Actually, the electric motor makes for very good acceleration. The Prius, at least, is not a wimpy underpowered small cramped car. I think, based on the sales demand for the Prius and other hybrids, that your experiences and friends are not typical. There is a hell of a lot of interest in hybrids, fuel cells (even if a lot of people don't understand what a "hydrogen economy" really is), etc. Lots of people are worried about global warming, lots of people are worried about running out of oil.
The Prius battery pack is warranted for 100,000 miles or 8 years; in California and some other states, 150,000 miles or 10 years. The expected lifetime is 15 years. Based on the price of NiMH batteries I can pick up at Sam's Club, the raw cost of the batteries should be under $1200 (I can pick up 8 2AH NiMH AA cells for about $10; Prius is 6.5 AH at 288V (for the earlier model, less for later models), which works out to 240 x 4 batteries, or $1200 for somewhat higher capacity than needed, for later models it goes down to about $840). That's less than $.01/mile (at 150,000 miles).
If they got rid of all the complicated mechanical stuff and just made it an electric vehicle with a diesel generator, I think it would be more efficient, less expensive, and cheaper maintenance. You'd need higher capacity motors and generator, higher electrical power handling capability, but that shouldn't be any more expensive than a transmission or the complex power coupling, and would probably be significantly lighter and require less maintenance.
Where I live, garbage pickup is done by private companies, who charge different rates. They have different choices for where they take their garbage as well, with different fees, costs for transportation, etc. The local government has mandated that all garbage companies do recycling pickup as well. Garbage companies charge by the size and number of cans they'll pick up, so if you recycle and can use fewer cans or a smaller can, then you save (since recycled amounts aren't part of any limit). When people won't pay the extra $1/pickup so they don't have to take the can out to the curb, they'll also do recycling so they can pay for the cheaper service.
One way to make "free market" work better would be to add in additional taxes for the landfill to account for the "true cost" of dumping - free market doesn't work there because it tends to be too short-sighted, reasoning that future costs of dealing with full landfill will be dealt with by someone else later down the line. As landfill fills up, eventually free market demand increases the price, and increasingly it becomes cost effective to do "waste stream sorting" to pull recyclables out, but it is even less expensive to have the originators of the waste stream keep it separate. We just didn't have enough foresight to charge people more for not doing that sorting until we get to a crisis.
If people really aren't going to recycle unless they're hit with a hefty increase in garbage pickup cost, AND the "true cost" of not recycling is indeed higher than that time is worth, then the cost of garbage pickup SHOULD be made higher. Besides, I think your basic premise is wrong - plenty of people recycle voluntarily, with NO direct economic incentive. I've seen people worth mega-millions who compost their kitchen scraps, keep several containers for recycling cans and plastic, bag newspapers and magazines, haul stuff to the recycling dropoff, even collect their "California redemption value" cans and bottles - it's not like they NEED that $1.75 they get back.
"When you add in the intangibles like more free time the auto workers had to spend with their family instead of building you a hybrid" -- I'm not sure where you're going there. Free market works fine when the costs are up front. To the extent it takes more time to build a hybrid (if that's even true), it will be reflected in the sticker price.
As people get more experience with hybrids, the "cost of disposal" tends to get built in to the price/demand. Disposal will be part of the maintenance price (of replacing the batteries when required), and ultimate disposal will be reflected in resale -> resale ->... -> junking price (except where the junk yard isn't properly regulated or taxed to reflect real long term costs). Ability to resell does trickle up to the amount you're willing to buy it for. IF true costs for externalities were added in (to the price of the car and the price of the gas) through taxes, it should be a neutral economic decision whether to buy one type or another - you'll pay for what it really costs.
"By using more gas you're contributing more to society monetarily" - no, not if your long-term cost to society is higher than the taxes on that gas.
Then you're in favor of adding a "consumption" tax to account for the real cost of products. Add on a tax per pound for garbage, but allow free drop off of recyclables. Add on extra gas taxes. Add on disposal taxes on sale of anything that poses specific problems (e.g. CRTs).
The free market won't do it, but those are real costs that a "rational" person would pay (unless it is voluntary, thus allowing freeloaders). If most people would just do it voluntarily, a few freeloaders who don't want to "unless there's something in it for me" won't hurt things too much. But you're right, most people are too short-sighted to do it unless you give them an economic incentive. I'd rather not have such taxes, it just makes things more complicated, but since the majority of people apparently are like you, we probably need them.
Diesel is not necessarily more expensive than gasoline, it all has to do with immediate supply and demand. At one point, the big attraction of diesel cars (Mercedes, VW) was the lower price of diesel fuel. In addition, diesel has a higher energy content than gasoline, thus higher MPG. Finally, a diesel engine is more efficient overall.
Diesel engines have become much cleaner. A solution I'd like to see is to eliminate the complex mechanical interlocks current hybrids use to transfer power from engine or electric motor to wheels, from engine to generator for charging the batteries, and from wheels to regenerative braking. Go as if in a straight electric, except use a diesel engine for charging the system. The diesel can power off when stopped or for short trips, and the battery can provide an extra boost for acceleration.
Probably not conscious. A lot of drivers seem to inch forward every few seconds no matter what, and most of them don't seem to even notice. Others will move forward because they see someone else move forward. You moving forward to make a right turn triggered him to move forward. Doesn't make it any less irritating, but it probably wasn't "I'm so big, I can block his view if I just move forward a bit".
It should be pretty easy to check for. Remember all those "college music stores" that were reporting loss of sales due to "piracy by college students using Napster on high speed Internet access provided on campus"? Then all those schools that locked down Napster? Any figures on sales at those stores compared to sales where the networks weren't locked down? I'd think if there was any indication to support their contention, the RIAA would have trumpeted such figures as widely as possible.
Not only that, when I'm there I want my watch to let me know how long it is until lunch time. I don't want to have to remember that lunch is now at 20:00 instead of the usual 17:00 I'm used to, or that I should set my alarm clock to 16:00 instead of 14:00 when I want to make a morning meeting.
Also, Linux originally started out as a terminal emulator which Linus wrote, in part to learn how interrupts worked in the '386. It grew from there. Minix was just the development environment that Linus used to write, compile, configure and write to floppy disk for booting. He also used the minix file system format originally (which is still available, last I looked).
Reading his letter, where he says that OpenServer 6 costs $599-1399 one-time and you never have to pay support costs again (since he's comparing it to support costs for Red Hat), then goes on later to boast about how good their tech support is, he never says how much you have to pay for that 24x365 tech support. If all you want is what that one-time price gives you ("license, software fixes, and access to their online knowledge base", then why are you buying a support contract from Red Hat? Just download it from the Internet or buy a CD or even buy a "Dummies Guide to Fedora" which includes a CD, and off you go.
He does have a few good points. Point 1 ("Linux isn't free") isn't one of them.
Point 2 ("Kernel is better than Linux") is debatable - but should be easy to show in a shoot-out. Check memory management, forking processes, creating and destroying threads, file system performance, network performance, context switching performance - all running on identical machines (probably on several configurations), and each one as installed, and as tuned to do the best it can.
Point 3 (better security) has meaningless stats. You'd need to compare similarly managed machines with similar tasks and network infrastructure, and the stat given (the ridiculously over-precise 65.64%) is meaningless without comparative number of systems.
Point 4 has some validity - as a single company, they can probably put out a much more definitive roadmap. They can also just as easily change direction rendering the old roadmap meaningless.
Point 5 (backwards compatibility) is also somewhat valid (I'm taking them on their word that they really do attempt and succeed at it) - but a Linux installation can be just as "backwards compatible" as you need, in several ways: kernel interfaces rarely change old behaviors, libraries can have multiple versions (and dependency tracking handles it), most things are in source and can be recompiled if needed. If you need to do regression testing on your applications every time the system changes, then don't change it very often. Just because you CAN update things on a daily basis doesn't mean you should. And "who is verifying..."? Why, the distributors are!
Point 6 (So You Don't Want To Be An OS Expert) is irrelevant - if you need to buy support, buy it - installing a standard commercial distribution requires the same level of expertise as installing an SCO system.
Point 7 (Warranteed) is just FUD - if you need a warrantee, buy your distribution from someone who will give you one. Maybe you can then argue on price, but don't confuse the support you get on a free download with the support you get when you buy a distribution or a support contract. For example, how long a warrantee period do you get when you buy OpenServer 6 from SCO for $599 without any further support contracts?
Point 8 (No Forks) just barely has a valid point: if you want or need a system that is certified to meet certain standards, then buy a distribution that does so. Linux-the-Kernel and Linux-the-tools and Linux-the-desktop aren't scattering all over the place - there are alternate choices for all sorts of things, and maybe even a few of those alternates are forks, but the system as a whole is pretty well "unified".
Point 9 (reliability) has no evidence to support it - plenty of Linux installations are "locked in a closet and hide the key". It's true that a company's reliability is valued by customers, which is why Red Hat and others have an incentive to be reliable. No different from SCO.
Point 10 (Customer Support) is once again pure FUD - I don't know what reputation various Linux support companies have compared to SCO, but there's no reason they wouldn't have support engineers who know how things work every bit as much as SCO engineers (unless SCO has K & R working for them now), and that's not even getting into the possibility of getting help without needing a support contract.
Whew, this wasn't going to be a point-by-point critique, just a comment on the "Linux isn't free" thing.
Just as some VCRs pick up the program information and some don't, some will pick up time from a station (often PBS) and some won't. I don't know if the protocol includes a DST flag or if the unit is supposed to figure out when to switch on its own.
My ReplayTV doesn't have any option to set the time, only the time zone. Again, I don't know if the time protocol it uses each night when it connects to the server includes a DST flag, nor what it is supposed to do if it doesn't talk to the server on the day before the switch (if it does have a DST-pending flag). All I'll be able to do is change timezones for the three weeks before and 1 week after the automatic switch, if it doesn't compensate properly. It may just be simpler to turn off DST and leave it set to a different timezone for the entire time. I'm not sure if US units will have a way to adjust like that for people already in EST, though - what do you switch it to? I'm also not sure if the schedule info is stored in UCT or "local time" - if in UCT, then I don't have to worry about it not recording the right show, I just get confused by the time displays. I'd guess it stores everything internally by UCT, or it would get very confusing in the fall when the same time repeats itself.
Java's bloat doesn't seem to be that much. I opened up a fresh copy of Safari, and loaded the current time page from time.gov (e.g. for my time zone). Real memory usage was about 13.5MB (about 161 MB virtual). Loading the Java version of that page boosted that to about 30MB (383 MB virtual). What's interesting about that page is that they use Java only for doing the polling of the current time from the server, while doing the display updating (including figuring out the light/dark display on the map) in Javascript.
By comparison, with two windows, each with several tabs, and with the JVM already initialized, I'm using a total of 188MB real memory and 630MB of virtual - and now it's also taking up about 25% of my CPU on this Slashdot response page (either from the text box, or from the animated GIF, I'm not sure), even without doing any typing. Typing into it brings it up another 5%. 5 Freaking Percent just to process my keys and plot some text! How ever did we get anything done on 1MHz 8-bit processors? It would take about 100 of them just to do the processing needed to type some text and word-wrap it. Now it's running at 70% CPU for some reason. Crazy. Running Google Maps, scrolling around or zooming in or out with the satellite view uses about 50% CPU. Just loading the map brings real memory use up above that needed for loading Java (though only increasing virtual by.about 20MB), though that obviously grows as you scroll around and it caches more of the map.
Let's see, Macs don't matter because they are such a small market share, but Linux and BSD are suddenly so important that Java not being available on them matters (because somehow it is too difficult to install for some reason, even though it is available as an RPM, and available as source, so anyone could build a package for any distribution; I can see why Debian wouldn't include it, the SCSL is a little bit too controlling, but they could include it as part of non-free, for instance). So how is it not Microsoft's fault that the only platform that really matters, Microsoft Windows, doesn't support it properly?
And yes, I've installed Java, from source and from binary packages. I don't recall any particular problems with it.
Mac OSX also comes with Python, PERL and Tcl (though not Tcl/Tk, for some reason - but there's a very easy install of that as well - I particularly recommend the "Batteries Included" release, it comes with almost every extension you could ever want already loaded and ready to go), not to mention the usual suite of Unix tools (awk, sed, shells, grep, locate, find, xargs, etc.).
So that basically supports my thesis that the reason AJAX is being used is that a better solution isn't available, on of those better solutions being Java, because Microsoft has torpedoed anything that could threaten their hegemony of the desktop. Look for IE7 to add "compelling new features" that, if you use them, make you incompatible with anyone else, and also subtly start changing things so that developers without a clue will "fix" AJAX applications the easy way, breaking them totally for any other browser.
Why would Java on Mac be so much faster than anything else? I just can't fathom a JVM that is so slow that a Javascript interpreter can be faster. 165K of Javascript (Google Maps) does parse and execute "fast enough" on a fast enough computer, but I'm reminded of the saying that with enough power, even a brick can fly.
As for "what do you fall back to", I'm saying that the existence of a JRE on a machine "ought to be" as ubiquitous as "a browser". I mean, what's your fallback if the user doesn't have a browser? I'm tired of having my choices limited by what Microsoft chooses to support. It's great that Firefox is complete enough to be mostly compatible with IE, but as long as Microsoft holds a lock, Firefox won't be able to introduce any new features (at least not any that will be widely used), and will have to track whatever Microsoft does. How is that less proprietary than Sun? And where the heck is Kaffe (I mean that figuratively, I know literally where it is).
I don't recall the exact terms of the agreement between Microsoft and Sun regarding Java - why wouldn't Sun allow Microsoft to include a stock (unmodified by Microsoft) version of Java with a Windows install? I thought the disagreement was with Microsoft making unauthorized changes to the base classes.
Javascript isn't "just there", it is part of a runtime also. My operating system happens to come with that runtime (Safari), along with Java applets (that can run inside or outside a browser), Java Web Start, and Java application support. I'm sorry you're using a backwards OS. If everyone had Tiny BASIC on their machine, would you advocate using that as a common base as well, just because it is there?
Safari loads in less than half a second (fresh boot, so nothing even cached). Initializing Java for an applet take less than 2 seconds for the first time, not perceptible for later times. Applet class files are cached, just like the 165K Google Maps Javascript file is cached. I just don't see the problem. My machine is supposedly not all that fast (1.8GHz G5 with 1.25GB memory).
It's called "indexing". Same way they search the Internet. You didn't think they go out and read all the sites on the Internet when you search for "Britney Spears nude", did you? They do the same thing with your e-mail as it comes in. A decent e-mail client or operating system can do the same thing (see, e.g. Tiger). I haven't used the e-mail client from Tiger, but the one from the previous version does full-text indexing of e-mail, and can do full-text indexing of specified directories, for fast searching.
Google doesn't have magic hardware, they aren't reading through 800MB of your e-mail every time you do a search.
The logistics of moving that many people isn't overwhelming. There are over 650 million "enplanements" per year in the United States alone. 10 years at that rate would move all of the people currently here. In 10 years time, by some projections, world population will be less than 1 billion more than it is now - so about 2 more years would finish the job off.
As far as energy required, it would be large, but (assuming perfect efficiency, and only talking about lifting ~180 lb people, not including luggage, supplies or vehicles) about 3% of current worldwide electrical energy generation capacity (3600 GW) would keep up with that rate (at 24 hours/day, 365 days/year). Obviously, we'd need a more efficient method than rockets to even approach that, but it means it isn't impossible.
Of course, to merely remove a "significant" number of people would take much less. The above was for 100% removal. 25% of that rate is still twice the current worldwide population growth rate, and it would still only take about 60 years to remove the entire population.
Doesn't mean its practical, or that it is desirable, but it certainly isn't impossible (given reasonable advances in technology over the next 50 years).
Yes, it should be clear from my comments about using Safari that I have a Mac. On a 1.8GHz G5, which is apparently now not considered very fast. I've used Java applications, delivered over a network, and they were quite acceptable even on slow machines. I found Java speed to be reasonable even on a machine so slow that I could type about 10-15 times faster than Mozilla could accept text into a text box (a 200MHz PPC).
Why doesn't Gmail use Java? I don't know, possibly because it isn't the correct buzzword anymore. Does IE on Windows machines come with Java installed? One that works properly and is reasonably recent?
The only reason more people don't have Java easily available is because of Microsoft. I don't see that as a good reason to try to shoehorn the Web model and a hodgepodge language into something it isn't very good at.
There is absolutely no reason that Java applications can't be delivered and updated every bit as easily as downloading gobs of Javascript on every page. I don't understand the comment about Applets taking forever to load - bytes is bytes, Java bytecode is not much different in size than the source code it came from. Javascript isn't some radically different language where you can magically do really complex things in enormously less code than anything else. I certainly don't find that Applets take forever to load.
Versioning issues with regard to a JVM (and, more importantly, the standard class libraries expected to be there) is not more of a problem than versioning issues with different browsers, and there's no reason that different implementations of Java would be more incompatible than different implementations of Web browsers and Javascript.
Slashdot Mirror
Actually, the electric motor makes for very good acceleration. The Prius, at least, is not a wimpy underpowered small cramped car. I think, based on the sales demand for the Prius and other hybrids, that your experiences and friends are not typical. There is a hell of a lot of interest in hybrids, fuel cells (even if a lot of people don't understand what a "hydrogen economy" really is), etc. Lots of people are worried about global warming, lots of people are worried about running out of oil.
The Prius battery pack is warranted for 100,000 miles or 8 years; in California and some other states, 150,000 miles or 10 years. The expected lifetime is 15 years. Based on the price of NiMH batteries I can pick up at Sam's Club, the raw cost of the batteries should be under $1200 (I can pick up 8 2AH NiMH AA cells for about $10; Prius is 6.5 AH at 288V (for the earlier model, less for later models), which works out to 240 x 4 batteries, or $1200 for somewhat higher capacity than needed, for later models it goes down to about $840). That's less than $.01/mile (at 150,000 miles).
If they got rid of all the complicated mechanical stuff and just made it an electric vehicle with a diesel generator, I think it would be more efficient, less expensive, and cheaper maintenance. You'd need higher capacity motors and generator, higher electrical power handling capability, but that shouldn't be any more expensive than a transmission or the complex power coupling, and would probably be significantly lighter and require less maintenance.
Where I live, garbage pickup is done by private companies, who charge different rates. They have different choices for where they take their garbage as well, with different fees, costs for transportation, etc. The local government has mandated that all garbage companies do recycling pickup as well. Garbage companies charge by the size and number of cans they'll pick up, so if you recycle and can use fewer cans or a smaller can, then you save (since recycled amounts aren't part of any limit). When people won't pay the extra $1/pickup so they don't have to take the can out to the curb, they'll also do recycling so they can pay for the cheaper service.
One way to make "free market" work better would be to add in additional taxes for the landfill to account for the "true cost" of dumping - free market doesn't work there because it tends to be too short-sighted, reasoning that future costs of dealing with full landfill will be dealt with by someone else later down the line. As landfill fills up, eventually free market demand increases the price, and increasingly it becomes cost effective to do "waste stream sorting" to pull recyclables out, but it is even less expensive to have the originators of the waste stream keep it separate. We just didn't have enough foresight to charge people more for not doing that sorting until we get to a crisis.
If people really aren't going to recycle unless they're hit with a hefty increase in garbage pickup cost, AND the "true cost" of not recycling is indeed higher than that time is worth, then the cost of garbage pickup SHOULD be made higher. Besides, I think your basic premise is wrong - plenty of people recycle voluntarily, with NO direct economic incentive. I've seen people worth mega-millions who compost their kitchen scraps, keep several containers for recycling cans and plastic, bag newspapers and magazines, haul stuff to the recycling dropoff, even collect their "California redemption value" cans and bottles - it's not like they NEED that $1.75 they get back.
"When you add in the intangibles like more free time the auto workers had to spend with their family instead of building you a hybrid" -- I'm not sure where you're going there. Free market works fine when the costs are up front. To the extent it takes more time to build a hybrid (if that's even true), it will be reflected in the sticker price.
As people get more experience with hybrids, the "cost of disposal" tends to get built in to the price/demand. Disposal will be part of the maintenance price (of replacing the batteries when required), and ultimate disposal will be reflected in resale -> resale -> ... -> junking price (except where the junk yard isn't properly regulated or taxed to reflect real long term costs). Ability to resell does trickle up to the amount you're willing to buy it for. IF true costs for externalities were added in (to the price of the car and the price of the gas) through taxes, it should be a neutral economic decision whether to buy one type or another - you'll pay for what it really costs.
"By using more gas you're contributing more to society monetarily" - no, not if your long-term cost to society is higher than the taxes on that gas.
Then you're in favor of adding a "consumption" tax to account for the real cost of products. Add on a tax per pound for garbage, but allow free drop off of recyclables. Add on extra gas taxes. Add on disposal taxes on sale of anything that poses specific problems (e.g. CRTs).
The free market won't do it, but those are real costs that a "rational" person would pay (unless it is voluntary, thus allowing freeloaders). If most people would just do it voluntarily, a few freeloaders who don't want to "unless there's something in it for me" won't hurt things too much. But you're right, most people are too short-sighted to do it unless you give them an economic incentive. I'd rather not have such taxes, it just makes things more complicated, but since the majority of people apparently are like you, we probably need them.
Diesel is not necessarily more expensive than gasoline, it all has to do with immediate supply and demand. At one point, the big attraction of diesel cars (Mercedes, VW) was the lower price of diesel fuel. In addition, diesel has a higher energy content than gasoline, thus higher MPG. Finally, a diesel engine is more efficient overall.
Diesel engines have become much cleaner. A solution I'd like to see is to eliminate the complex mechanical interlocks current hybrids use to transfer power from engine or electric motor to wheels, from engine to generator for charging the batteries, and from wheels to regenerative braking. Go as if in a straight electric, except use a diesel engine for charging the system. The diesel can power off when stopped or for short trips, and the battery can provide an extra boost for acceleration.
Probably not conscious. A lot of drivers seem to inch forward every few seconds no matter what, and most of them don't seem to even notice. Others will move forward because they see someone else move forward. You moving forward to make a right turn triggered him to move forward. Doesn't make it any less irritating, but it probably wasn't "I'm so big, I can block his view if I just move forward a bit".
I think the lack of effective security was all explained in that unaired episode, "The Revenge of RMS".
It should be pretty easy to check for. Remember all those "college music stores" that were reporting loss of sales due to "piracy by college students using Napster on high speed Internet access provided on campus"? Then all those schools that locked down Napster? Any figures on sales at those stores compared to sales where the networks weren't locked down? I'd think if there was any indication to support their contention, the RIAA would have trumpeted such figures as widely as possible.
I've always used the alternate old adage, "If the only problem you have is a nail, all your tools start to look like hammers."
Not only that, when I'm there I want my watch to let me know how long it is until lunch time. I don't want to have to remember that lunch is now at 20:00 instead of the usual 17:00 I'm used to, or that I should set my alarm clock to 16:00 instead of 14:00 when I want to make a morning meeting.
Also, Linux originally started out as a terminal emulator which Linus wrote, in part to learn how interrupts worked in the '386. It grew from there. Minix was just the development environment that Linus used to write, compile, configure and write to floppy disk for booting. He also used the minix file system format originally (which is still available, last I looked).
Reading his letter, where he says that OpenServer 6 costs $599-1399 one-time and you never have to pay support costs again (since he's comparing it to support costs for Red Hat), then goes on later to boast about how good their tech support is, he never says how much you have to pay for that 24x365 tech support. If all you want is what that one-time price gives you ("license, software fixes, and access to their online knowledge base", then why are you buying a support contract from Red Hat? Just download it from the Internet or buy a CD or even buy a "Dummies Guide to Fedora" which includes a CD, and off you go.
He does have a few good points. Point 1 ("Linux isn't free") isn't one of them.
Point 2 ("Kernel is better than Linux") is debatable - but should be easy to show in a shoot-out. Check memory management, forking processes, creating and destroying threads, file system performance, network performance, context switching performance - all running on identical machines (probably on several configurations), and each one as installed, and as tuned to do the best it can.
Point 3 (better security) has meaningless stats. You'd need to compare similarly managed machines with similar tasks and network infrastructure, and the stat given (the ridiculously over-precise 65.64%) is meaningless without comparative number of systems.
Point 4 has some validity - as a single company, they can probably put out a much more definitive roadmap. They can also just as easily change direction rendering the old roadmap meaningless.
Point 5 (backwards compatibility) is also somewhat valid (I'm taking them on their word that they really do attempt and succeed at it) - but a Linux installation can be just as "backwards compatible" as you need, in several ways: kernel interfaces rarely change old behaviors, libraries can have multiple versions (and dependency tracking handles it), most things are in source and can be recompiled if needed. If you need to do regression testing on your applications every time the system changes, then don't change it very often. Just because you CAN update things on a daily basis doesn't mean you should. And "who is verifying ..."? Why, the distributors are!
Point 6 (So You Don't Want To Be An OS Expert) is irrelevant - if you need to buy support, buy it - installing a standard commercial distribution requires the same level of expertise as installing an SCO system.
Point 7 (Warranteed) is just FUD - if you need a warrantee, buy your distribution from someone who will give you one. Maybe you can then argue on price, but don't confuse the support you get on a free download with the support you get when you buy a distribution or a support contract. For example, how long a warrantee period do you get when you buy OpenServer 6 from SCO for $599 without any further support contracts?
Point 8 (No Forks) just barely has a valid point: if you want or need a system that is certified to meet certain standards, then buy a distribution that does so. Linux-the-Kernel and Linux-the-tools and Linux-the-desktop aren't scattering all over the place - there are alternate choices for all sorts of things, and maybe even a few of those alternates are forks, but the system as a whole is pretty well "unified".
Point 9 (reliability) has no evidence to support it - plenty of Linux installations are "locked in a closet and hide the key". It's true that a company's reliability is valued by customers, which is why Red Hat and others have an incentive to be reliable. No different from SCO.
Point 10 (Customer Support) is once again pure FUD - I don't know what reputation various Linux support companies have compared to SCO, but there's no reason they wouldn't have support engineers who know how things work every bit as much as SCO engineers (unless SCO has K & R working for them now), and that's not even getting into the possibility of getting help without needing a support contract.
Whew, this wasn't going to be a point-by-point critique, just a comment on the "Linux isn't free" thing.
Just as some VCRs pick up the program information and some don't, some will pick up time from a station (often PBS) and some won't. I don't know if the protocol includes a DST flag or if the unit is supposed to figure out when to switch on its own.
My ReplayTV doesn't have any option to set the time, only the time zone. Again, I don't know if the time protocol it uses each night when it connects to the server includes a DST flag, nor what it is supposed to do if it doesn't talk to the server on the day before the switch (if it does have a DST-pending flag). All I'll be able to do is change timezones for the three weeks before and 1 week after the automatic switch, if it doesn't compensate properly. It may just be simpler to turn off DST and leave it set to a different timezone for the entire time. I'm not sure if US units will have a way to adjust like that for people already in EST, though - what do you switch it to? I'm also not sure if the schedule info is stored in UCT or "local time" - if in UCT, then I don't have to worry about it not recording the right show, I just get confused by the time displays. I'd guess it stores everything internally by UCT, or it would get very confusing in the fall when the same time repeats itself.
It should be obvious that I already know that, so what point are you trying to make?
Java's bloat doesn't seem to be that much. I opened up a fresh copy of Safari, and loaded the current time page from time.gov (e.g. for my time zone). Real memory usage was about 13.5MB (about 161 MB virtual). Loading the Java version of that page boosted that to about 30MB (383 MB virtual). What's interesting about that page is that they use Java only for doing the polling of the current time from the server, while doing the display updating (including figuring out the light/dark display on the map) in Javascript.
By comparison, with two windows, each with several tabs, and with the JVM already initialized, I'm using a total of 188MB real memory and 630MB of virtual - and now it's also taking up about 25% of my CPU on this Slashdot response page (either from the text box, or from the animated GIF, I'm not sure), even without doing any typing. Typing into it brings it up another 5%. 5 Freaking Percent just to process my keys and plot some text! How ever did we get anything done on 1MHz 8-bit processors? It would take about 100 of them just to do the processing needed to type some text and word-wrap it. Now it's running at 70% CPU for some reason. Crazy. Running Google Maps, scrolling around or zooming in or out with the satellite view uses about 50% CPU. Just loading the map brings real memory use up above that needed for loading Java (though only increasing virtual by.about 20MB), though that obviously grows as you scroll around and it caches more of the map.
Let's see, Macs don't matter because they are such a small market share, but Linux and BSD are suddenly so important that Java not being available on them matters (because somehow it is too difficult to install for some reason, even though it is available as an RPM, and available as source, so anyone could build a package for any distribution; I can see why Debian wouldn't include it, the SCSL is a little bit too controlling, but they could include it as part of non-free, for instance). So how is it not Microsoft's fault that the only platform that really matters, Microsoft Windows, doesn't support it properly?
And yes, I've installed Java, from source and from binary packages. I don't recall any particular problems with it.
Mac OSX also comes with Python, PERL and Tcl (though not Tcl/Tk, for some reason - but there's a very easy install of that as well - I particularly recommend the "Batteries Included" release, it comes with almost every extension you could ever want already loaded and ready to go), not to mention the usual suite of Unix tools (awk, sed, shells, grep, locate, find, xargs, etc.).
So that basically supports my thesis that the reason AJAX is being used is that a better solution isn't available, on of those better solutions being Java, because Microsoft has torpedoed anything that could threaten their hegemony of the desktop. Look for IE7 to add "compelling new features" that, if you use them, make you incompatible with anyone else, and also subtly start changing things so that developers without a clue will "fix" AJAX applications the easy way, breaking them totally for any other browser.
Why would Java on Mac be so much faster than anything else? I just can't fathom a JVM that is so slow that a Javascript interpreter can be faster. 165K of Javascript (Google Maps) does parse and execute "fast enough" on a fast enough computer, but I'm reminded of the saying that with enough power, even a brick can fly.
As for "what do you fall back to", I'm saying that the existence of a JRE on a machine "ought to be" as ubiquitous as "a browser". I mean, what's your fallback if the user doesn't have a browser? I'm tired of having my choices limited by what Microsoft chooses to support. It's great that Firefox is complete enough to be mostly compatible with IE, but as long as Microsoft holds a lock, Firefox won't be able to introduce any new features (at least not any that will be widely used), and will have to track whatever Microsoft does. How is that less proprietary than Sun? And where the heck is Kaffe (I mean that figuratively, I know literally where it is).
I don't recall the exact terms of the agreement between Microsoft and Sun regarding Java - why wouldn't Sun allow Microsoft to include a stock (unmodified by Microsoft) version of Java with a Windows install? I thought the disagreement was with Microsoft making unauthorized changes to the base classes.
Javascript isn't "just there", it is part of a runtime also. My operating system happens to come with that runtime (Safari), along with Java applets (that can run inside or outside a browser), Java Web Start, and Java application support. I'm sorry you're using a backwards OS. If everyone had Tiny BASIC on their machine, would you advocate using that as a common base as well, just because it is there?
Safari loads in less than half a second (fresh boot, so nothing even cached). Initializing Java for an applet take less than 2 seconds for the first time, not perceptible for later times. Applet class files are cached, just like the 165K Google Maps Javascript file is cached. I just don't see the problem. My machine is supposedly not all that fast (1.8GHz G5 with 1.25GB memory).
It's called "indexing". Same way they search the Internet. You didn't think they go out and read all the sites on the Internet when you search for "Britney Spears nude", did you? They do the same thing with your e-mail as it comes in. A decent e-mail client or operating system can do the same thing (see, e.g. Tiger). I haven't used the e-mail client from Tiger, but the one from the previous version does full-text indexing of e-mail, and can do full-text indexing of specified directories, for fast searching.
Google doesn't have magic hardware, they aren't reading through 800MB of your e-mail every time you do a search.
The logistics of moving that many people isn't overwhelming. There are over 650 million "enplanements" per year in the United States alone. 10 years at that rate would move all of the people currently here. In 10 years time, by some projections, world population will be less than 1 billion more than it is now - so about 2 more years would finish the job off.
As far as energy required, it would be large, but (assuming perfect efficiency, and only talking about lifting ~180 lb people, not including luggage, supplies or vehicles) about 3% of current worldwide electrical energy generation capacity (3600 GW) would keep up with that rate (at 24 hours/day, 365 days/year). Obviously, we'd need a more efficient method than rockets to even approach that, but it means it isn't impossible.
Of course, to merely remove a "significant" number of people would take much less. The above was for 100% removal. 25% of that rate is still twice the current worldwide population growth rate, and it would still only take about 60 years to remove the entire population.
Doesn't mean its practical, or that it is desirable, but it certainly isn't impossible (given reasonable advances in technology over the next 50 years).
Yup, really impressive, especially the rotating cube with a different running movie clip on each face, all on a 200MHz PPC.
Why do you say that "no significant portion of the population will ever move off-world"?
Yes, it should be clear from my comments about using Safari that I have a Mac. On a 1.8GHz G5, which is apparently now not considered very fast. I've used Java applications, delivered over a network, and they were quite acceptable even on slow machines. I found Java speed to be reasonable even on a machine so slow that I could type about 10-15 times faster than Mozilla could accept text into a text box (a 200MHz PPC).
Why doesn't Gmail use Java? I don't know, possibly because it isn't the correct buzzword anymore. Does IE on Windows machines come with Java installed? One that works properly and is reasonably recent?
The only reason more people don't have Java easily available is because of Microsoft. I don't see that as a good reason to try to shoehorn the Web model and a hodgepodge language into something it isn't very good at.
There is absolutely no reason that Java applications can't be delivered and updated every bit as easily as downloading gobs of Javascript on every page. I don't understand the comment about Applets taking forever to load - bytes is bytes, Java bytecode is not much different in size than the source code it came from. Javascript isn't some radically different language where you can magically do really complex things in enormously less code than anything else. I certainly don't find that Applets take forever to load.
Versioning issues with regard to a JVM (and, more importantly, the standard class libraries expected to be there) is not more of a problem than versioning issues with different browsers, and there's no reason that different implementations of Java would be more incompatible than different implementations of Web browsers and Javascript.
Why do you need to target 1.5? If you really do need 1.5, then I don't see how Javascript in a Web browser is going to be adequate.