Each vendor should release a Java VM that they optimize using their own compiler, hand coded assembler and any dirty trick they can think about, including hardware implementation of bytecode. If I can then take real-life Java apps and they actually run faster on a given CPU, the results are accurate - at least for Java apps.
With this method, I don't have to recompile every app used in the benchmark for each CPU and can be still sure that their execution is optimized in every way the vendor can think about.
is how grayscale Palms have always displayed shades of gray. The LCD driver will turn pixels on and off on alternative frames. Under some conditions, the flicker becomes visible and then the grayscale screens look pretty annoying. But anyway, it's still a hardware solution so you can kind of claim it's part of your color support. Dithering of cause is another story.
In any case, nobody buys a Palm for a nice-looking screen. That's what a CLIE is for.
When you are fast-forwarding through a particular commercial it's your personal choice. In this case, a link directly to MP3 could misled people who do not know you want them to buy stuff in order to support the artist. A better analogy is posting a commercial-free TV episode on the web and not telling people that it originally had commercials.
It probably still costs less than 37 cents for ISP to deliver a single e-mail. Otherwise none of $19.95 per month ISPs would survive, given all the spam we get. So if some of political campaigning could be moved to e-mail we can save some government money to be used for more productive activities.
The real problem is how the e-mail address list was obtained - either through sites that didn't warn users that their e-mail will be used for marketing or through ones that tricked people with checkboxes set by default. Not too many people go to buy something on the web and agree to get political messages.
I think the real solution is to encourage people to fill in their e-mail address when they register to vote. If they do, they agree to recieve a limited number of e-mail messages (let's say, 3 per candidate for a given election). There will be a convinient way for candidates to check the database so that they can save money by not sending a dead-tree flyer to these people. Candidates will be publically ostricized for not checking the database and wasting government/donation money or for exceeding the limit on the number of messages.
Let's say I want to offer something on my web site, but I want everyone to go through my front page to get it. Say I am giving away my music and want a chance to sell my CDs and T-shirts before people download MP3. If people make links directly to all my MP3s, I will loose my chance to earn money. I could serve everything with CGIs that generate one time URLs, but maybe I just don't have the money to buy a fast dedicated box I would need to serve large files through scripts. I might also not have enough money to hire lawyers and get you to remove the links in this way.
Sure, there are technical solutions to unwanted (by the author) copying, security and so on. But let's say there is a guy who wants to sell t-shirts and offer his MP3 files on an unpatched IIS NT4.0 server. Let's say his front page has a sign saying that he can make the thing more secure, but then he'll have to charge (more) money to pay IT consultants and hosting services. Shouldn't we try to respect his wish and just link to the poor guy's front page?
I haven't met anyone who enjoyed studying in grade school and the reason is pretty simple. We are only eager to study if it makes us happy right away or at least we can see how it will help us eventually. Memorizing state capitals or hundreds of exact days in the history class or proving an abstract theorem
just makes us feel that everything else we learn must be useless too. An ideal education system will teach skills according to a hierarchy of human needs, starting with the most basic ones and only at the end teaching the disposable extras. The only catch is, each skill should be taught only when a kid can understand why it might be important. The hierarchy itself would vary between countries, but in US it's something like that
Very basic job skills - enough to get hired in safeway and avoid being homeless. Personal hygeine, respect for your manager, showing up on time, looking for cheap housing etc.
Dating. Sure, slashdot crowd is suffering from lack of skills in this area in particular, but for most people that's what makes difference between just surviving and real happiness. Meeting people, balance of getting what you want and making sure the other person is happy too, raising children, handling a divorce if it has to happen, etc, etc. Even a very happy relationship could be more happy in some area and that's the happiness that counts most. Teaching something that might help at an age when you can understand it would be super cool.
After basic survival and interperson skills are taught, students should be offered a sampling of subjects that are currently in hot demand or are likely to be soon. If they happen to have a liking/talent for one of these subjects, they should be taught skills that would help them land a job that pays well.
Finally, students should be taught how to be good citizens in a larger society beyond their family - volunteering, donating to charities, conservation and so on. Of course it's important but people are not likely to do it until they are satisfied with their own life.
Only after someone has mastered all these things, it makes sense to teach abstract enjoyment of math, physics and so on. For most people, this will only happen after 30. There is a small minority though which is critical to the society. A reasonable number of people will trade high-paying job for the one they intelectually (or physically for that matter) enjoy. This is a large percentage of slashdot I guess. Several people I know will give up dating to do what they are doing. And really gifted people will rather be hungry than give up their quest. It's important that the school keeps the door open for this kind of people, because they will contribute much more to the society than most of us who are happy to sip a beer and sit on the couch with a date. Perhaps offer small samples of each subject and encourage students to study ones that fascinate them. But it's equally important not to torture the majority of us that just think about a date and a beer (hopefully in this order) and expect us to enjoy the stupid math.
It appears that Andrew has no objections to using physical or electronic violence against annoying slashdot posters. So lets hack slashdot IP logs+his ISP registeration databases for real life address and go swat him on the ass a couple of times. For the good of society of course.
For those reading while humor-inpaired: it's a joke of course. People just trolling about beating someone only deserve verbal abuse. But if you actually see someone mugging a cell-phone user or hitting their kid... well you have all the right to remind them how it feels like.
It depends on what you are doing. If you are not out to make big money and just want some mathematical recreations on an $5 table, they pretty much ignore you. In Tahoe, there are still single-deck tables where they go most of the way through the deck and the dealer will even agree to reshuffle if too many aces are gone in the beginning (yeah, theoretically the deck can still turn favourable, but it's too boring to play with no chance to get a blackjack).
Because basically it's equivalent to a plain HTTP connection. Even if you already have a valid key for a site stored, after I spoof it I can just send you a new one and you'll probably accept it as well. Even for the first use, the risk of attack may be small for you but not for the website owner that may see a lot of users sign up and be owned during a few hours when it's spoofed.
Sorry, but I am not ordering from you with my credit card if you saved $150 rather than give me at least minimum assurance you are who you claim to be. At least Verisign has your credit card or bank account information. It could be that we need better CAs, ability to block ones we don't trust and at least one of them that issues free keys for non-profit sites that need security. But in the meantime, I would be happier if any well known organization can vouch that you are who you claim to be - even if you use CAs such as Microsoft, RIAA and Church of Scientology.
Well, I look at how it's not valid. If it's just expired recently or it's for ssl.bankamerica.com rather than www.bankamerica.com I will let it go. If it's not signed by CA or is for bankam3r1ca.com, I'll close the window right away, in case I have any ActiveX controls signed by the real site that talk to the server and perhaps can be spoofed to do interesting things. Hmm.. What were the websites where you order stuff again?
Just announce and prove your discovery without telling how you did it. Then if it's only a matter of time because other people duplicate your work - even Fermat's last theorem was eventually proven. If you really think it can not be duplicated - say an alien shuttle crashed in your backyard - leave some hints for others to follow. Or encrypt it in a way that currently will take 1M years to break so that there is a second way it can be discovered - either by duplicating your work or by building a radically faster computer.
What if all open source alternatives suck? I don't want to be the one browsing a government website that stores my tax information in MySQL database and let's me view it through PHP frontend that generates PostScript reports. With DB2/J2EE/PDF I can be more sure that my data doesn't get lost, front end doesn't crash and I can find some place to view/print the report.
It would be better to let government to choose any software based on total cost (purchase+support) and the quality of service to the public. We can still require the vendor to write a full-featured, open-source client that citizens can use to interface with government's server. For example, one can either generate web pages to support Mozilla or write their own open source browser to support in addition to IE. It's pretty clear what most companies would rather do.
Well, if someone tried to break into my house and is now wondering about neighborhood and trying to break into other homes at random? What if those of my neighbors that have not been inoculated are getting bitten by that guy and also turning into blood-thirsty zombies?
I guess your comment makes sense when a crime is over and authorities can step in on time to prevent additional crimes. This is pretty unlikely on Internet. For Code Red, the worm will keep trying to attack hosts in my organization whereas by automatically patching every attacking node I can put a stop to it. For DOS attack in progress, I am still suffering from a "crime". And if someone just stole my credit card list, well if I can hack into that machine and delete it, it's much better than wait until it is posted on an IRC channel.
Also remember that if there is no danger to life from hacking, it also applies to my counter-attack, as long as I limit myself to stopping the crime in progress rather than doing format c:. If I am wrong I can just say sorry and your machine will be left with little damage.
If they can sell several hundred phones to the same geeks who by CLIEs and then all the things actors do will happen naturally for free and with less potential embarassment (what if real phone owners will be harassed in a bar because people think they are advertisers?). If they can not even sell a few by putting them in Fry's, ad campaign will not do much good either
IBM used to make 2.88M drives that could be tortured to write 4M. It seams that there must be some technology advanced to write at least twice more these days. A drive that can write, say, 8M to existing floppies and perhaps more on new ones with the same form factor will be extemely useful. The problem with CF cards is that they are easy to loose and also too expensive for me to just give you one with something and forget about it. CD-RW comes close but has annoying problems (need a carrying case, slow to write small amounts of data, expensive drives, uncertain lifetime).
In the old times companies just payed workers in coupons that could only be used to buy stuff in the company's store, where they of course jacked up the prices. People took an issue with that and it doesn't happen anymore. We don't trust an arbitary private company to print it's own currency when even elected governments can be barely trusted not to abuse it by causing inflation or (in some countries) taking over sectors of economy and doing the same thing as old-time companies.
Entertainment products are in no way a currency. They don't have any stable, universal value. After I watched a movie, there is nothing I can do with the ticket stub. They are exactly what the name implies - products and services. But anyway, society generally controls both what government does with currency and what companies must let you do with their products. If you buy a car, you want to be able to drive anywhere you want, disassemble it to make improvements, sell it and so on. True, it a crime for me to steal a car from dealer's parking lot but it's also the same crime for him/her to steal it from mine after I payed for it. It's also Ok for me to look at someones car and, should I have the expertise, make a similar one for myself.
What enterntainment industry is asking for though is a dictatorship, power without control. It will hardly agree to be controlled by public like a government - we will not get to elect CEOs of Hollywood companies. It also hasn't agreed to the same restrictions as makers of physical products. I can not get my money back if Pressplay deactivates my music collection. There is no reason for us to cooperate and give them such a priviliged position.
Seriously though, this gadget could be useful in remote areas of the world. Maybe some village without electricity can get one to listen to weather forecasts, report emergencies and so on. Most people can do better with an AA cell phone adapter though.
Come on, how are they going to control spread of the source developed by 18 universities? It's far more likely that they just don't want billions of dollars to flow out of the country and into the hands of Microsoft. Or make warez an offical policy. I bet students will mostly do the work for free and anyway they'll start on top of Linux and don't have to pay developers even a fraction of US salaries. It's a good idea for any country, even US. I doubt that the number of jobs Microsoft creates justifies all the cash they pocket away.
I work for an unnamed big company that has converted to clearcase. On the other hand, I invented various excuses to keep CVS for my group. The other people on my floor are not having fun. They have to check out every file they change in advance rather than just letting source control figure out what they have changed. Then they have to check in each file a lot of times - to their private branch, then to shared branch and other places I don't understand. To merge the file, they have to look at some mind-boggling chart with dozens of boxes and prolifiration of arrows. The central server constantly goes down and for a few days they get to do source control with floppies. When it is running, ct co -nc filename takes anywhere from 2 seconds to 1 minite. We never had a fulltime source control admin when we used RCS - I just used to write a new shell script once in a few month when they wanted some new feature. Now we have a fulltime admin and even people who used clearcase for 3 years talk to him every few days. In my life, I only saw a few pieces of software that were so screwed. This one has earned a place in my heart (or somewhere) right next to RealOne player, DOS 4.0, Solaris 2.1 and Thexder 95. If they make me move to this thing, I'll just set up a shell script to zip my cvs repository every week and check it in to clearcase.
Then they should hire programmers to fix it for them. Just because people gave you a mug of free beer doesn't mean you don't have to buy another one later or brew your own.
If they are trying to avoid copyright lawsuits, they are actually making it worse for themselves. By censoring my online communications, they also assume responsablity if I send hate mail, download warez and so on. On the other hand, if they are worried about bandwidth - well why would people get high-speed access if they were not going to use bandwidth? I bet most customers will at least occasionally download audio or video. They can cap the total bandwidth and document the limits but it has nothing to do with what exactly I am doing - sharing files or videoconferencing.
I use several drivers that do not come with default RH7.3 (Lucent winmodem, NVIDIA and NTFS) and even installing official RPMed updates is a pain. For example, I need to keep gcc-2.96 around - 3.1 causes Lucent driver to reboot the machine. I gave up some other patches, like Zaurus because they always caused panics. As for development 2.5 kernels, the number of config options is bewildering (can I just install the newer sound subsystem or do I need both), some defaults really suck (like no keyboard support) and every time I tried I got compile errors.
Back in the days of 1.1.x, I actually installed every release for fun to see what's new. Now, I could be running Windows for all I care, since I need to wait for someone's binary updates. I guess building kernels is now just for some elite club of core developers and distribution builders who know what magic version of compiler to use, which modules can compile cleanly, what to patch vs kernel.org etc.
The only way out I see is to split kernel into many components with well defined interfaces that do not change often. kernel.org should have just a bootstrap kernel. Let's say you download one file with a "core" source and one with x86 basics such as EIDE and VESA SVGA. Then you can go to, say, kernel-audio.org and download your sound subsystem and a sound blaster driver. Or you should be able to use the old binary ones you already installed if you are satisfied with your sound.
Someone will say that fixed interfaces for drivers will hurt efficiency and increase memory footprint (because a kernel might need to support several versions of the interface to let older drivers work). Yet the same is true for every other core system component (X-windows protocol, libc binary interface, PPP protocol) and people are always asking for more modularity, not less. It's better to have a system which is 30% slower, but that let's you watch TV on your tuner card, network with your CLIE and your Zaurus and output sound to your digital speakers (all these drivers are much more likely to be written, especially by vendors, if they can be compiled once and continue working for long time). I am sure current design is more fun to hack for ultimate efficiency (apparently so fun there are 3 different branches of the same kernel on kernel.org), but it's really not fun for the rest of the people to customize.
If you write free software you either do this for fun or want to make money on something else, like support, other products or hardware. Just not on people using this particular software. Otherwise, you should charge for software and maybe give source to registered users (Minix used to be like this for long time). Ads by definition are designed to distract me from what I am doing and make me look at them. When I am browsing the web, I am already distracted, so it doesn't matter. But I don't want a computer that distracts me from work, action games or writting serious e-mail. I would rather have one with proprietory software that allows core OS to remain free. I would love if Microsoft released a Linux distro with, let's say, open source VB and.NET and then made money by selling office on top of it. By the way, if I contribute some small piece to Linux, ad-supported desktops are unfair for me because basically people are still using my work to get money.
Because there is no documentation telling you how to write web pages for all the "standard" browsers equally easily - with nice, dynamic cross-reference of objects and methods. Eiether that, or maybe there is no way to do this because even a simple script written to standard will run into a lot of bugs. Or w3c model is much more complicated than IE. Anyway, I don't think people would deliberately ignore standards if they were easy to use. I would guess most people use open() instead of CreateFile when writting Win32 programs.
Slashdot Mirror
With this method, I don't have to recompile every app used in the benchmark for each CPU and can be still sure that their execution is optimized in every way the vendor can think about.
Come on, a standard to prevent wardriving called PEEP? Sounds like another product that will live up to Microsoft reputation for security.
In any case, nobody buys a Palm for a nice-looking screen. That's what a CLIE is for.
When you are fast-forwarding through a particular commercial it's your personal choice. In this case, a link directly to MP3 could misled people who do not know you want them to buy stuff in order to support the artist. A better analogy is posting a commercial-free TV episode on the web and not telling people that it originally had commercials.
The real problem is how the e-mail address list was obtained - either through sites that didn't warn users that their e-mail will be used for marketing or through ones that tricked people with checkboxes set by default. Not too many people go to buy something on the web and agree to get political messages.
I think the real solution is to encourage people to fill in their e-mail address when they register to vote. If they do, they agree to recieve a limited number of e-mail messages (let's say, 3 per candidate for a given election). There will be a convinient way for candidates to check the database so that they can save money by not sending a dead-tree flyer to these people. Candidates will be publically ostricized for not checking the database and wasting government/donation money or for exceeding the limit on the number of messages.
Sure, there are technical solutions to unwanted (by the author) copying, security and so on. But let's say there is a guy who wants to sell t-shirts and offer his MP3 files on an unpatched IIS NT4.0 server. Let's say his front page has a sign saying that he can make the thing more secure, but then he'll have to charge (more) money to pay IT consultants and hosting services. Shouldn't we try to respect his wish and just link to the poor guy's front page?
- Very basic job skills - enough to get hired in safeway and avoid being homeless. Personal hygeine, respect for your manager, showing up on time, looking for cheap housing etc.
- Dating. Sure, slashdot crowd is suffering from lack of skills in this area in particular, but for most people that's what makes difference between just surviving and real happiness. Meeting people, balance of getting what you want and making sure the other person is happy too, raising children, handling a divorce if it has to happen, etc, etc. Even a very happy relationship could be more happy in some area and that's the happiness that counts most. Teaching something that might help at an age when you can understand it would be super cool.
- After basic survival and interperson skills are taught, students should be offered a sampling of subjects that are currently in hot demand or are likely to be soon. If they happen to have a liking/talent for one of these subjects, they should be taught skills that would help them land a job that pays well.
- Finally, students should be taught how to be good citizens in a larger society beyond their family - volunteering, donating to charities, conservation and so on. Of course it's important but people are not likely to do it until they are satisfied with their own life.
Only after someone has mastered all these things, it makes sense to teach abstract enjoyment of math, physics and so on. For most people, this will only happen after 30. There is a small minority though which is critical to the society. A reasonable number of people will trade high-paying job for the one they intelectually (or physically for that matter) enjoy. This is a large percentage of slashdot I guess. Several people I know will give up dating to do what they are doing. And really gifted people will rather be hungry than give up their quest. It's important that the school keeps the door open for this kind of people, because they will contribute much more to the society than most of us who are happy to sip a beer and sit on the couch with a date. Perhaps offer small samples of each subject and encourage students to study ones that fascinate them. But it's equally important not to torture the majority of us that just think about a date and a beer (hopefully in this order) and expect us to enjoy the stupid math.For those reading while humor-inpaired: it's a joke of course. People just trolling about beating someone only deserve verbal abuse. But if you actually see someone mugging a cell-phone user or hitting their kid... well you have all the right to remind them how it feels like.
It depends on what you are doing. If you are not out to make big money and just want some mathematical recreations on an $5 table, they pretty much ignore you. In Tahoe, there are still single-deck tables where they go most of the way through the deck and the dealer will even agree to reshuffle if too many aces are gone in the beginning (yeah, theoretically the deck can still turn favourable, but it's too boring to play with no chance to get a blackjack).
Because basically it's equivalent to a plain HTTP connection. Even if you already have a valid key for a site stored, after I spoof it I can just send you a new one and you'll probably accept it as well. Even for the first use, the risk of attack may be small for you but not for the website owner that may see a lot of users sign up and be owned during a few hours when it's spoofed. Sorry, but I am not ordering from you with my credit card if you saved $150 rather than give me at least minimum assurance you are who you claim to be. At least Verisign has your credit card or bank account information. It could be that we need better CAs, ability to block ones we don't trust and at least one of them that issues free keys for non-profit sites that need security. But in the meantime, I would be happier if any well known organization can vouch that you are who you claim to be - even if you use CAs such as Microsoft, RIAA and Church of Scientology.
Well, I look at how it's not valid. If it's just expired recently or it's for ssl.bankamerica.com rather than www.bankamerica.com I will let it go. If it's not signed by CA or is for bankam3r1ca.com, I'll close the window right away, in case I have any ActiveX controls signed by the real site that talk to the server and perhaps can be spoofed to do interesting things. Hmm.. What were the websites where you order stuff again?
Just announce and prove your discovery without telling how you did it. Then if it's only a matter of time because other people duplicate your work - even Fermat's last theorem was eventually proven. If you really think it can not be duplicated - say an alien shuttle crashed in your backyard - leave some hints for others to follow. Or encrypt it in a way that currently will take 1M years to break so that there is a second way it can be discovered - either by duplicating your work or by building a radically faster computer.
It would be better to let government to choose any software based on total cost (purchase+support) and the quality of service to the public. We can still require the vendor to write a full-featured, open-source client that citizens can use to interface with government's server. For example, one can either generate web pages to support Mozilla or write their own open source browser to support in addition to IE. It's pretty clear what most companies would rather do.
I guess your comment makes sense when a crime is over and authorities can step in on time to prevent additional crimes. This is pretty unlikely on Internet. For Code Red, the worm will keep trying to attack hosts in my organization whereas by automatically patching every attacking node I can put a stop to it. For DOS attack in progress, I am still suffering from a "crime". And if someone just stole my credit card list, well if I can hack into that machine and delete it, it's much better than wait until it is posted on an IRC channel.
Also remember that if there is no danger to life from hacking, it also applies to my counter-attack, as long as I limit myself to stopping the crime in progress rather than doing format c:. If I am wrong I can just say sorry and your machine will be left with little damage.
If they can sell several hundred phones to the same geeks who by CLIEs and then all the things actors do will happen naturally for free and with less potential embarassment (what if real phone owners will be harassed in a bar because people think they are advertisers?). If they can not even sell a few by putting them in Fry's, ad campaign will not do much good either
IBM used to make 2.88M drives that could be tortured to write 4M. It seams that there must be some technology advanced to write at least twice more these days. A drive that can write, say, 8M to existing floppies and perhaps more on new ones with the same form factor will be extemely useful. The problem with CF cards is that they are easy to loose and also too expensive for me to just give you one with something and forget about it. CD-RW comes close but has annoying problems (need a carrying case, slow to write small amounts of data, expensive drives, uncertain lifetime).
Entertainment products are in no way a currency. They don't have any stable, universal value. After I watched a movie, there is nothing I can do with the ticket stub. They are exactly what the name implies - products and services. But anyway, society generally controls both what government does with currency and what companies must let you do with their products. If you buy a car, you want to be able to drive anywhere you want, disassemble it to make improvements, sell it and so on. True, it a crime for me to steal a car from dealer's parking lot but it's also the same crime for him/her to steal it from mine after I payed for it. It's also Ok for me to look at someones car and, should I have the expertise, make a similar one for myself.
What enterntainment industry is asking for though is a dictatorship, power without control. It will hardly agree to be controlled by public like a government - we will not get to elect CEOs of Hollywood companies. It also hasn't agreed to the same restrictions as makers of physical products. I can not get my money back if Pressplay deactivates my music collection. There is no reason for us to cooperate and give them such a priviliged position.
Seriously though, this gadget could be useful in remote areas of the world. Maybe some village without electricity can get one to listen to weather forecasts, report emergencies and so on. Most people can do better with an AA cell phone adapter though.
Come on, how are they going to control spread of the source developed by 18 universities? It's far more likely that they just don't want billions of dollars to flow out of the country and into the hands of Microsoft. Or make warez an offical policy. I bet students will mostly do the work for free and anyway they'll start on top of Linux and don't have to pay developers even a fraction of US salaries. It's a good idea for any country, even US. I doubt that the number of jobs Microsoft creates justifies all the cash they pocket away.
I work for an unnamed big company that has converted to clearcase. On the other hand, I invented various excuses to keep CVS for my group. The other people on my floor are not having fun. They have to check out every file they change in advance rather than just letting source control figure out what they have changed. Then they have to check in each file a lot of times - to their private branch, then to shared branch and other places I don't understand. To merge the file, they have to look at some mind-boggling chart with dozens of boxes and prolifiration of arrows. The central server constantly goes down and for a few days they get to do source control with floppies. When it is running, ct co -nc filename takes anywhere from 2 seconds to 1 minite. We never had a fulltime source control admin when we used RCS - I just used to write a new shell script once in a few month when they wanted some new feature. Now we have a fulltime admin and even people who used clearcase for 3 years talk to him every few days. In my life, I only saw a few pieces of software that were so screwed. This one has earned a place in my heart (or somewhere) right next to RealOne player, DOS 4.0, Solaris 2.1 and Thexder 95. If they make me move to this thing, I'll just set up a shell script to zip my cvs repository every week and check it in to clearcase.
Then they should hire programmers to fix it for them. Just because people gave you a mug of free beer doesn't mean you don't have to buy another one later or brew your own.
If they are trying to avoid copyright lawsuits, they are actually making it worse for themselves. By censoring my online communications, they also assume responsablity if I send hate mail, download warez and so on. On the other hand, if they are worried about bandwidth - well why would people get high-speed access if they were not going to use bandwidth? I bet most customers will at least occasionally download audio or video. They can cap the total bandwidth and document the limits but it has nothing to do with what exactly I am doing - sharing files or videoconferencing.
I use several drivers that do not come with default RH7.3 (Lucent winmodem, NVIDIA and NTFS) and even installing official RPMed updates is a pain. For example, I need to keep gcc-2.96 around - 3.1 causes Lucent driver to reboot the machine. I gave up some other patches, like Zaurus because they always caused panics. As for development 2.5 kernels, the number of config options is bewildering (can I just install the newer sound subsystem or do I need both), some defaults really suck (like no keyboard support) and every time I tried I got compile errors. Back in the days of 1.1.x, I actually installed every release for fun to see what's new. Now, I could be running Windows for all I care, since I need to wait for someone's binary updates. I guess building kernels is now just for some elite club of core developers and distribution builders who know what magic version of compiler to use, which modules can compile cleanly, what to patch vs kernel.org etc. The only way out I see is to split kernel into many components with well defined interfaces that do not change often. kernel.org should have just a bootstrap kernel. Let's say you download one file with a "core" source and one with x86 basics such as EIDE and VESA SVGA. Then you can go to, say, kernel-audio.org and download your sound subsystem and a sound blaster driver. Or you should be able to use the old binary ones you already installed if you are satisfied with your sound. Someone will say that fixed interfaces for drivers will hurt efficiency and increase memory footprint (because a kernel might need to support several versions of the interface to let older drivers work). Yet the same is true for every other core system component (X-windows protocol, libc binary interface, PPP protocol) and people are always asking for more modularity, not less. It's better to have a system which is 30% slower, but that let's you watch TV on your tuner card, network with your CLIE and your Zaurus and output sound to your digital speakers (all these drivers are much more likely to be written, especially by vendors, if they can be compiled once and continue working for long time). I am sure current design is more fun to hack for ultimate efficiency (apparently so fun there are 3 different branches of the same kernel on kernel.org), but it's really not fun for the rest of the people to customize.
If you write free software you either do this for fun or want to make money on something else, like support, other products or hardware. Just not on people using this particular software. Otherwise, you should charge for software and maybe give source to registered users (Minix used to be like this for long time). Ads by definition are designed to distract me from what I am doing and make me look at them. When I am browsing the web, I am already distracted, so it doesn't matter. But I don't want a computer that distracts me from work, action games or writting serious e-mail. I would rather have one with proprietory software that allows core OS to remain free. I would love if Microsoft released a Linux distro with, let's say, open source VB and .NET and then made money by selling office on top of it. By the way, if I contribute some small piece to Linux, ad-supported desktops are unfair for me because basically people are still using my work to get money.
Because there is no documentation telling you how to write web pages for all the "standard" browsers equally easily - with nice, dynamic cross-reference of objects and methods. Eiether that, or maybe there is no way to do this because even a simple script written to standard will run into a lot of bugs. Or w3c model is much more complicated than IE. Anyway, I don't think people would deliberately ignore standards if they were easy to use. I would guess most people use open() instead of CreateFile when writting Win32 programs.