Part of the problem is in how RPM allows requirements to be met; it can lead to quite a few different "can not install this package, can't install requirements as it conflicts with something else". In contrast, DEB has "x requires y" - it's not as open-ended, and trying to deviate from it is at-your-own-risk.
Not quite. IIRC diesel engines use the heat/compression to ignite while gasoline direct injection engines still require a spark plug.
Yes. I said that.
Wikipedia also states that GDI engines can reach compression ratios as high as 65:1 in certain conditions.
Yes, but there are clearly DI diesels with 14:1 compression. I was just saying the physical engine itself could very likely be easily converted to run diesel with different injectors and glowplugs instead of sparkplugs.
Crash tests (and casualty rates) beg to differ. That are different from "fender benders" where heavily built vehicles win at low velocity impacts.
Maybe. I've seen these older vehicles hit some pretty immovable objects (1500lb+ bull cow) at high speeds without destroying them. In fact, a little radiator goop and new headlamps (in the above case) and the vehicle drove away. The passenger was fine.
The Diesel passenger van with the best mileage is the E250 with 7.3 IDI diesel (late eighties/early nineties.) People get real-world mileage over 20 MPG by taking it easy. You are NOT going to get 30 mpg by turbocharging, although you MIGHT get 30 mpg by switching to a 3:1 axle (stock is typ. 3.55:1, sometimes 3.73:1, occasionally 4.11) and by driving it like a little girl.
I wasn't thinking stock; I was thinking custom E150 - different trans, different engine. Something like a turbo'd inline 4 Windsor on a manual 5 speed or a Cummins 4BT.
I really wish Ford had come out with that 4 cylinder powerstroke for the Ranger.
Not true. In the past 10 years, SUVs have gotten significantly smaller. The large vehicles in 1990 were larger still (as were the medium-sized ones). A 4-door sedan from 1980 would more comfortably seat 6 than a modern "sits 7" minivan or SUV.
To get any sort of size, you're going to have to get one of the largest vehicles available today. Hell, you could easily fuck in the back seats of some of those sedans - laying down.
This is true. The safety of all vehicles is not the same, however. But then, I'm likely subconsciously thinking "old truck" vs. "new sedan" in terms of safety, not "same model of a different year".
Get rear-ended often enough and you'll change your tune. For whatever reason, my wife and I have the worst fortunes in this regard. Insurance companies do not like it, and it has made our rate go up as a result. Assholes in sports cars running red lights, people not stopping for a red light at all, or even just a mild fender bender at a stop sign - almost invariably, the damage to a modern car seems to be in the 2-3k range.
For similar accidents, I'll take an old solid vehicle any day of the week. More often You might get the other vehicle's paint on your bumper or a dented quarter panel. For every potentially fatal accident, you'll have 50+ mild ones (guessing). More often than not, the injuries are incurred from the airbags, not the the accident itself.
I've seen people carted away in an ambulance with severe cuts and blood everywhere due to the front of their vehicles caving in on them like a tin can - modern cars, of course. I've seen these accidents: an older vehicle may have been (mostly) damaged, and there might be broken limbs, but certainly nothing requiring more than a paramedic.
In September, Symbian 3 was Nokia's latest great hope for becoming relevant in the modern smartphone market.
Are we paying attention to the same market? Symbian hasn't been relevant for a long time for smartphone use; they've been feature phones, as far as I'm concerned, and not terribly good ones at that.
As for MeeGo... why? I can understand the motivation behind it, but not why they're essentially throwing out Maemo to do so. Maemo was (is) already a long way to being mature, and Meego looks like a complete start from scratch. (Architecturally, it's also better - being based on deb packages instead of RPMs). I don't know if it's Intel stalling or Nokia, but Meego just isn't going anywhere fast enough for those interested. ("Soon enough to matter" isn't significant; a new market player with well-designed utility and good marketing will always have room to play.)
I'm quite disappointed in Nokia in this regard. They had a lot of potential when the N900 came out - and then they fell flat on their face. Nokia, with all their different execs leaving and various other product turmoil, looks like it's starting to flounder - and will soon be failing unless they pull their ship together from the top.
Windows Phone 7? Don't even go there. It's the same ball of suck that the iPhone is with regards to privacy and control, with none of the benefits. (Android isn't much better due to vendors locking the phones.)
I'm not sure, but I believe the 10.10 kernel does not have the as IO scheduler: it was removed from mainline in 31 or 32, or something like that, for some dumb reason.
It hasn't always been like that. I remember that, around 2000-2001, I was quite able to go about my business without much more than slightly perceivable lag doing:
* kernel compiling * XFree86 compiling * large file transfers * and so on
At the time, I actually didn't worry so much about having RAM, because I knew I had swap. It was mostly quite efficient for me. Performance degraded gracefully and gradually.
Remember the alternative at the time? Yeah, Windows - where copying files over the network would result in similar performance. Or MacOS, where you might as well just go make some coffee if you want to do anything IO intensive.
I'm not quite sure what it was that I was doing that was different than you (and others, apparently) were doing, but I do remember the almost immediate change (for me): it was around 2.6.18. Unfortunately, I don't remember anymore what it was that I'd moved from. It may have been a 2.4 kernel/distro (I was a big debian user, but I don't think I waited that long to upgrade:P). I know that current RedHat based distros still have 2.6.18 as their base (with massive backports), and it's plagued by the same I/O issues.
One thing I can say for certain, and that's that Stormix, Debian, and Progeny Debian didn't suffer from these problems in the 1999-2003 timeframe - for me. I had no problems.
I did find a couple things to help improve the situation, for my purposes, in the past couple years. They run contrary to convention, but agree with what I've heard from others running Linux in larger environments.
* Run the anticipatory scheduler instead of CFQ. CFQ sucks. (Unfortunately, they killed the as scheduler in.30 something.) * Don't use ext3. Ext3 + cfq = death, for whatever reason. I haven't tested ext4 yet enough to know one way or the other, but Iv'e historically used XFS (which, even with fragmentation, is better).
As for the cause, I can only guess. A couple things I have guessed, however:
* the orthogonal performance increases of RAM/CPU to storage. Were I to use the old software on new systems, maybe I'd see the same performance degradation - but in the late 90s, the performance differences were not as emphasized as rotating rust is to modern CPUs. * Changes in the kernel timing and multitasking/scheduling itself, to give more preference to disk IO in general. I haven't looked through change logs, but I'd suspect this has happeened on account of the big-enterprise contribution to linux in the last decade: they fix what's important to them. * The 2.6 kernel. I remember 2.4 was pretty awesome in terms of performance when it came out, but 2.6 didn't really do all that much for me. This goes with the previous possibility, in that 2.6 had a lot more enterprisey emphasis than 2.4.
Why would you run a vehicle if you're stopped? Just turn it off when appropriate.
That said, there are reasons why you might actually want to keep running your engine. For instance, if it's -20F below, and you'd like to keep the interior temperature comfortable enough to only wear a jacket and hat.
Are you serious? Have you actually seen an accident before? How about a vehicle made in the 90s or 80s?
Those older vehicles were not light. The bodies were made from cold rolled steel, with solid I-beam construction. They were much, much safer than most modern unibody designs, if only due to mass. They got similar or better fuel mileage due to the lack of restrictive emission add-ons.
Furthermore, newer cars aren't "safer". They handle better and are more controllable due to innovations in suspension and steering, and have a safer compartment resulting in better safety, but the vehicles themselves are less likely to survive even a 'mild' fender bender without thousands of dollars in a rebuild.
I've yet to see a modern passenger vehicle in a collision that didn't total the modern vehicle. A friend's 91 suburban was hit by a modern Honda Odyssey (late model): the Honda hit his rear passenger side quarter section. After replacing two sheered bolts and redoing the rear body panel, his Suburban was as good as new.
The Honda, on the other hand, looked like a box of legos had exploded all the way back to the middle passenger seats. It was a complete write-off, though I suppose you may have been able to salvage the rear gate and the seats. (The children in the back had to crawl out the rear gate, because the side doors would not open due to the frame twisting from the impact).
Meanwhile, a full-size 80s conversion van with an inline diesel is going to: * get over 20mpg, possibly 30mpg is you turbocharge it
I've also been in, or seen several other accidents. My dad had someone rear-end his '78 Oldsmobile Delta 88 a number of years ago: his bumper got bent in. The newer half-ton pickup was totaled.
I had someone crash into my van a couple of months ago in a parking lot - while I was parked and standing outside the vehicle. They hit my front bumper as they pulled into the spot across from me, probably no faster than 10-15mph (but not paying attention). How you can miss such a large vehicle is beside me, but whatever. Their late-model Civic was totaled.
Their hood crinkled like a tin can, their bumper was ripped to the ground, and the cap was torn off their radiator/their radiator was damaged/etc., and there was fluid everywhere. Thank God there were witnesses, because I doubt the cop would've believed what had happened on my account alone. My van had no damage other than some of the paint from their car scraping off onto mine (yay, chrome).
It boggles my mind that 25 years later most cars I can buy in the US get half of what my 25 year old car got. If that. It also means that getting 70 shouldn't be impossible. Thats 3.3l/100km, and it's been done.
Modern vehicles would be getting signfiicantly better mileage if it were not for all the emission-saving devices we've strapped to our vehicles in the meantime, because system efficiency has improved quite a bit. Diesels seem the most crippled by these things: you can easily gain 20-40% fuel economy improvements by removing or disabling the "emission controls" on a newer diesel engine, which is already getting comparable performance to your 85 Golf diesel.
I've got a friend with an 80s military Blazer with a diesel engine. He gets about 18-25mpg in it. It's got 1 ton axles and weighs quite a bit - yet still manages to get similar mileage to the newer, half-as-heavy (or less) hybrids and crossovers which utilize modern tech gas engines. Keep in mind that diesels have always had better MPG due to the energy density superiority of diesel, so comparing a diesel engine to a gas engine has never been (and never will be) fair to gas engines. (On the other hand, maybe that's the way it should be - gasoline takes more energy to produce per gallon than diesel does).
Climbing mountains is stupid in just about every single case, so that's actually a fitting example. So you've climbed the mountain and you have the scars to prove it. The rest of us are doing things that actually matter more than your temporary ego boost.
Look up 'metaphor' in the dictionary; it may prove useful in the future for purposes of not drooling all over your footwear.
Though, I doubt it. If you were able to see long-term picture gains or benefits, we'd not be having this discussion.
What do you mean "not supposed" to share addresses? Is that not what the unrouted address blocks (10, 172, and 192 prefixes) were intended for? Just because they started things out by carving out a single public IP for every machine does not mean that it's good practice or that it works well.
Non-routable address space makes sense to me from a number of vantage points.
* If, on IPv6, I move my office to another location and get a different upstream provider, how do I contend with internal address assignment? Complete redo? * In a web server farm, you gain nothing from multiple public addresses (and, tentatively, lose things like security certs). * Forwarding to hosts within a DMZ makes more sense from a security perspective as you're still able to have not-publicly-routed broadcast to those machines, on those interfaces.
Where, exactly, would a 'required 1:1 mapping' of IP to hosts improve things at all?
It's a bit old at this point, but it proved itself true as of 10.5 as well (at least on PPC hardware). The performance is noticeably bad in many cases.
It's significant to the point where people running FileMaker on Mac workstations have moved to the Windows Server version, server-side.
So how do you explain, say, the lack of Linux support for all those iPod owning Linux geeks. A winelib version would do. Or at least not bricking an iPod after loading music from a non iTunes app.
You misunderstood me. Those iPad using linux geeks moved over to apple as soon as they could justify the cost of an Apple workstation or laptop (around 10.3 or 10.4, I think it was).
I seem to recall early iPods worked just fine in Linux; I had a 2gig nano which I could sync with (I think it was) apod or similar. aptitude shows a number of ipod/iphone related tools, though I've never tried them (nor will I - I traded my iPod for a nice pair of boots).
I wasn't aware of the capabilities of the card in question, so included two.
I'm sorry, but a 'cinema' display means nothing to me. I prefer a wide screen format display for many terminals and side/side windows. Two of them is better, and I'd rather have a good LCD than a crap one for a workstation. That's what makes sitting in front of one all day tolerable.
So it is, in essence, a diesel engine - that runs on gasoline. IIRC, diesel engines are around 14:1-16:1 for a DI diesel. I'd wager a guess that they offer (or will offer) a Mazda 2 overseas with the same engine running diesel (with glow instead of spark plugs, of course).
The problem with IPv6 is that it relies upon local systems - private subnets, the endpoints of the internet, etc. - supporting IPv6.
This doesn't. It only requires the routers between the endpoints supporting it, which is a significantly smaller cost for adoption, putting the burden completely on upstream providers instead of putting a disproportionate burden on the end user.
I have seen several proposed implementations similar to this which allow LAN topology to remain working as-is.
Slashdot Mirror
What the hell. Somehow I managed to read your post as:
Connection: Sigourney Weaver Handheld
Need... more... sleep. Also, skimming and jumping about is bad.
Part of the problem is in how RPM allows requirements to be met; it can lead to quite a few different "can not install this package, can't install requirements as it conflicts with something else". In contrast, DEB has "x requires y" - it's not as open-ended, and trying to deviate from it is at-your-own-risk.
Sorry, I must have missed something.
I can only install OSX on Apple hardware.
I'm assuming this applies to OS X Server, too.
Am I missing the server-class hardware (minimum requirement: redundant PSUs) on their site?
Indeed, nice to meet you as well.
I've been doing IT for a bit over a decade now, and starting to feel old. :) Been on /. since before the days of Slashdot Radio. :)
Not quite. IIRC diesel engines use the heat/compression to ignite while gasoline direct injection engines still require a spark plug.
Yes. I said that.
Wikipedia also states that GDI engines can reach compression ratios as high as 65:1 in certain conditions.
Yes, but there are clearly DI diesels with 14:1 compression. I was just saying the physical engine itself could very likely be easily converted to run diesel with different injectors and glowplugs instead of sparkplugs.
Crash tests (and casualty rates) beg to differ. That are different from "fender benders" where heavily built vehicles win at low velocity impacts.
Maybe. I've seen these older vehicles hit some pretty immovable objects (1500lb+ bull cow) at high speeds without destroying them. In fact, a little radiator goop and new headlamps (in the above case) and the vehicle drove away. The passenger was fine.
The Diesel passenger van with the best mileage is the E250 with 7.3 IDI diesel (late eighties/early nineties.) People get real-world mileage over 20 MPG by taking it easy. You are NOT going to get 30 mpg by turbocharging, although you MIGHT get 30 mpg by switching to a 3:1 axle (stock is typ. 3.55:1, sometimes 3.73:1, occasionally 4.11) and by driving it like a little girl.
I wasn't thinking stock; I was thinking custom E150 - different trans, different engine. Something like a turbo'd inline 4 Windsor on a manual 5 speed or a Cummins 4BT.
I really wish Ford had come out with that 4 cylinder powerstroke for the Ranger.
And we insist on driving larger cars.
Not true. In the past 10 years, SUVs have gotten significantly smaller. The large vehicles in 1990 were larger still (as were the medium-sized ones). A 4-door sedan from 1980 would more comfortably seat 6 than a modern "sits 7" minivan or SUV.
To get any sort of size, you're going to have to get one of the largest vehicles available today. Hell, you could easily fuck in the back seats of some of those sedans - laying down.
This is true. The safety of all vehicles is not the same, however. But then, I'm likely subconsciously thinking "old truck" vs. "new sedan" in terms of safety, not "same model of a different year".
Get rear-ended often enough and you'll change your tune. For whatever reason, my wife and I have the worst fortunes in this regard. Insurance companies do not like it, and it has made our rate go up as a result. Assholes in sports cars running red lights, people not stopping for a red light at all, or even just a mild fender bender at a stop sign - almost invariably, the damage to a modern car seems to be in the 2-3k range.
For similar accidents, I'll take an old solid vehicle any day of the week. More often You might get the other vehicle's paint on your bumper or a dented quarter panel. For every potentially fatal accident, you'll have 50+ mild ones (guessing). More often than not, the injuries are incurred from the airbags, not the the accident itself.
I've seen people carted away in an ambulance with severe cuts and blood everywhere due to the front of their vehicles caving in on them like a tin can - modern cars, of course. I've seen these accidents: an older vehicle may have been (mostly) damaged, and there might be broken limbs, but certainly nothing requiring more than a paramedic.
a) News of symbians death are IMHO highly exaggerated.
Dollars to doughnuts, FreeBSD dies before Symbian.
In September, Symbian 3 was Nokia's latest great hope for becoming relevant in the modern smartphone market.
Are we paying attention to the same market? Symbian hasn't been relevant for a long time for smartphone use; they've been feature phones, as far as I'm concerned, and not terribly good ones at that.
As for MeeGo... why? I can understand the motivation behind it, but not why they're essentially throwing out Maemo to do so. Maemo was (is) already a long way to being mature, and Meego looks like a complete start from scratch. (Architecturally, it's also better - being based on deb packages instead of RPMs). I don't know if it's Intel stalling or Nokia, but Meego just isn't going anywhere fast enough for those interested. ("Soon enough to matter" isn't significant; a new market player with well-designed utility and good marketing will always have room to play.)
I'm quite disappointed in Nokia in this regard. They had a lot of potential when the N900 came out - and then they fell flat on their face. Nokia, with all their different execs leaving and various other product turmoil, looks like it's starting to flounder - and will soon be failing unless they pull their ship together from the top.
Windows Phone 7? Don't even go there. It's the same ball of suck that the iPhone is with regards to privacy and control, with none of the benefits. (Android isn't much better due to vendors locking the phones.)
I'm not sure, but I believe the 10.10 kernel does not have the as IO scheduler: it was removed from mainline in 31 or 32, or something like that, for some dumb reason.
It hasn't always been like that. I remember that, around 2000-2001, I was quite able to go about my business without much more than slightly perceivable lag doing:
* kernel compiling
* XFree86 compiling
* large file transfers
* and so on
At the time, I actually didn't worry so much about having RAM, because I knew I had swap. It was mostly quite efficient for me. Performance degraded gracefully and gradually.
Remember the alternative at the time? Yeah, Windows - where copying files over the network would result in similar performance. Or MacOS, where you might as well just go make some coffee if you want to do anything IO intensive.
I'm not quite sure what it was that I was doing that was different than you (and others, apparently) were doing, but I do remember the almost immediate change (for me): it was around 2.6.18. Unfortunately, I don't remember anymore what it was that I'd moved from. It may have been a 2.4 kernel/distro (I was a big debian user, but I don't think I waited that long to upgrade :P). I know that current RedHat based distros still have 2.6.18 as their base (with massive backports), and it's plagued by the same I/O issues.
One thing I can say for certain, and that's that Stormix, Debian, and Progeny Debian didn't suffer from these problems in the 1999-2003 timeframe - for me. I had no problems.
I did find a couple things to help improve the situation, for my purposes, in the past couple years. They run contrary to convention, but agree with what I've heard from others running Linux in larger environments.
* Run the anticipatory scheduler instead of CFQ. CFQ sucks. (Unfortunately, they killed the as scheduler in .30 something.)
* Don't use ext3. Ext3 + cfq = death, for whatever reason. I haven't tested ext4 yet enough to know one way or the other, but Iv'e historically used XFS (which, even with fragmentation, is better).
As for the cause, I can only guess. A couple things I have guessed, however:
* the orthogonal performance increases of RAM/CPU to storage. Were I to use the old software on new systems, maybe I'd see the same performance degradation - but in the late 90s, the performance differences were not as emphasized as rotating rust is to modern CPUs.
* Changes in the kernel timing and multitasking/scheduling itself, to give more preference to disk IO in general. I haven't looked through change logs, but I'd suspect this has happeened on account of the big-enterprise contribution to linux in the last decade: they fix what's important to them.
* The 2.6 kernel. I remember 2.4 was pretty awesome in terms of performance when it came out, but 2.6 didn't really do all that much for me. This goes with the previous possibility, in that 2.6 had a lot more enterprisey emphasis than 2.4.
Why would you run a vehicle if you're stopped? Just turn it off when appropriate.
That said, there are reasons why you might actually want to keep running your engine. For instance, if it's -20F below, and you'd like to keep the interior temperature comfortable enough to only wear a jacket and hat.
Are you serious? Have you actually seen an accident before? How about a vehicle made in the 90s or 80s?
Those older vehicles were not light. The bodies were made from cold rolled steel, with solid I-beam construction. They were much, much safer than most modern unibody designs, if only due to mass. They got similar or better fuel mileage due to the lack of restrictive emission add-ons.
Furthermore, newer cars aren't "safer". They handle better and are more controllable due to innovations in suspension and steering, and have a safer compartment resulting in better safety, but the vehicles themselves are less likely to survive even a 'mild' fender bender without thousands of dollars in a rebuild.
I've yet to see a modern passenger vehicle in a collision that didn't total the modern vehicle. A friend's 91 suburban was hit by a modern Honda Odyssey (late model): the Honda hit his rear passenger side quarter section. After replacing two sheered bolts and redoing the rear body panel, his Suburban was as good as new.
The Honda, on the other hand, looked like a box of legos had exploded all the way back to the middle passenger seats. It was a complete write-off, though I suppose you may have been able to salvage the rear gate and the seats. (The children in the back had to crawl out the rear gate, because the side doors would not open due to the frame twisting from the impact).
Meanwhile, a full-size 80s conversion van with an inline diesel is going to:
* get over 20mpg, possibly 30mpg is you turbocharge it
I've also been in, or seen several other accidents. My dad had someone rear-end his '78 Oldsmobile Delta 88 a number of years ago: his bumper got bent in. The newer half-ton pickup was totaled.
I had someone crash into my van a couple of months ago in a parking lot - while I was parked and standing outside the vehicle. They hit my front bumper as they pulled into the spot across from me, probably no faster than 10-15mph (but not paying attention). How you can miss such a large vehicle is beside me, but whatever. Their late-model Civic was totaled.
Their hood crinkled like a tin can, their bumper was ripped to the ground, and the cap was torn off their radiator/their radiator was damaged/etc., and there was fluid everywhere. Thank God there were witnesses, because I doubt the cop would've believed what had happened on my account alone. My van had no damage other than some of the paint from their car scraping off onto mine (yay, chrome).
It boggles my mind that 25 years later most cars I can buy in the US get half of what my 25 year old car got. If that. It also means that getting 70 shouldn't be impossible. Thats 3.3l/100km, and it's been done.
You're forgetting something significant: emissions laws/regulations.
Modern vehicles would be getting signfiicantly better mileage if it were not for all the emission-saving devices we've strapped to our vehicles in the meantime, because system efficiency has improved quite a bit. Diesels seem the most crippled by these things: you can easily gain 20-40% fuel economy improvements by removing or disabling the "emission controls" on a newer diesel engine, which is already getting comparable performance to your 85 Golf diesel.
I've got a friend with an 80s military Blazer with a diesel engine. He gets about 18-25mpg in it. It's got 1 ton axles and weighs quite a bit - yet still manages to get similar mileage to the newer, half-as-heavy (or less) hybrids and crossovers which utilize modern tech gas engines. Keep in mind that diesels have always had better MPG due to the energy density superiority of diesel, so comparing a diesel engine to a gas engine has never been (and never will be) fair to gas engines. (On the other hand, maybe that's the way it should be - gasoline takes more energy to produce per gallon than diesel does).
Climbing mountains is stupid in just about every single case, so that's actually a fitting example. So you've climbed the mountain and you have the scars to prove it. The rest of us are doing things that actually matter more than your temporary ego boost.
Look up 'metaphor' in the dictionary; it may prove useful in the future for purposes of not drooling all over your footwear.
Though, I doubt it. If you were able to see long-term picture gains or benefits, we'd not be having this discussion.
What do you mean "not supposed" to share addresses? Is that not what the unrouted address blocks (10, 172, and 192 prefixes) were intended for? Just because they started things out by carving out a single public IP for every machine does not mean that it's good practice or that it works well.
Non-routable address space makes sense to me from a number of vantage points.
* If, on IPv6, I move my office to another location and get a different upstream provider, how do I contend with internal address assignment? Complete redo?
* In a web server farm, you gain nothing from multiple public addresses (and, tentatively, lose things like security certs).
* Forwarding to hosts within a DMZ makes more sense from a security perspective as you're still able to have not-publicly-routed broadcast to those machines, on those interfaces.
Where, exactly, would a 'required 1:1 mapping' of IP to hosts improve things at all?
http://www.anandtech.com/show/1778
It's a bit old at this point, but it proved itself true as of 10.5 as well (at least on PPC hardware). The performance is noticeably bad in many cases.
It's significant to the point where people running FileMaker on Mac workstations have moved to the Windows Server version, server-side.
So how do you explain, say, the lack of Linux support for all those iPod owning Linux geeks. A winelib version would do. Or at least not bricking an iPod after loading music from a non iTunes app.
You misunderstood me. Those iPad using linux geeks moved over to apple as soon as they could justify the cost of an Apple workstation or laptop (around 10.3 or 10.4, I think it was).
I seem to recall early iPods worked just fine in Linux; I had a 2gig nano which I could sync with (I think it was) apod or similar. aptitude shows a number of ipod/iphone related tools, though I've never tried them (nor will I - I traded my iPod for a nice pair of boots).
I wasn't aware of the capabilities of the card in question, so included two.
I'm sorry, but a 'cinema' display means nothing to me. I prefer a wide screen format display for many terminals and side/side windows. Two of them is better, and I'd rather have a good LCD than a crap one for a workstation. That's what makes sitting in front of one all day tolerable.
Young folks? Yeah, that's me. I haven't been doing IT for long at all...
No, I just couldn't remember the specifics of how it transpired (my assumption is usually: Microsoft was doing their usual thing).
So it is, in essence, a diesel engine - that runs on gasoline. IIRC, diesel engines are around 14:1-16:1 for a DI diesel. I'd wager a guess that they offer (or will offer) a Mazda 2 overseas with the same engine running diesel (with glow instead of spark plugs, of course).
The problem with IPv6 is that it relies upon local systems - private subnets, the endpoints of the internet, etc. - supporting IPv6.
This doesn't. It only requires the routers between the endpoints supporting it, which is a significantly smaller cost for adoption, putting the burden completely on upstream providers instead of putting a disproportionate burden on the end user.
I have seen several proposed implementations similar to this which allow LAN topology to remain working as-is.