I don't know where the GP poster is from, but in Ontario, Canada, the basic everyone-has-to-know-what's-in-it Driver's Handbook has the phrase "shift to neutral/de-clutch" in quite a number of places when describing emergency situations and some correct responses to them.
The problem is, so many people behind the wheel aren't drivers. They have no desire to be there, they don't care about knowing how to operate a car, they just want to get to where they're going. So they squeak by enough testing to get a license, never bother with Driver's Ed or any advanced driver skills course.
And then they say no-one ever told them what to do.
Which, of course, is wrong; they were told, they just weren't listening. And sure, it's hard to care about something when you don't have a need for it. But it's almost like wilful ignorance, like people who say, "Oh I'm not good with math" and never try.
The default host on my servers comes back with an empty page for all URLs. If you don't provide the Host: header, you don't get any host at all.
Originally, that was so my family couldn't stumble in to the Wrong VHost with a bad URL. (Mainly because, at the time, the Wrong VHost was the default one.) But I liked some of the side-effects, so I do it on other servers, too.
And I haven't cared about browsers who don't send a Host: header since about 2004. lynx sends a Host: header. The only time it's a bit irritating is when I'm using telnet to debug something, then I have to type an extra line.
But remember who wrote Commodore's BASIC on every machine they made: Microsoft.
The really sad part about the 64's BASIC was the PET/CBM line already had BASIC 4.0, and Commodore went back to 2.0 for the VIC-20 and 64. Which meant there was no disk drive support in BASIC, you had to remember to use file channel numbers and load a special filename to do a directory listing. DOS operations were done with funny open commands. (Back when DOS really was the operating system on the disk drive, not a product from Microsoft written for IBM. Most of the PET/CBM disk drives had two CPUs, though less RAM than the computer.)
Still, it was simple enough that you could load a "wedge" that would use the BASIC line-editor to write assembler programs. When you're that close to the silicon, you learn how computers work or you don't program them.
And Sun does, indeed, include the source for GPL software on their media.
It would be up to you to keep the patches current on the sources as well as the binaries. Sun's done their bit.
Not that I like this; it's maddening enough having to use Red Hat Network for Red Hat Enterprise patches. (It's not that I don't want to pay for the contracts, I'm more than happy to do that. It's the resulting hoops that have to be jumped through. CentOS with "yum update whateverIwant" is so much easier--and no, I'm not registering any systems with Red Hat Network. I don't register for CentOS or Fedora.)
Serial is still well-served. I can log in to the serial console on my Linux server using minicom on Mac OS X and a Keyspan USB-Serial adapter. zmodem should be good to go. Linux to Linux is slightly easier, as all my Linux boxes have built-in RS-232 support (though some need a header-to-DE9/DB25 adaptor).
Though I've got those 8p8c modular adaptors on everything now so I don't have to find the right cables.
Depending on what's around, Kermit and patience might be the best solution. Ahh, memories of the HP-48 calculator... and mainframes... and lots of other things, now that I remember.
In a similar vein, you can bolt a newer desktop environment onto that Red Hat 7.2 box. At least, you can leave the kernel and glibc behind in ancient history, as the GUI is not part of the operating system. (It's part of the Red Hat distribution, but you can replace it with anything you like. Like replacing OS X Finder with PathFinder or something.)
Now, if you want suck, how about an NFS server that freezes all client connections because one filesystem it is serving is over 95% full and the filesystem driver has a disastrous big-O complexity problem in the free block location algorithm? One that isn't solved by the normal "reserved for superuser" hack that's been in UNIX since the days of a 14-byte filename limit.
Still, one could hope that people learned from Apple's experience with the Classic VM inside OS X.
The one thing I don't like from the transition is how Carbon/Classic:-separated paths are still hanging around in some interfaces.
But between Classic.app on OS X, Cygwin on Windows, and WINE on anything POSIX-ish (source API compatibility, not binary), there's plenty of work out there to serve as a template.
None of which Microsoft can use, I guess, because it's either Apple's or (L)GPLed.
How about, I've got a subcompact that's easy to drive into town and park, and my neighbour has a pick-up truck that's great for hauling stuff from the garden store and lumber yard.
When my neighbour borrows my car to go into town, I don't charge him gas. In return, though, he lets me borrow his truck to go to the lumber yard, and he doesn't charge me for gas. Some months I do a lot of gardening and use a bit more of his gas. Other months he has a lot of paperwork to file in town and uses a bit more of my gas. But it's not worth fighting over, and we solve it by bring beer to each other's backyard BBQ parties.
That's what peering is all about: I send my data on your network in return for you sending your data on my network. Since the amounts of data are nearly equal, any work spent to figure out exactly how much is an expense to both our businesses. If, instead, we just use a share-and-share-alike agreement, we each maintain our own plant, but are both better for being able to use the two networks instead of just one.
Hmmmm... aren't a couple of Jigoro Kano's (founder of Judo) favourite phrases (translated) "Maximum efficient use of power" and "mutual welfare and benefit"? Of course, with peering, both networks win.
You know, there's another way to solve that problem: Make the channel line up "lock" for X days after each change, where X is chosen so you can afford to run it.
I'd put money on the fact that they were being _charged_ bundle fees to carry the signals anyway, and used the customer service thing as an excuse to drop a la carte programming. ("Pick-and-pay" in Canadian.)
But are strings "code"? If they're not, then that's your two lines.
And if you say assuming a "STROUT" routine is present is cheating (there's something in the BASIC ROM you can use to do that, the KERNAL ROM only has CHROUT at $FFD2 I wish I could forget that), then you'd better make sure you count LOC in your compiler, assembler, linker, runtime libraries, and so on to make your higher-level 2-liner program run.
Which is all just to help point out how truly bogus such LOC counts are when talking about system complexity. Throw in microcoded processors and PLCs and DSPs, and where do you switch from "counting code" back to "counting transistors"?
LOCs are vaguely useless when measuring output of a programming team. For system complexity, you need a proper definition of "system" for it to make sense, and even then, comparing the complexity of two systems by "LOC" is almost certainly useless.
Other way around; the key zone on the disc is readable by any player. But on consumer writable media, it is pre-burned with zeros.
It is also part of the "meta information"; you can't see it with normal I/O commands like 'dd'. You'd have to have a device driver that implements a DVD-specific ioctl to retrieve the keys.
Since.iso is a headerless format, you'd need a lookaside file to contain the metadata needed to make a virtual drive that lets you mount an encrypted.iso. Or use an image file format that supports metadata.
Of course, the people writing the sort of stuff aren't trying to stay within the letter of a law that either (a) does not apply in their country or (b) they don't like, no-one bothers. AFAIK and all that.
In fact, implementing such a system may expose you to PATENT claims in the U.S.!
Don't confuse availability with archive and backup.
If S3 automatically replicates changes across all the remote stores, like RAID does, then any "damage" that's injected to the system will be replicated across all the copies automatically without any form of recovery.
This is (part of) why RAID is not a replacement for BACKUP.
HandBrake can do MKV files with DTS and AC3 pass-thru audio, and DVD "image" subtitles. (MP4 can only do AC3 pass-through and "text" subtitles.)
Between HandBrake and XBMC Live, I've got some lovely Blu-Ray-sourced.mkv files; DTS sound, multiple language subtitle channels all switchable, and so on....
Ugh. That reminds me of a disk formatter on QNX in the mid-80s. It asked 3 variations on "are you sure?", but worded so the answers were "y" (Yes, I'm sure), "n" (No, I don't want to abort) and "c" (Continue with the operation).
Which was fine once or twice. But it was very slow to load (as it had to load from floppy: no formatter on the hard disk because you can't format the only hard disk you can have). And if you get in the "are you sure" zone and keep typing Y you have to start all over booting from the floppy.
Had the drive not trashed itself once every 8 or 12 weeks I wouldn't have cared. Fortunately, it eventually filled up the bad block map and we could trash the whole system, as there was no budget for repairs.
So the way to test that is to grab a cheap Ethernet hub or switch. (Easily found for less than twenty bucks these days if you don't have a couple of spares already.) Hook the computer to the switch, the switch to the router, and the router to the Internet.
Disconnect the switch from the router while leaving the computer to switch cable connected. Windows will still see the media as valid, but the connection to the Internet will be dead.
Or use the MAC address filter feature on most home routers; at least the Little Blue Routers. Block the PC at the router, or even set up a schedule to turn it off and on automatically.
Not that I'm a fan of terminating all connections when the media goes down on the PC; I've re-cabled live UNIX servers with TCP just re-sending all the packets that got missed while the media was down.
Like an IT group that can't run their own systems, so is always trying to shift blame to other groups. Since they're always looking at problems the wrong way up (like rebooting UNIX servers rather than figuring out what the fault is), they burn a lot of time. This makes it politically bad for them to then actually admit that they were doing something wrong. Especially after getting large capital budgets for installing equipment that doesn't solve the problems.
Two things. On your first paragraph, all it takes is someone unfamiliar with the terminology to say something that doesn't mean what she thinks it means and you're in the wrong part of the call script. Given how confused the average end-user gets when faced with terms like CPU, PC, memory and so on, this is fairly easy to happen. Say Microsoft asks if the CPU was replaced, and the caller thinks that means the whole case thing, which hasn't been changed. (There's plenty of people who call the whole box the CPU.) Immediately you're going the wrong way, because the activation codes say "new CPU" but the customer says "no, same CPU".
On the second, a DVD player is a physical product. I have absolutely no qualms about replacing DVD DISCS that have failed with copies I made from friend's originals or rentals. About 2% of my collection has failed in this manner, more 20 discs, including the very first James Bond DVD releases. (I couldn't find a trace of the UPCs on imdb.com or amazon.ca; it's like the studios disavowed those particular pressings.)
So making a pirated copy of Windows when you have the retail box, store receipt and original disc? What's wrong with that? You've met the terms of the license agreement, but a technical fault is preventing you from using the software. Overcoming that fault does not make you a software pirate.
I have done similar things with commercial software. The company took our money for a particular configuration, but their license manager won't allow that configuration to be enabled. So I decompiled the code and removed the license manager. We pay our annual maintenance and are well within the legal terms of our support contract, it's just a defective technical limitation that's in the way.
Yeah, they patched the engine management software on my Subaru while it was in for an oil change. It apparently takes longer to do the patch than the oil change and vehicle inspection combined....
(Problem was, it would go into fast idle for a few seconds after declutching. So much for PZEV....)
That, and similar programs, have dumbed down public school curriculum so that quick learning kids aren't challenged, slow learning kids aren't threatened, and teachers don't have to actually do anything but follow a script. (I'm happy to allow learning speeds to be unrelated to ability. Especially during childhood; development rates varying the way they do.)
I was one of the last people through public school in Ontario where you could fail a year and be kept back to do it again. And I graduated from grade 8 in 1985. They've effectively eliminated grading since then; and back when I did it, all they had for "grades" was "excellent", "satisfactory", and "needs improvement".
High school, 9 to OAC (like 13) was a little better, especially since there were actual marks. Proper percentage marks, even. But, again, that's a while ago too.
Heck, my mom used to get into arguments with the teachers about me reading books in class. "He's not paying attention to the lesson," the teachers would complain. "But he's quiet and getting good marks in everything," my mom would say. "That's not the point." (Mind you, I only found out about this a few years ago.)
And now they complain about the literacy rate. Coincidence?
Treating everyone _fairly_ doesn't mean treating everyone _identically_. Especially in education, people learn in different ways, have different aptitudes, and so on. Any system which thinks every student is identical has already failed those students.
But yes, leaving it off the non-3G version doesn't just make manufacturing simpler for that model, it also lets people know you're cheap... or live somewhere that doesn't play nice with 3G... I'm already hearing we're not getting the 3G model until long after the U.S. gets it.
Since I _don't_ deal with a lot of Windows PCs, I _am_ running everything on 5.0 GHz. Well, OK, there's a "guest WiFi" at 2.4 GHz for friends, but I don't care if that's broken.
Most PCs I've seen with 802.11n only have the 2.4GHz support, and same with a lot of add-on cards and access points. The dual-band stuff is starting to be a little easier to get, though still a bit of a price-premium. Of course, since my machines are either Macs on Wi-Fi or Linux on gigabit copper, I'm already used to a price premium.
However, there's a good side effect of this: a certain eggy on-line store had a whole bunch of refurbished Netgear 5.0 GHz _only_ access points/bridges on for less than $40. So the couple of devices I have that don't do WAP and/or don't do 5.0 GHz are now using those things. Similarly, I set up a friend's office to use one so his Macs aren't drowned out by all the 2.4 GHz chatter in the area. (Downtown, right beside a condo and hotel, across the road from an office building... and almost everyone on channel 1, too.)
Heck, $40 is less than the price of a USB 802.11n adapter. So I bought a couple extra just in case.
So, if everyone else would just _stay off_ 5.0 GHz, I'll be very happy.
It also means we can browse over SSH to our home machines in case we want to check something that might be NSFW and don't want to risk any lag between "oops" and Cmd-W. I hate open concept offices.
Slashdot Mirror
I don't know where the GP poster is from, but in Ontario, Canada, the basic everyone-has-to-know-what's-in-it Driver's Handbook has the phrase "shift to neutral/de-clutch" in quite a number of places when describing emergency situations and some correct responses to them.
The problem is, so many people behind the wheel aren't drivers. They have no desire to be there, they don't care about knowing how to operate a car, they just want to get to where they're going. So they squeak by enough testing to get a license, never bother with Driver's Ed or any advanced driver skills course.
And then they say no-one ever told them what to do.
Which, of course, is wrong; they were told, they just weren't listening. And sure, it's hard to care about something when you don't have a need for it. But it's almost like wilful ignorance, like people who say, "Oh I'm not good with math" and never try.
The default host on my servers comes back with an empty page for all URLs. If you don't provide the Host: header, you don't get any host at all.
Originally, that was so my family couldn't stumble in to the Wrong VHost with a bad URL. (Mainly because, at the time, the Wrong VHost was the default one.) But I liked some of the side-effects, so I do it on other servers, too.
And I haven't cared about browsers who don't send a Host: header since about 2004. lynx sends a Host: header. The only time it's a bit irritating is when I'm using telnet to debug something, then I have to type an extra line.
Troll? Maybe.
But remember who wrote Commodore's BASIC on every machine they made: Microsoft.
The really sad part about the 64's BASIC was the PET/CBM line already had BASIC 4.0, and Commodore went back to 2.0 for the VIC-20 and 64. Which meant there was no disk drive support in BASIC, you had to remember to use file channel numbers and load a special filename to do a directory listing. DOS operations were done with funny open commands. (Back when DOS really was the operating system on the disk drive, not a product from Microsoft written for IBM. Most of the PET/CBM disk drives had two CPUs, though less RAM than the computer.)
Still, it was simple enough that you could load a "wedge" that would use the BASIC line-editor to write assembler programs. When you're that close to the silicon, you learn how computers work or you don't program them.
I put a WYSE sticker on my Mac and now it's a dumb terminal!
Or maybe I'm just terminally dumb....
And Sun does, indeed, include the source for GPL software on their media.
It would be up to you to keep the patches current on the sources as well as the binaries. Sun's done their bit.
Not that I like this; it's maddening enough having to use Red Hat Network for Red Hat Enterprise patches. (It's not that I don't want to pay for the contracts, I'm more than happy to do that. It's the resulting hoops that have to be jumped through. CentOS with "yum update whateverIwant" is so much easier--and no, I'm not registering any systems with Red Hat Network. I don't register for CentOS or Fedora.)
Serial is still well-served. I can log in to the serial console on my Linux server using minicom on Mac OS X and a Keyspan USB-Serial adapter. zmodem should be good to go. Linux to Linux is slightly easier, as all my Linux boxes have built-in RS-232 support (though some need a header-to-DE9/DB25 adaptor).
Though I've got those 8p8c modular adaptors on everything now so I don't have to find the right cables.
Depending on what's around, Kermit and patience might be the best solution. Ahh, memories of the HP-48 calculator... and mainframes... and lots of other things, now that I remember.
In a similar vein, you can bolt a newer desktop environment onto that Red Hat 7.2 box. At least, you can leave the kernel and glibc behind in ancient history, as the GUI is not part of the operating system. (It's part of the Red Hat distribution, but you can replace it with anything you like. Like replacing OS X Finder with PathFinder or something.)
Now, if you want suck, how about an NFS server that freezes all client connections because one filesystem it is serving is over 95% full and the filesystem driver has a disastrous big-O complexity problem in the free block location algorithm? One that isn't solved by the normal "reserved for superuser" hack that's been in UNIX since the days of a 14-byte filename limit.
Still, one could hope that people learned from Apple's experience with the Classic VM inside OS X.
The one thing I don't like from the transition is how Carbon/Classic :-separated paths are still hanging around in some interfaces.
But between Classic.app on OS X, Cygwin on Windows, and WINE on anything POSIX-ish (source API compatibility, not binary), there's plenty of work out there to serve as a template.
None of which Microsoft can use, I guess, because it's either Apple's or (L)GPLed.
How about, I've got a subcompact that's easy to drive into town and park, and my neighbour has a pick-up truck that's great for hauling stuff from the garden store and lumber yard.
When my neighbour borrows my car to go into town, I don't charge him gas. In return, though, he lets me borrow his truck to go to the lumber yard, and he doesn't charge me for gas. Some months I do a lot of gardening and use a bit more of his gas. Other months he has a lot of paperwork to file in town and uses a bit more of my gas. But it's not worth fighting over, and we solve it by bring beer to each other's backyard BBQ parties.
That's what peering is all about: I send my data on your network in return for you sending your data on my network. Since the amounts of data are nearly equal, any work spent to figure out exactly how much is an expense to both our businesses. If, instead, we just use a share-and-share-alike agreement, we each maintain our own plant, but are both better for being able to use the two networks instead of just one.
Hmmmm... aren't a couple of Jigoro Kano's (founder of Judo) favourite phrases (translated) "Maximum efficient use of power" and "mutual welfare and benefit"? Of course, with peering, both networks win.
Indeed; and Pink Floyd albums are just one example of where those "black holes" (to cite Bob & Doug McKenzie) actually contained music, not silence.
You know, there's another way to solve that problem: Make the channel line up "lock" for X days after each change, where X is chosen so you can afford to run it.
I'd put money on the fact that they were being _charged_ bundle fees to carry the signals anyway, and used the customer service thing as an excuse to drop a la carte programming. ("Pick-and-pay" in Canadian.)
I would like to subscribe to your team of people chipping away at the Windows machines with hammers service, please.
I'd like them to start with 4oz tack hammers and work up to 10lb sledgehammers over the course of a year.
Ah heck, just bring in the sidewalk drill and get it over with.
I can do a CBM/PET/VIC/C64-type helloworld subroutine in 3 lines of assembly:
But are strings "code"? If they're not, then that's your two lines.
And if you say assuming a "STROUT" routine is present is cheating (there's something in the BASIC ROM you can use to do that, the KERNAL ROM only has CHROUT at $FFD2 I wish I could forget that), then you'd better make sure you count LOC in your compiler, assembler, linker, runtime libraries, and so on to make your higher-level 2-liner program run.
Which is all just to help point out how truly bogus such LOC counts are when talking about system complexity. Throw in microcoded processors and PLCs and DSPs, and where do you switch from "counting code" back to "counting transistors"?
LOCs are vaguely useless when measuring output of a programming team. For system complexity, you need a proper definition of "system" for it to make sense, and even then, comparing the complexity of two systems by "LOC" is almost certainly useless.
Other way around; the key zone on the disc is readable by any player. But on consumer writable media, it is pre-burned with zeros.
It is also part of the "meta information"; you can't see it with normal I/O commands like 'dd'. You'd have to have a device driver that implements a DVD-specific ioctl to retrieve the keys.
Since .iso is a headerless format, you'd need a lookaside file to contain the metadata needed to make a virtual drive that lets you mount an encrypted .iso. Or use an image file format that supports metadata.
Of course, the people writing the sort of stuff aren't trying to stay within the letter of a law that either (a) does not apply in their country or (b) they don't like, no-one bothers. AFAIK and all that.
In fact, implementing such a system may expose you to PATENT claims in the U.S.!
Don't confuse availability with archive and backup.
If S3 automatically replicates changes across all the remote stores, like RAID does, then any "damage" that's injected to the system will be replicated across all the copies automatically without any form of recovery.
This is (part of) why RAID is not a replacement for BACKUP.
HandBrake can do MKV files with DTS and AC3 pass-thru audio, and DVD "image" subtitles. (MP4 can only do AC3 pass-through and "text" subtitles.)
Between HandBrake and XBMC Live, I've got some lovely Blu-Ray-sourced .mkv files; DTS sound, multiple language subtitle channels all switchable, and so on....
Ugh. That reminds me of a disk formatter on QNX in the mid-80s. It asked 3 variations on "are you sure?", but worded so the answers were "y" (Yes, I'm sure), "n" (No, I don't want to abort) and "c" (Continue with the operation).
Which was fine once or twice. But it was very slow to load (as it had to load from floppy: no formatter on the hard disk because you can't format the only hard disk you can have). And if you get in the "are you sure" zone and keep typing Y you have to start all over booting from the floppy.
Had the drive not trashed itself once every 8 or 12 weeks I wouldn't have cared. Fortunately, it eventually filled up the bad block map and we could trash the whole system, as there was no budget for repairs.
So the way to test that is to grab a cheap Ethernet hub or switch. (Easily found for less than twenty bucks these days if you don't have a couple of spares already.) Hook the computer to the switch, the switch to the router, and the router to the Internet.
Disconnect the switch from the router while leaving the computer to switch cable connected. Windows will still see the media as valid, but the connection to the Internet will be dead.
Or use the MAC address filter feature on most home routers; at least the Little Blue Routers. Block the PC at the router, or even set up a schedule to turn it off and on automatically.
Not that I'm a fan of terminating all connections when the media goes down on the PC; I've re-cabled live UNIX servers with TCP just re-sending all the packets that got missed while the media was down.
You get people like that in the real world.
Like an IT group that can't run their own systems, so is always trying to shift blame to other groups. Since they're always looking at problems the wrong way up (like rebooting UNIX servers rather than figuring out what the fault is), they burn a lot of time. This makes it politically bad for them to then actually admit that they were doing something wrong. Especially after getting large capital budgets for installing equipment that doesn't solve the problems.
Two things. On your first paragraph, all it takes is someone unfamiliar with the terminology to say something that doesn't mean what she thinks it means and you're in the wrong part of the call script. Given how confused the average end-user gets when faced with terms like CPU, PC, memory and so on, this is fairly easy to happen. Say Microsoft asks if the CPU was replaced, and the caller thinks that means the whole case thing, which hasn't been changed. (There's plenty of people who call the whole box the CPU.) Immediately you're going the wrong way, because the activation codes say "new CPU" but the customer says "no, same CPU".
On the second, a DVD player is a physical product. I have absolutely no qualms about replacing DVD DISCS that have failed with copies I made from friend's originals or rentals. About 2% of my collection has failed in this manner, more 20 discs, including the very first James Bond DVD releases. (I couldn't find a trace of the UPCs on imdb.com or amazon.ca; it's like the studios disavowed those particular pressings.)
So making a pirated copy of Windows when you have the retail box, store receipt and original disc? What's wrong with that? You've met the terms of the license agreement, but a technical fault is preventing you from using the software. Overcoming that fault does not make you a software pirate.
I have done similar things with commercial software. The company took our money for a particular configuration, but their license manager won't allow that configuration to be enabled. So I decompiled the code and removed the license manager. We pay our annual maintenance and are well within the legal terms of our support contract, it's just a defective technical limitation that's in the way.
Yeah, they patched the engine management software on my Subaru while it was in for an oil change. It apparently takes longer to do the patch than the oil change and vehicle inspection combined....
(Problem was, it would go into fast idle for a few seconds after declutching. So much for PZEV....)
No Child Left Behind.
That, and similar programs, have dumbed down public school curriculum so that quick learning kids aren't challenged, slow learning kids aren't threatened, and teachers don't have to actually do anything but follow a script. (I'm happy to allow learning speeds to be unrelated to ability. Especially during childhood; development rates varying the way they do.)
I was one of the last people through public school in Ontario where you could fail a year and be kept back to do it again. And I graduated from grade 8 in 1985. They've effectively eliminated grading since then; and back when I did it, all they had for "grades" was "excellent", "satisfactory", and "needs improvement".
High school, 9 to OAC (like 13) was a little better, especially since there were actual marks. Proper percentage marks, even. But, again, that's a while ago too.
Heck, my mom used to get into arguments with the teachers about me reading books in class. "He's not paying attention to the lesson," the teachers would complain. "But he's quiet and getting good marks in everything," my mom would say. "That's not the point." (Mind you, I only found out about this a few years ago.)
And now they complain about the literacy rate. Coincidence?
Treating everyone _fairly_ doesn't mean treating everyone _identically_. Especially in education, people learn in different ways, have different aptitudes, and so on. Any system which thinks every student is identical has already failed those students.
It's almost certainly the 3G antenna location.
But yes, leaving it off the non-3G version doesn't just make manufacturing simpler for that model, it also lets people know you're cheap... or live somewhere that doesn't play nice with 3G... I'm already hearing we're not getting the 3G model until long after the U.S. gets it.
Since I _don't_ deal with a lot of Windows PCs, I _am_ running everything on 5.0 GHz. Well, OK, there's a "guest WiFi" at 2.4 GHz for friends, but I don't care if that's broken.
Most PCs I've seen with 802.11n only have the 2.4GHz support, and same with a lot of add-on cards and access points. The dual-band stuff is starting to be a little easier to get, though still a bit of a price-premium. Of course, since my machines are either Macs on Wi-Fi or Linux on gigabit copper, I'm already used to a price premium.
However, there's a good side effect of this: a certain eggy on-line store had a whole bunch of refurbished Netgear 5.0 GHz _only_ access points/bridges on for less than $40. So the couple of devices I have that don't do WAP and/or don't do 5.0 GHz are now using those things. Similarly, I set up a friend's office to use one so his Macs aren't drowned out by all the 2.4 GHz chatter in the area. (Downtown, right beside a condo and hotel, across the road from an office building... and almost everyone on channel 1, too.)
Heck, $40 is less than the price of a USB 802.11n adapter. So I bought a couple extra just in case.
So, if everyone else would just _stay off_ 5.0 GHz, I'll be very happy.
It also means we can browse over SSH to our home machines in case we want to check something that might be NSFW and don't want to risk any lag between "oops" and Cmd-W. I hate open concept offices.