cellular air links don't have "net neutrality". The pricing for voice, web browsing, SMS, video, and non-Web data connections is totally different. That's what it's like without net neutrality.
Checking the "snopes.com" page source code, there's Javascript from "as.casalemedia.com". This is Casale Media, a Toronto-based firm. "We provide web users with relevant, personalized advertising that adds value to the browsing experience.... The network serves more than 30 billion ads every month to users in 200 countries." Something else to add to the list of advertising server domains.
I'm not getting a Zango popup, or any popups, or even a Firefox popup warning, though.
But if you have a real-world name and address, do you need to associate it with a business, even if a site is operated by an individual? Case in point: my personal web site's contact page [pineight.com]. Are personal web sites supposed to be rated "do not enter"?
If it has ads, we consider it "commercial":
Rating: "Site ownership unknown or questionable."
No Location
...
Advertising
2 advertising links found.
Advertising link to "pagead2.googlesyndication.com" found on page Pin Eight: Welcome
Advertising link to "pagead2.googlesyndication.com" found on page Pin Eight: Terms of Use ...
Commercial (has advertising) site.
Slashdot Mirror
His main complaint is really an IPv6 conversion problem. Facebook probably works on Vista from network connections that can't get an IPv6 connection at all, because the client presumably tries IPv4. But he was apparently testing from some connection that could pass IPv6 packets to Facebook, and Vista properly tried to use IPv6. It's a legitimate complaint if this problem isn't properly reported to the user.
We're going to be seeing more of this. Rollout of IPv6 to consumers only started a few weeks ago.
JSON is almost exactly equivalent to LISP S-expressions. Unfortunately, JSON has major security problems due to a Javascript design error. In LISP, there's the "reader", which takes in a string, parses it, and generates a tree structure (or graph; there's a way to express cross-links), and just returns the data structure. Then lISP has "eval", which takes a data structure created by the reader and runs it.
Javascript combines both functions into one, called "eval", which takes a string, parses it, and runs it. Giant security hole. There are attempts to patch around this, but it's ugly.
Sometimes they come back.
Roland is off in bogosity land, as usual. The wingspan of a barn swallow is about 0.3m. The wingspan of an A-4 Skyhawk is 8.1m, which is 27x larger. So, scaled for size, an A-4 Skyhawk actually has about 4x the roll rate of a sparrow.
Historically, aircraft that looked or worked like birds have been spectacularly unsuccessful. Little ornithopter UAVs do work, but the ornithopter concept does not scale up well.
Zombietime's Mohammed Image Archive has a collection of most of the available images of Mohammed. The oldest dates from 67 years after his death, and is from a coin in the British Museum.
The site also has an archive of their incoming hate mail on this subject, some of which is quite funny.
You don't have to make your own key. Diebold will sell you one. "Replacement Access Keys", part number GS-567311-1000, $5.90/set of 2. Order by phone, 1-800-769-3246. Operators are standing by.
It's possible to filter out the bottom-feeders, as we do at SiteTruth. We're looking at this mostly from the user side. But there are also serious complaints about "domaining" from the advertiser side.
Clicks on "typosquatting" sites don't lead to many sales. Basically, they're targeting users who click on random stuff. That doesn't mean those users actually buy based on their mis-aimed clicks. More likely, some real company that advertised via Google AdWords is getting money sucked out of their ad budget without much return. The analytics people are skeptical of the claims of domainers.
The Direct Marketing Association has a white paper for advertisers which recommends that advertisers filter those sites out of their campaigns. "The traffic produced by sites utilizing the practices described above is almost always absolutely worthless. To ensure contextual advertising effectiveness, advertisers should eliminate these sites from their campaigns." Google, however, makes this difficult, because Google doesn't tell the advertiser where their ads are running, and requires excluding each individual domainer site by name, from Google's user interface. There's no "disable all bottom feeders" option. This is a problem.
The DMA's white paper suggests ways an advertiser can defend their ad costs against domainers, automatically accumulating a list of domainers feeding them clicks, discovering which sites generate poor returns, and excluding them. But with clicks coming in randomly from hundreds of thousands (maybe millions) of constantly changing bottom-feeder sites, blacklisting the bogus sites is like spam filtering by source address - it's a losing battle.
The advertiser community is getting wise to this. We may see some pushback from that side.
Child molestation is mostly by friends and family, plus the occasional priest. 80% friends and family, 20% strangers. So, kids, get out of the house, stay away from churches, and head for the mall.
Can I then grab back that batch of models I sent off for physics and then send the updated coordinates to the players computers? Or is it just part of the graphics rendering pipeline?
It's mostly part of the graphics rendering pipeline. Ageia's "physics engine" is mostly used for particle effects (smoke, fire, rain, snowflakes, etc.) which don't affect the gameplay at all. There were attempts to use it for actual game physics, but the performance was no better than doing that on the main CPU. For particle physics, massive uncoordinated parallelism (the easy case) is possible; the particles don't affect other objects or even each other.
The Republican candidates left are the dud ones. McCain is too old (72), and Huckabee is too weird. Neither has any management experience. Romney is the best of the lot, and would probably do OK.
Both Edwards and Giuliani looked better than the three remaining candidates.
None of these men came out of the Bush administration. Coming from "Team Bush" is a political kiss of death at this point.
I notice that the parent post reached 5, and then was marked "overrated" hours after posting to bring it down to 3. This happens consistently with articles critical of Apple, and always hours after the original posting. It's almost as if Apple has a system which watches for criticism of Apple and then sends someone in to mute the criticism.
I can't see a Google/Yahoo alliance getting antitrust approval. Even with the out-to-lunch antitrust division at Justice under the Bush administration. If you're #1 with over half the market and want to buy #2, it's pretty clear. A merger between #2 and #3 which leaves the merged entity with less than #1 is likely to be approved, though.
It's really not that hard to do this right.
This really isn't that hard.
"and then the ability to seamlessly activate via iTunes"
"Seamlessly"? You have to have a computer connected to the Internet just to activate your phone? That is so lame. There's a huge population of people, especially outside the US, who have mobile phones but not computers. I wonder what percentage of those un-activated iPhones were bought by people who didn't realize they had to mess with a PC just to turn the phone on.
And you still can't download music over the air link, can you?
It makes sense for Microsoft to acquire parts of Yahoo, but $44 billion in cash for the whole thing? That's way overpriced. Yahoo, based on their declining earnings, is worth $12-$15 billion. If that. Microsoft will have to issue stock or take on debt to buy Yahoo for cash.
I suspect the hostile offer of $44 billion is a bluff. The real idea is to pressure Yahoo's board into a friendly merger structured as a stock swap. Then Microsoft becomes a bigger company and Yahoo shareholders are issued Microsoft stock to replace their Yahoo stock.
Somebody already tried this, as I discovered during a patent search.
Other weird search ideas included adjusting search preferences based on what other applications are running. If you seach for "gold", you get different responses depending on whether you're running Everquest or Excel.
What's more likely to work is ad personalization based on your social network. If your friends bought something, then promoting it to you is a promising idea. That's been proposed as PriceKut.
There's surprisingly little in Python 3.x that's really needed, and much that isn't. The approach to parameter typing (optional and unenforced) is silly. Having it and not enforcing it is just asking for trouble.
It probably would have been more productive to standardize on 2.6 and get a formal standard out the door, instead of using the CPython implementation as the standard. With a formal standard that couldn't be casually changed, the other implementations, all of which are faster than CPython, would have a firm target to implement. Python is twenty years old, and there still aren't multiple, compatible implementations.
Python could be a much higher performance language. Take a look at the Shed Skin implementation. One guy is trying to implement a hard-code compiler for Python that does type inference to determine types at compile time whenever possible. That yields a 10x-30x speedup. If you have rackmount servers running Python, that's a big win - one rack instead of ten.
Python has some optimization gotchas that really should be fixed. The big problem is "undetectable dynamism", or "if the only tool you have is a dictionary, everything looks like a hash". It's rare to store into a local variable of a function, or modify a method of a class from the outside of the class. Most classes don't have attributes added to them after creation. Python allows all these things, which can occasionally be useful. The trouble is that it's really tough to tell at compile time if the hard cases are going to be needed, and thus code has to be pessimized for the worst case.
This could be fixed with a few restrictions:
- Classes which can be dynamically modified from outside themselves should be subclasses of "dynamicobject" instead of "object". This makes everything dynamic but reduces performance. For most objects, the compiler can then find all the variables during
compilation, assign them fixed slots, and avoid having a dictionary in each object.
If an object indulges in self-modification or attribute creation, the compiler can see that at compile time and generate the slow code for the hard case. This is only needed for objects which are patched from outside themselves, something the compiler can't now detect and needs to know about.
- Variables cannot change major type during execution. If a variable is initialized with an integer or float value, it cannot thereafter be changed to an object type. Shed Skin imposes this restriction, which means it doesn't have to "box" numbers in objects and can hard-compile arithmetic.
This would make it possible to boost Python performance up to the Java level, and get it within striking distance of C/C++, yet not require declarations.Last week, the "breakups" column in the Wall Street Journal suggested that Yahoo be broken up and sell off or outsource their search operation to Microsoft. The WSJ columnist thought that YHOO shareholder value could be increased 60% that way. I can see Microsoft wanting Yahoo's search and ad operations. But what will Microsoft do with the rest of Yahoo?
$44 billion is overpaying, too. Yahoo is profitable and made about $750 million in 2006. But they made $1.8 billion in 2005, and the trend is down, as is their stock. The number for 2007 should be around $600 million. The company is probably worth only about $9-$12 billion, figuring that a P/E of 15-20 is reasonable for a company that isn't growing.
Does that provision really already exist?
Yes. Paragraph 3.7.9 of the registrar agreement: "Registrar shall abide by any ICANN adopted specifications or policies prohibiting or restricting warehousing of or speculation in domain names by registrars."
Actually, it's probably time to get rid of paging out to disk. At best, virtual memory creates the illusion of having twice as much memory, while seriously reducing performance. With DRAM at $64/GB, it's now more trouble than it's worth. Disk rotation isn't getting any faster.
(QNX actually does have limited support for virtual memory; it's used for big GCC compiles and not much else.)
Most of the "ICANN accredited registrars" are fronts for domain tasting. There are only a few real registrars; the rest are dummies for picking up dropped domains. Enom has a huge number of dummy fronts - "Enom1, Inc" through "Enom469, Inc".
One step needed is for ICANN to enforce the provision of the registrar agreement which allows ICANN to prohibit registrars from owning or speculating in domains. And the provision which requires that a registrar have assurance of payment before activating a domain. With that, the end of the "grace period", and Google refusing to monetize domains for the first five days, we should see this problem decrease. The .org TLD recently got rid of their grace period, and domain transactions dropped 90%.
We're working on this from the browser end. The general idea of our SiteTruth system is to filter out the bottom-feeders. It's the next step after ad-blocking - make the link pages, directory pages, typosquatters, and similar junk far less visible.
It's not even clear that advertisers benefit from all those junk pages. If you advertise with Google ads, and get clicks from junk pages, do they really result in sales? Or is this just a way to take money from the real advertiser and divert it to some bottom-feeder?
As someone who's done operating system internals work and has written extensively for QNX, I should comment.
Down at the bottom, microkernels are about interprocess communication. The key problem is getting interprocess communication right. Botch that, from a performance or functionality standpoint, and your system will be terrible. In a world where most long-running programs now have interprocess communication, it's amazing that most operating systems still do it so badly.
For interprocess communication, the application usually needs a subroutine call, and the operating system usually gives it read and write. Pipes, sockets, and System V IPC are all queues. So clunky subroutine call systems are built on top of them. Many different clunky subroutine call systems: SOAP, JSON, XMLHttpRequest, CORBA, OpenRPC, MySQL protocol, etc. Plus all Microsoft's stuff, from OLE onward. All of this is a workaround for the mess at the bottom. The performance penalty of those kludges dwarfs that of microkernel-based interprocess communication.
I've recently been writing a web app that involves many long-running processes on a server, and I wish I had QNX messaging. I'm currently using Python, pickle, and pipes, and it is not fun. Most notably, handling all the error cases is much harder than under QNX.
Driver overhead for drivers in user-space isn't that bad. I wrote a FireWire camera driver for QNX, and when sending 640 x 480 x 24 bits x 30 FPS, it used about 3% of a Pentium III, with the uncompressed data going through QNX messaging with one frame per message. So quit worrying about copying cost.
The big problem with microkernels is that the base design is very tough. Mach is generally considered to have been botched (starting from BSD was a mistake). There have been very few good examples anyone could look at. Now that QNX source is open, developers can see how it's done. (The other big success, IBM's VM, is still proprietary.)
Incidentally, there's another key feature a microkernel needs that isn't mentioned much - the ability to load user-space applications and shared libraries during the boot process. This removes the temptation to put stuff in the kernel because it's needed during boot. For example, in QNX, there are no display drivers in the kernel, not even a text mode driver. A driver is usually in the boot image, but it runs in user space. Also, program loading ("exec") is a subroutine in a shared object, not part of the kernel. Networking, disk drivers, and such are all user-level applications but are usually part of the boot image.
Incidentally, the new head of Palm's OS development team comes from QNX, and I think we'll be seeing a more microkernel-oriented system from that direction.
U2's good stuff would be public domain by now if we had reasonable copyright lengths, like we used to.
Cellular air links don't have "net neutrality". The pricing for voice, web browsing, SMS, video, and non-Web data connections is totally different. That's what it's like without net neutrality.
Checking the "snopes.com" page source code, there's Javascript from "as.casalemedia.com". This is Casale Media, a Toronto-based firm. "We provide web users with relevant, personalized advertising that adds value to the browsing experience. ... The network serves more than 30 billion ads every month to users in 200 countries." Something else to add to the list of advertising server domains.
I'm not getting a Zango popup, or any popups, or even a Firefox popup warning, though.
But if you have a real-world name and address, do you need to associate it with a business, even if a site is operated by an individual? Case in point: my personal web site's contact page [pineight.com]. Are personal web sites supposed to be rated "do not enter"?
If it has ads, we consider it "commercial":
Rating: "Site ownership unknown or questionable."
...
No Location
...
Advertising
2 advertising links found.
Advertising link to "pagead2.googlesyndication.com" found on page Pin Eight: Welcome
Advertising link to "pagead2.googlesyndication.com" found on page Pin Eight: Terms of Use
Commercial (has advertising) site.