I always love "estimates" without any evidence to back them up. Of course, that whole article is completely lacking in any supporting evidence for ANYTHING!
Problem #2:
so we'll need to construct one new nuclear fission plant every other day
That is assuming a full switch-over to fission. I guess all of our current power plants are going to swallowed by the earth. I guess he is insisting that we get rid of ALL of our current hydro power plants, solar/wind facilities, natural gas, coal, oil, etc. I guess we are also incapable of building MORE of those non-fission plants (particularly the non-polluting ones) in the next 50 years.
Oh yeah... He also assumes technology isn't going to advance, AT ALL, in the next 50 years. So, the first (50-year-old) fission plant will produce exactly as much power as the last (brand-new/most-recent) plant.
Problem #3:
we can expect about 600 TW or so fairly easily
Yes, because it's of-course fairly easy to cover the entire surface of the earth with solar cells. Building and installing them won't result in any pollution either.
His "Terawatt" figures include things like home heating as well, where technologies like geothermal can be easily used, needing very little actual electricity (no fission plants for that).
What's the point if any service that feeds video to you compresses the crap out of it?
There are a lot of reason digital video is in the sad state that it's in, but compression really isn't the cause.
Digital video looks like crap, first of all, because they insist on using hardware MPEG-2 encoders. I wish I knew why! With hardware encoders, like they use for digital TV and DVDs, the quality is crap, and needs more than double the bitrate to look good. Just about any software MPEG-2 encoder will produce output that looks unbelievably better than DTV/DVDs. Seriously, if anyone knows any good reasons for this, I'd like to know. It seems insane to me that they spend $100 million to make a movie like T-2, and yet cheap-out on mastering the DVDs, making it look like staticy, noisy, blocky, crap.
The second cause is that they cheap-out on the decoder hardware, and don't include any codecs more modern than MPEG-2. If they would just switch to a decent modern video codec (I suggest VP3, but any will do) they could get the video down to less than 1/4 the bitrate they need now, and it would still look great. Of course they might even then decide that 1/4 isn't small enough, and lower the bitrate further, but modern codecs (VP3/VP7/H.264/etc) decay far more graciously than the more primitive codecs like MPEG-2 and MPEG-4, so people wouldn't be so offended by the results.
You're obviously not using the hardware acceleration of the videocard. Under Unix it's called XVMC, and AFAIK, no ATI drivers support it. NVidia's binary drivers do, fortunately, so I've watched plenty of 1080 material with only about 20% CPU utilization on a slower system than you've got.
It works so well because video is one of the strange areas of computing. MPEG-2 takes more CPU power to DISPLAY than it does to decode. This could be helped TREMENDOUSLY if general-purpose computers just shipped with one FGPA/DSP/ASIC chip on-board to handle tasks (like interrupts) which normal CPUs are TERRIBLE at.
My other point being that switching to a much more CPU intensive codec, which shrinks the datasize greatly, will not adversely affect speed. So much of the CPU power is going to displaying the video, that increasing the processing needed for decoding it would hardly be noticable. The big side-effect being that hardware which decodes MPEG-2 (like your videocard) will become obsolete, and you'll HAVE TO have a fast enough CPU to do everything on it's own.
Besides, if $100 is expensive to people in the 3rd world (and you can bet on that) they're not going to want to carry a laptop around where it can get stolen or damaged. What do you all think?
Every one of these that comes along is always over-arching, over-expensive, and still too inflexible to be worthwhile. It seems like they're not sure whether they're designing it for CEOs or poor children.
For example: They talk about how expensive the hinges are to allow the screen to rotate around, so you can use the laptop as a tablet. WHO THE FUCK CARES? Who really thinks: if it can't be folded out into a tablet, poor African school-children won't use it?
The same for the display. If they would just use a simple black&white LCD like digital watches (and the old Palm handhelds) it would be dirt cheap. Instead, they insist on color, backlighting, etc. You don't really need color to read a book, it just makes the occasional illustrations a bit nicer. And a small light that clips on the top of the laptop would be better and cheaper than a LCD backlight. Backlights don't light-up the keyboard for one thing.
Long after the United States is gone, there will still be the Internet.
Personally, I doubt it.
There will still be computer networking, of course. And there will probably be one (or more) global networks. That said, considering how craptacular many of the government-established telco monopolies are in many foreign countries, I think it might be one big network with the reliability of a cellphone call.
Considering how much political in-fighting there is across countries in Europe, I wouldn't be surprised if there were multiple internets, and/or certain countries would require fees from others to hook up to their internet... with interesting effects like all internet traffic into France being shut-off every day at 4pm:-)
Of course, that's all assuming the current distribution of world power we have today. If some country (that doesn't censor information like China) becomes a real economic powerhouse before the US internet gets turned-off, it just might be able to directly take the place of the USA, and keep the internet alive.
I think the former is the more likely senario, though.
Are you worth $65,000 per year? Maybe you're worth more or you value your time more?
The price you expect to be paid depends entirely on the structure of the job. If I have to drive there, put in my 40+ hours doing things I dislike, etc., you can bet it will take a lot of money to keep me there.
OTOH, if it's something easy, generally enjoyable, that I can do as much as I want, and whenever I want... then I would be willing to take a lot less money for my time.
How many people here get a 6-digit salary, but will happily spend a couple days fixing an old $40 VCR/DVD-player?
Someone needs to create an air interconnect standard that lets server room designers snap-on cold air supplies onto a standard "air-port" on the box or blade.
I did... It's called 3" ducting. Pretty much the same stuff that exhausts the hot air from your dryer, but an inch smaller, and available everywhere. Nice round 3" ducting fits nicely around 80mm exhaust fans.
The standard duct collars are quite easy to adapt to fit to computers. Drill 4 holes in it that line-up with the fan screws, and simply attach it.
Best thing about it is that almost every adapter you could possibly need is available, cheap, in any hardware store. Anything custom you can easily make with large (3") PVC tubing, a few tools, and pipe glue.
It would be absolutely perfect if companies would just ship their computers with the collars already attached.
Eventually you get to a point where very little cold air makes it to those servers in the middle of the room because it's flowing up through other vents before it can get there. What's the solution?
The solution, IMHO, is to have a much more powerful fan pushing that air around. You don't really need more AC, just more pressure...
Alternatively, you can make the vents in the middle much larger than the rest to compensate.
OS X drivers are not going to be available for 95% of your periphrials.
Which is a good reason for people not to buy Intel-based Macs right away, but has nothing at all to do with whether you should run it on non-Apple hardware or not.
Major companies will very quickly make available OS X drivers for their hardware to be sure. I hate Realtek as much as anybody, but dammed if their RTL8139 cards don't come with drivers for every platform you can imagine...
Open source drivers may help fill the driver vaccume as well.
No cure for cancer......but we get a subwoofer that can shake your fillings out. Sigh.
That's such a ridiculous argument... slightly worse than "in soviet russia".
Do you really believe these guys could have cured cancer if only they hadn't built this subwoofer?
We have the ability to do more than one thing at a time, and stopping all other activity will NOT help. In fact, if this subwoofer makes money, that may incidentally lead to slightly more money for cancer research. Not getting out of the bed in the morning, just because what you're doing isn't saving people's lives, is crazy.
Which probably is a type of false advertising, as they advertise the 14 day trial period as just that, the type of service you should expect.
Now could they POSSIBLY give you an accurate representation of the service you can expect, when you have no rental history at all to tell them which level of service you will normally qualify for? Their "trial" service is just that... It's not expected to (and can't) be a perfect and exact representation of regular service.
Also this algorithm creates artificial shortages in films that they don't have a large stock of as those films will end up on the top of the DVD players of people who simply never send back their films.
No, it's not an artifical shortage... It's a real, actual shortage. Your solution is to send the rarest movies to those who rent the most, and completely screw-over those who rent the least?
If somebody only rents 4 movies a month, it only stands to reason they should get first-shot at rare titles. If you rent 20 movies a month, the rare title might be unavailable the first 5 times, but odds are good that soon you'll be returning a movie right when they have excess copies of that single title, and you'll get it after a week or so, making wait times about equal, overall. I would know, I've been on both sides of it after years with netflix.
Your solution, to give the rare titles to more active renters, is going to ensure that those who keep netflix's profit margins up are NEVER going to be able to get the movie they want, which means your monthly fee is going to have to be raised.
I think everyone is just being overly hard on Netflix, because everyone wants something for nothing.
Apparently, the editor (samzenpus) doesn't know the difference between the POWER and the FREQUENCY of a speaker.
Re:Anyone else wondering the following?
on
Pixar For Sale?
·
· Score: 1
Why not Video iPod and downloadable "Pixar" films.
Umm, there are lots of downloadable Pixar (short) films on iTunes... They can't sell the familiar Disney/Pixar films you know and love because the rights are owned by Disney.
Besides, I don't understand the video iPod... Tiny 2.5" screen, and it's NOT EVEN 16:9! Letterboxed films on a 2.5" screen??? Who the hell would want that?
Use netflix for six months and watch how long it takes them to send you the next DVD. Watch this take longer and longer as time goes by.
See, now you're completely wrong about this. It has nothing at all to do with how much time has gone by, and everything to do with how many movies you rented in the past month. This also affects the availability of the DVDs in your queue that they don't have enough copies of.
When you first sign-up, they do ship as fast as they possibly can, and will continue to do that if you rent a relatively small number of DVDs each month. If you rent a lot, the turnaround time is only slightly longer, as they add your DVDs to the queue below the less-frequent renters.
Personally, I find this to be a prefectly appropriate trade-off. My DVDs are slowed-down only slightly, so somebody that is helping subsidize Netflix is slightly happier. I'll admit I think it's very unfortunate they don't mention this fact anywhere on their site, but it is widely available info.
I use it on my laptop and have no problem with this alleged "weakness" you speak of.
Really? How's the ACPI support on that laptop? Oh, right... non-existant. How many times can you hot-plug USB devices (on average) before it dumps you into the UKC> (kernel crash debug) prompt? How's the cardbus support? Does it lock-up your system on boot-up, or only when you plug-in a card?
I don't think you've used it for anything but a firewall, and that's where your views are skewed.
You couldn't be more WRONG. I used it almost exclusively as my desktop for YEARS. Since about 2.4 IIRC, up to 3.6, before I finally switched to FreeBSD. My annoyance with OpenBSD was borne out of many, many years of systematic frustration with it. And I still use it for firewalls and one of my fileservers, although I'm likely going to replace it with something else when security updates cease, and I have to chose between upgrading and manually merging config-file changes, or installing clean.
But it makes a decent workstation too. And it gets better with every release.
For varying levels of "decent". My standard of "decent" is clearly a lot higher than yours.
As recently as a year ago the port of Mozilla/Firefox was damn-near unusable. Unless you loaded-up all of KDE to use Konq (which still isn't a very good browser IMHO) there wasn't a single decent, non-crashing browser available. I suffered through using the incredibly buggy and tiny-font-inducing Netscape 4.X for far, far longer than users of any other OS. Other browser projects simply could not compile on OpenBSD.
I think it was only 2 years ago that OpenBSD first got a working port of MPlayer. And even then it barely worked. Even today, there's still no VCD/SVCD/Mode2 support on OpenBSD.
There's been plenty of driver problems, too. Soundcard drivers that would lock-up the system after a few minutes. DVD-ROM drivers that would freeze-up at about the 7GB mark for no good reason. Network interfaces that would go down for no reason. An XF86 distribution that was lacking modern videocard drivers, requiring a recompiling of the kernel to get a compiled-from-source X11 working. Non-existant USB support for a long-time after it was widely in-use. And many, many more things that don't come to mind immediately.
As far as improvements, OpenBSD does technically "improve" from one version to the next, as bugs get fixed and features get added. However, there's always a new and equally bad bug added, and plenty that just get passed-down through the versions for years.
Back in the FreeBSD 4.x days, I would call it a toss-up between the two BSDs (and that's mainly why I used OpenBSD), but 5.X has been a huge improvement. Much more user-friendly, less crufty, more hardware support and features, etc.
3.8 was a very easy upgrade, very comparable to say, upgrading a Debian stable release.
I've tried upgrading OpenBSD before. It's mostly smooth, but there's always some major "gotchas". Upgrading from about 3.4 to 3.5 (IIRC) while files from the new XF86 were installed to their correct location, the OLD XF86 install was still the one that would start-up, and the one programs were linking against. This wasn't noticable right away, but once you try to install a package that depends on X, it fails. Once you try to compile a port that needed the newer X11, it would spew out endless errors. Fixing that problem by hand was not easy. It would have been easier, and faster over-all, to install from scratch, copy over the config files, and re-build all the ports to get a working OpenBSD system. I'm willing to bet you've seen problem from your upgrade, that you're just selectively leaving out. The same goes for your OpenBSD experience as-a-whole. Zealotry is bad.
Anyway, as NetBSD and OpenBSD share a lot of code,
Denmark now generates 20% of its electricity from wind. Is that not a fraction?
For Denmark it is, for the US it wouldn't even be a noticable ammount. Even though there have been many attempts to make "alternative" energy sources practical, they are still only making up something like 5%. If you think you can do better, you're welcome to try.
Do you honestly think we can build %20+ of new capacity coming from solar at a cost even remotely comparable to wind?
Well, ignoring the specifics, yes. Solar is right at the point of being competitive with wind. Look at the latest 7 square-mile solar field in the works by So.Cal. Edison. Hydro and Nuclear are quite simply the only practical clean power sources that can provide power in the quantities needed in the US. Wind may work in some places, and geothermal may work in some places, but not here.
I'd be more than happy to encourage the expansion of current solar and wind stations, but they can't possibly meet the needs. Even if you spent massive ammounts of money to deploy them. The question is, and has been for 50 years, whether most of our electricity is going to come from coal, oil, natural gas, or nuclear. Given the choice (and yes, those are the only choices in the forseeable future) nuclear is by far the cleanest and most environmentally friendly option.
Wind gets going much faster and can be added incrementally.
Yes, but it can't provide anywhere near the levels of electricity nuclear can. Nor at anywhere near the cost.
Either way most of the gains are still to be made in conservation, which end-runs the whole generation discussion.
No, it doesn't. You can't conserve your way down to zero. Even if you could get effeciency up near 100% on everything, the increasing numbers of people in this country would still drive electricity demand. What if fully-electric cars replace conventional vehicles? Do you want coal to supply those electricity needs, or nuclear? Wind is going to need a revolution to be able to practically supply even a double-digit percentage of the electricity we use now. Just try to imagine how many wind-turbines would have to be installed to meet near-future needs.
Slashdot Mirror
I always love "estimates" without any evidence to back them up. Of course, that whole article is completely lacking in any supporting evidence for ANYTHING!
Problem #2:
That is assuming a full switch-over to fission. I guess all of our current power plants are going to swallowed by the earth. I guess he is insisting that we get rid of ALL of our current hydro power plants, solar/wind facilities, natural gas, coal, oil, etc. I guess we are also incapable of building MORE of those non-fission plants (particularly the non-polluting ones) in the next 50 years.
Oh yeah... He also assumes technology isn't going to advance, AT ALL, in the next 50 years. So, the first (50-year-old) fission plant will produce exactly as much power as the last (brand-new/most-recent) plant.
Problem #3:
Yes, because it's of-course fairly easy to cover the entire surface of the earth with solar cells. Building and installing them won't result in any pollution either.
His "Terawatt" figures include things like home heating as well, where technologies like geothermal can be easily used, needing very little actual electricity (no fission plants for that).
This "troll" mod is completely ridiculous.
Every complaint I stated about OpenBSD are/were known-bugs, all of which I experienced first-hand over the years.
If anyone has a problem with what I've said, speak-up and I'll respond. Post AC if you don't want your moderation reversed.
There are a lot of reason digital video is in the sad state that it's in, but compression really isn't the cause.
Digital video looks like crap, first of all, because they insist on using hardware MPEG-2 encoders. I wish I knew why! With hardware encoders, like they use for digital TV and DVDs, the quality is crap, and needs more than double the bitrate to look good. Just about any software MPEG-2 encoder will produce output that looks unbelievably better than DTV/DVDs. Seriously, if anyone knows any good reasons for this, I'd like to know. It seems insane to me that they spend $100 million to make a movie like T-2, and yet cheap-out on mastering the DVDs, making it look like staticy, noisy, blocky, crap.
The second cause is that they cheap-out on the decoder hardware, and don't include any codecs more modern than MPEG-2. If they would just switch to a decent modern video codec (I suggest VP3, but any will do) they could get the video down to less than 1/4 the bitrate they need now, and it would still look great. Of course they might even then decide that 1/4 isn't small enough, and lower the bitrate further, but modern codecs (VP3/VP7/H.264/etc) decay far more graciously than the more primitive codecs like MPEG-2 and MPEG-4, so people wouldn't be so offended by the results.
You're obviously not using the hardware acceleration of the videocard. Under Unix it's called XVMC, and AFAIK, no ATI drivers support it. NVidia's binary drivers do, fortunately, so I've watched plenty of 1080 material with only about 20% CPU utilization on a slower system than you've got.
It works so well because video is one of the strange areas of computing. MPEG-2 takes more CPU power to DISPLAY than it does to decode. This could be helped TREMENDOUSLY if general-purpose computers just shipped with one FGPA/DSP/ASIC chip on-board to handle tasks (like interrupts) which normal CPUs are TERRIBLE at.
My other point being that switching to a much more CPU intensive codec, which shrinks the datasize greatly, will not adversely affect speed. So much of the CPU power is going to displaying the video, that increasing the processing needed for decoding it would hardly be noticable. The big side-effect being that hardware which decodes MPEG-2 (like your videocard) will become obsolete, and you'll HAVE TO have a fast enough CPU to do everything on it's own.
Oh yes, like "The Great Leap Forward"... What a slow, but sure, step that was for China...
Every one of these that comes along is always over-arching, over-expensive, and still too inflexible to be worthwhile. It seems like they're not sure whether they're designing it for CEOs or poor children.
For example: They talk about how expensive the hinges are to allow the screen to rotate around, so you can use the laptop as a tablet. WHO THE FUCK CARES? Who really thinks: if it can't be folded out into a tablet, poor African school-children won't use it?
The same for the display. If they would just use a simple black&white LCD like digital watches (and the old Palm handhelds) it would be dirt cheap. Instead, they insist on color, backlighting, etc. You don't really need color to read a book, it just makes the occasional illustrations a bit nicer. And a small light that clips on the top of the laptop would be better and cheaper than a LCD backlight. Backlights don't light-up the keyboard for one thing.
Yes, because IP addresses just assign themselves...
It was a much bigger landmark when FreeBSD (finally) first got ported over to non-x86 systems. Not long ago actually.
This isn't as significant, IMHO.
Personally, I doubt it.
There will still be computer networking, of course. And there will probably be one (or more) global networks. That said, considering how craptacular many of the government-established telco monopolies are in many foreign countries, I think it might be one big network with the reliability of a cellphone call.
Considering how much political in-fighting there is across countries in Europe, I wouldn't be surprised if there were multiple internets, and/or certain countries would require fees from others to hook up to their internet... with interesting effects like all internet traffic into France being shut-off every day at 4pm
Of course, that's all assuming the current distribution of world power we have today. If some country (that doesn't censor information like China) becomes a real economic powerhouse before the US internet gets turned-off, it just might be able to directly take the place of the USA, and keep the internet alive.
I think the former is the more likely senario, though.
The price you expect to be paid depends entirely on the structure of the job. If I have to drive there, put in my 40+ hours doing things I dislike, etc., you can bet it will take a lot of money to keep me there.
OTOH, if it's something easy, generally enjoyable, that I can do as much as I want, and whenever I want... then I would be willing to take a lot less money for my time.
How many people here get a 6-digit salary, but will happily spend a couple days fixing an old $40 VCR/DVD-player?
I did... It's called 3" ducting. Pretty much the same stuff that exhausts the hot air from your dryer, but an inch smaller, and available everywhere. Nice round 3" ducting fits nicely around 80mm exhaust fans.
The standard duct collars are quite easy to adapt to fit to computers. Drill 4 holes in it that line-up with the fan screws, and simply attach it.
Best thing about it is that almost every adapter you could possibly need is available, cheap, in any hardware store. Anything custom you can easily make with large (3") PVC tubing, a few tools, and pipe glue.
It would be absolutely perfect if companies would just ship their computers with the collars already attached.
The solution, IMHO, is to have a much more powerful fan pushing that air around. You don't really need more AC, just more pressure...
Alternatively, you can make the vents in the middle much larger than the rest to compensate.
Which is a good reason for people not to buy Intel-based Macs right away, but has nothing at all to do with whether you should run it on non-Apple hardware or not.
Major companies will very quickly make available OS X drivers for their hardware to be sure. I hate Realtek as much as anybody, but dammed if their RTL8139 cards don't come with drivers for every platform you can imagine...
Open source drivers may help fill the driver vaccume as well.
I remember first playing DVDs on a PII clocked at 366MHz IIRC. It's not very CPU intensive at all.
If you can't do it with your 800MHz PIII, the software is bloated, or the cheap videochip is offloading lots of video processing to the CPU.
First make sure you've got the latest drivers for your video chipset, and that all possible hardware acceleration options are turned on.
Then try to play some DVDs with MPlayer, which is the fastest around: http://mplayerhq.hu/
Yes. By definition in-fact.
Oh yeah, you'd be laughing for a few years... Then your digital recorder will break down, and you'll stop laughing.
Better to start now.
That's such a ridiculous argument... slightly worse than "in soviet russia".
Do you really believe these guys could have cured cancer if only they hadn't built this subwoofer?
We have the ability to do more than one thing at a time, and stopping all other activity will NOT help. In fact, if this subwoofer makes money, that may incidentally lead to slightly more money for cancer research. Not getting out of the bed in the morning, just because what you're doing isn't saving people's lives, is crazy.
Yes, I know you're trying to be funny.
Oh yes they are... You just aren't able to see the infrared light they're putting out.
Be careful. The next OpenBSD release song is likely to be humppa music!
.
For those of you not laughing, it's impossible to explain: http://undeadly.org/cgi?action=article&sid=200507
Don't follow that link if you enjoy sanity.
Now could they POSSIBLY give you an accurate representation of the service you can expect, when you have no rental history at all to tell them which level of service you will normally qualify for? Their "trial" service is just that... It's not expected to (and can't) be a perfect and exact representation of regular service.
No, it's not an artifical shortage... It's a real, actual shortage. Your solution is to send the rarest movies to those who rent the most, and completely screw-over those who rent the least?
If somebody only rents 4 movies a month, it only stands to reason they should get first-shot at rare titles. If you rent 20 movies a month, the rare title might be unavailable the first 5 times, but odds are good that soon you'll be returning a movie right when they have excess copies of that single title, and you'll get it after a week or so, making wait times about equal, overall. I would know, I've been on both sides of it after years with netflix.
Your solution, to give the rare titles to more active renters, is going to ensure that those who keep netflix's profit margins up are NEVER going to be able to get the movie they want, which means your monthly fee is going to have to be raised.
I think everyone is just being overly hard on Netflix, because everyone wants something for nothing.
Apparently, the editor (samzenpus) doesn't know the difference between the POWER and the FREQUENCY of a speaker.
Umm, there are lots of downloadable Pixar (short) films on iTunes... They can't sell the familiar Disney/Pixar films you know and love because the rights are owned by Disney.
Besides, I don't understand the video iPod... Tiny 2.5" screen, and it's NOT EVEN 16:9! Letterboxed films on a 2.5" screen??? Who the hell would want that?
"Until DVDs are wiped off the face of the earth, the stand-alone DVD player will remain supreme.
If you want to play DVDs, you need a stand-alone DVD player...
Plain and simple.
Open Source will never change that."
--CaptScarlet22 circa 1998
See, now you're completely wrong about this. It has nothing at all to do with how much time has gone by, and everything to do with how many movies you rented in the past month. This also affects the availability of the DVDs in your queue that they don't have enough copies of.
When you first sign-up, they do ship as fast as they possibly can, and will continue to do that if you rent a relatively small number of DVDs each month. If you rent a lot, the turnaround time is only slightly longer, as they add your DVDs to the queue below the less-frequent renters.
Personally, I find this to be a prefectly appropriate trade-off. My DVDs are slowed-down only slightly, so somebody that is helping subsidize Netflix is slightly happier. I'll admit I think it's very unfortunate they don't mention this fact anywhere on their site, but it is widely available info.
Really? How's the ACPI support on that laptop? Oh, right... non-existant. How many times can you hot-plug USB devices (on average) before it dumps you into the UKC> (kernel crash debug) prompt? How's the cardbus support? Does it lock-up your system on boot-up, or only when you plug-in a card?
You couldn't be more WRONG. I used it almost exclusively as my desktop for YEARS. Since about 2.4 IIRC, up to 3.6, before I finally switched to FreeBSD. My annoyance with OpenBSD was borne out of many, many years of systematic frustration with it. And I still use it for firewalls and one of my fileservers, although I'm likely going to replace it with something else when security updates cease, and I have to chose between upgrading and manually merging config-file changes, or installing clean.
For varying levels of "decent". My standard of "decent" is clearly a lot higher than yours.
As recently as a year ago the port of Mozilla/Firefox was damn-near unusable. Unless you loaded-up all of KDE to use Konq (which still isn't a very good browser IMHO) there wasn't a single decent, non-crashing browser available. I suffered through using the incredibly buggy and tiny-font-inducing Netscape 4.X for far, far longer than users of any other OS. Other browser projects simply could not compile on OpenBSD.
I think it was only 2 years ago that OpenBSD first got a working port of MPlayer. And even then it barely worked. Even today, there's still no VCD/SVCD/Mode2 support on OpenBSD.
There's been plenty of driver problems, too. Soundcard drivers that would lock-up the system after a few minutes. DVD-ROM drivers that would freeze-up at about the 7GB mark for no good reason. Network interfaces that would go down for no reason. An XF86 distribution that was lacking modern videocard drivers, requiring a recompiling of the kernel to get a compiled-from-source X11 working. Non-existant USB support for a long-time after it was widely in-use. And many, many more things that don't come to mind immediately.
As far as improvements, OpenBSD does technically "improve" from one version to the next, as bugs get fixed and features get added. However, there's always a new and equally bad bug added, and plenty that just get passed-down through the versions for years.
Back in the FreeBSD 4.x days, I would call it a toss-up between the two BSDs (and that's mainly why I used OpenBSD), but 5.X has been a huge improvement. Much more user-friendly, less crufty, more hardware support and features, etc.
I've tried upgrading OpenBSD before. It's mostly smooth, but there's always some major "gotchas". Upgrading from about 3.4 to 3.5 (IIRC) while files from the new XF86 were installed to their correct location, the OLD XF86 install was still the one that would start-up, and the one programs were linking against. This wasn't noticable right away, but once you try to install a package that depends on X, it fails. Once you try to compile a port that needed the newer X11, it would spew out endless errors. Fixing that problem by hand was not easy. It would have been easier, and faster over-all, to install from scratch, copy over the config files, and re-build all the ports to get a working OpenBSD system. I'm willing to bet you've seen problem from your upgrade, that you're just selectively leaving out. The same goes for your OpenBSD experience as-a-whole. Zealotry is bad.
No, they shared a lot of code
For Denmark it is, for the US it wouldn't even be a noticable ammount. Even though there have been many attempts to make "alternative" energy sources practical, they are still only making up something like 5%. If you think you can do better, you're welcome to try.
Well, ignoring the specifics, yes. Solar is right at the point of being competitive with wind. Look at the latest 7 square-mile solar field in the works by So.Cal. Edison. Hydro and Nuclear are quite simply the only practical clean power sources that can provide power in the quantities needed in the US. Wind may work in some places, and geothermal may work in some places, but not here.
I'd be more than happy to encourage the expansion of current solar and wind stations, but they can't possibly meet the needs. Even if you spent massive ammounts of money to deploy them. The question is, and has been for 50 years, whether most of our electricity is going to come from coal, oil, natural gas, or nuclear. Given the choice (and yes, those are the only choices in the forseeable future) nuclear is by far the cleanest and most environmentally friendly option.
Yes, but it can't provide anywhere near the levels of electricity nuclear can. Nor at anywhere near the cost.
No, it doesn't. You can't conserve your way down to zero. Even if you could get effeciency up near 100% on everything, the increasing numbers of people in this country would still drive electricity demand. What if fully-electric cars replace conventional vehicles? Do you want coal to supply those electricity needs, or nuclear? Wind is going to need a revolution to be able to practically supply even a double-digit percentage of the electricity we use now. Just try to imagine how many wind-turbines would have to be installed to meet near-future needs.