NPR just reported this as "vouchers that are redeemable for software and hardware [...] even some products not made by Microsoft". The other reporting, say on News.COM, also suggests that the vouchers will only be redeemable for selected products. And why give "vouchers" if they could be redeemed for anything? Why not just give cash?
It's common in these kinds of settlements for manufacturers to give vouchers for their own products. This is likely no different: you can probably use the vouchers to buy Microsoft software and hardware running Microsoft software. Non-redeemed vouchers will be given as 1/3 in Microsoft software and 1/3 in hardware (presumably, PCs running Microsoft software) to schools.
Even $1.1 billion in cash would be a slap on the wrist. This "settlement" is an insult: it's a marketing promotion for Microsoft and a means by which they can get more of their software into the education market.
I recommend reading the news.com article. Major points:
It's for Windows 95 and Windows 98.
It seems to be focused on business licensees.
It's unclear whether the vouchers are really unrestricted; the formulation "Microsoft or other products" sounds like marketing speak for "selected products by Microsoft and a few other companies".
It appears that for anything that isn't redeemed, Microsoft gets to keep 1/3, give 1/3 in cash, and give 1/3 in Microsoft software to schools.
The kicker is, however, $1bn is about as much as Toshiba had to pay for shipping supposedly defective floppy disk drives on their laptops.
I think this is absolutely evil. Even if Microsoft had to pay everything in cash, it would be peanuts. Instead, they'll be able to further contaminate schools with their proprietary software, something they have already volunteered to do as a "donation"--a tax sheltered marketing ploy.
Sound files are not getting much bigger per minute. Totally uncompressed audio is no more than 5MB/min tops in a format like shn.
That's two-channel audio. You are going to get many more channels (in fact, DVD audio already does, I believe).
Video isn't going to get a heck of a lot bigger than DVD-Video sizes.
Sure it is: DVD resolution is really low. You are first going to see HDTV-resolution DVDs, and later probably 3-6 MPixel video. Beyond that, there will be many simultaneous video streams: half a dozen camera angles from a live event, etc.
Dual licensing works fine for some software, and it doesn't work so well for other software.
First of all, you can really only dual-license if there is something to sell. That means dual-licensed open source software needs to have one of the less "business-oriented" open source licenses, like the GPL. But for some software, the GPL just isn't appropriate. For example, for a GUI toolkit, if it is GPL'ed, that limits its adoption.
Also, with dual-licensed software, companies have other considerations in mind than just creating technically the best possible software or the best software for the users. They may want to try to dominate a particular market niche and exclude competition, for example.
I think in the case of Troll Tech, both of these apply. Companies like Sun and IBM have chosen Gnome because they did not want to hand the keys to commercial applications for their desktops to some small company that they have no control over. Furthermore, Troll Tech has a strong incentive to have something like Qt/Embedded take over the entire display--if people could use free or competing toolkits on something like the Sharp Zaurus, it would threaten their revenue stream.
I think dual licensing can make sense in some circumstances, but I think it's risky. If you are considering using dual-licensed software, I think you would do well to look at the license, business, and track record very carefully. Personally, I think for MySQL it works, but for Qt, it doesn't.
One big deal is that I don't want to give another company my credit card number and personal information. Such subscriptions can be a pain to cancel, and many people let them go on long after they stopped using the service.
Another concern about TiVo is that they get detailed information about my viewing habits and can correlate that with my credit card (whether they do it or not is another matter).
I want to reduce the number of companies that have their fingers in my finances.
I think subscription-based PVR models are really a bit expensive for the consumer in the long run. Tivo may be selling their hardware under cost, but enough to justify the subscription cost? I personally don't believe so.
I REALLY hate to see everyone bashing Microsoft every time they attempt something new
Trouble is: it isn't new. Smart wrist watches are an old idea, as are portable multimedia player boxes.
Microsoft wouldn't be ridiculed if they came up with something genuinely new, or if they did a really good job reimplementing some known idea. But mediocre copies of other people's ideas just invite ridicule; they'll have to deal with it. Gates can at least laugh all the way to the bank: mediocre copies do sell, after all, as many other companies also show us.
Until ZeroConf clients are pre-installed on Linux and easily installed on Windows, people won't have much motivation for installing this on the server. After all, what's the point of using ZeroConf if the first thing you need to do is install and configure a piece of software on the server and every client?
Last I looked, Apple's license on this also made it unclear whether Linux distributions would pick it up. So, widespread deployment may have to wait until GPL or BSD implementations.
That may well be "possible", but it's still multiple defects. First, the OS should ensure that/Users is not writable by non-root accounts. Second, no web browser should replace a directory with a file when downloading.
Safari is a beta software anyway. Use it at your own risk.
Beta software is handed out by companies to get useful feedback from users. It's of benefit to the companies. Users download it out of curiosity, but that only goes so far. Companies that take your attitude will soon run out of beta testers (and probably users).
Beta software may perform poorly, but if anything, it should have extensive error checks compiled into it.
You use any software "at your own risk" (check your EULA).
But beta software is deliberately released to customers for testing. This is for the benefit of the company, not the customers. Companies should go out of their way to make sure their beta software doesn't do anything seriously bad, and blaming the user for trying the software is self-defeating.
If anything, beta software should have lots of extra checking compiled into it; it might quit more often than released software and run slower, but it shouldn't crash in unexpected ways any more frequently, and it should definitely go out of its way to guard against data loss.
Microsoft's media formats are not documented, hence they are less valuable. That's why they charge less. If the MPEG-4 folks think that they can't compete, they should lower their licensing fees. It is really an outrage anyway that MPEG-4 requires licensing fees for its implementations; it's difficult to see what profoundly new ideas are represented by its standards body.
What we have here is two greedy organizations battling it out. If we want to avoid getting dragged into this, we really do need open video standards.
I'm a hardcore geek and have expert certification on everything from Windows 2000 to A+ certs to Novell Network certs to CISCO certs to _____ . You name it, I've done it. But I personally could probably just barely piece one of these "home media units" together. Furthermore, I wouldn't even know what to use it for.
I don't have any of those certifications (thank God!), but I have had no trouble pulling together a "home media unit" from scratch.
A standard Linux install pretty much has all you need, and you have lots of choices for how to set it up--all of them pretty simple. You can set up the box as a streaming media server, or you can make it part of your own in-home P2P network, or you can handle all the music through web interfaces. You can push audio to a Linux box or have it pull it from other systems with standard commands.
Another very simple approach is to get a Macintosh--iTunes pretty much does everything you need for that out of the box.
So, we have more need for 6 disc changes than we do for 10 GB discs of hard drives on which to store mostly-illegally-obtained mp3s.
I own all the CDs for the MP3s that I have. Why store them on-line? Because a computer is much, much more convenient than two 300 CD jukeboxes or, worse, lots of jewel cases and strange looking pieces of furniture.
No one aside from the most hardcore Slashdotter would even know what to do with one of these
My parents seem to have no trouble understanding the convenience of just selecting a CD from an on-screen list, as opposed to dealing with hundreds of jewel boxes.
Of course, little of this applies to this HP device, which does sound much more complex, less functional, and proprietary than just getting a Mini-ITX system.
I suspect that it simply won't enable "protected" functions of the computer: cryptographic keys and the like that you need in order to play/access content that is subject to DRM.
That would still be annoying as hell, but you wouldn't lose any functionality over what you get now; it would just become harder and harder to access things like the next generation of digital audio/video and Windows media from Linux--at least if the big studios and record companies get their way.
Overall, though, I still think that this will just flop: no hardware or software vendor really needs the hassle or additional support costs that result from this.
Allen's "Mini-PC" is about a stale and outdated a concept as they come. Between the Tiqit, Oqo, and the IBM spinout, we only have some of the recent companies trying this. Previously, of course, HP had its palmtops, which, back then, were full DOS machines (yeah).
If you want something today, get a mini laptop from Fujitsu, Sony, or Dynamism.
This looks like a largely proprietary system and a serious security problem. It goes around poking your network and making connections to who-knows-where over the Internet.
For $400, you can get a Mini-ITX system or a WalMart PC and a better set of powered speakers. That also gives you decent amounts of MP3 storage. The result will be more flexible and convenient than this thing.
Publishing has a notion of "archival publication". Traditionally, if several libraries held copies, it was archival. Today, that function is also fulfilled by services like Arxiv.ORG. An archival publication can be cited as such. Nature probably wants to avoid that there are multiple archival copies of a paper.
I'm not sure I like the restriction, but at least I can understand why a traditional publisher or librarian might want to impose it.
The Google cache shouldn't be a problem, and Citeseer shouldn't be a problem either (it doesn't try to be archival, as far as I can tell).
It doesn't matter how much each different job contributes to the bottom line. The reasons why managers get prestige and pay is because they have power over people, and that can be leveraged to get prestige and money. That won't change.
The route to power and better pay for the engineering profession can be (and already is, in many cases) is to work independently and hire out your services. But with that comes a level of risk and entrepreneurship that many technically inclined people don't feel comfortable with--otherwise they wouldn't be engineers in the first place.
The really gifted kids will be drawn to science no matter the financial rewards.
Really gifted kids aren't stupid. They will also figure out that they likely won't get to do science for decades if they follow standard career paths. And they'll figure out that there is a good chance that they end up poorly paid and without a reasonable job in their 40s. In physics and biophysics (two scientific careers that I was considering), in many subfields, you end up being someone's underpaid lab assistant for a decade or more.
The best of the best still get good jobs, and there's still a lot of jobs at 2nd & 3rd tier universities.
Jobs in academia and science are often not awarded based on the ability to do science; they are awarded based on the ability to attract funding, students, and attention, and to get good peer reviews. That's not the same. It may be the best measure of "good science" that we have, but that doesn't make it so. The past shows us that much of the best science was not the stuff that peers thought valuable at the time. And the only way to make sure enough of that happens is to make sure there is a lot of excess science funding for stuff beyond "the best of the best", according to current wisdom.
And academic positions are not primarily about science. Even in the ideal case, they should be about teaching. And in the real world, they often are about neither.
Finally, doing science at 2nd and 3rd tier universities is hard because funding is disproportionately difficult.
Let supply and demand sort it out.
It is sorting it out: the demand for scientists is actually quite low in the US (and even lower elsewhere). That's why people choose different careers. The question is: is that a good thing?
This is a marketing invention. Foveon's big claim to fame is that they measure three color components at spatially the same pixel. That's not what the human eye does at all. The human eye and brain are masters at the process that single CCD and CMOS cameras attempt: interpolation and filling in missing information. Foveon is a nifty engineering attempt at trying to avoid having to figure out how to duplicate that natural process.
The Foveon images are nice, but frankly, I'd much rather have a 9 megapixel regular sensor than a 3 megapixel Foveon sensor: on most images, the 9 megapixel sensor is going to give you better resolution, and on images where it is worse, it will be only slightly worse (assuming the anti-aliasing and color processing is done correctly).
An experiment like this one has the potential to disconfirm the hypothesis as well.
It's about probabilities, not potential. If the probability of alternative hypotheses is low, then that means that you learn very little from the outcome of the experiment if it goes as expected.
The slur against physicists is unjustified, particularly the "elevate it to dogma" line.
Oh, it's quite justified: hypothesis testing and the information derived from experiments are well understood subjects in mathematics, statistics, and information theory.
If you want to test a hypothesis, you first assume that it is correct, then try to prove the assumption wrong. If you have a better method, please share it.
There is nothing wrong with the methodology, there is something wrong with the interpretation. A diferent outcome would have given you a lot of information. As is, we learned almost nothing. Most importantly, this experiment should not strengthen our confidence in GR significantly.
Ever tried to use an app that emulates your OS's native widgets with skins? It doesn't look right, it ignores global color and font settings, it ignores UI guidelines, it behaves differently when you drag the scrollbars, it uses its own oddball keystroke commands, you can't drag-n-drop to or from it... bleh.
You mean like Carbon on OS X? Sure, Carbon apps are annoying, but I can live with them, as can apparently most other people.
Seriously, this isn't about "skinning". X11 widget sets are often highly reconfigurable: color and font settings will follow system guidelines and keystrokes are reconfigurable. Drag-and-drop integration is something Apple could handle in the X11 server (and many Carbon and Cocoa apps, unfortunately, have serious limitations on dnd anyway).
X11 support for OS X could be as good as Carbon support, if not better. Apple just has to want it and let it happen. And that's what they seem to be realizing, common prejudices like yours notwithstanding.
I agree with your points. Additionally, I think it's important to realize that there is value in simplicity and limited feature sets in itself. More features often make a software system worse. Getting more resources to add more features to Linux is not necessarily a good thing.
The original UNIX developers limited their feature sets because that was their philosophy. Open source projects limit their feature sets because they rationally decide that the time and effort it takes to implement and maintain the feature are not warranted by its utility.
Microsoft, however, has the resources to implement every nifty feature they can think of. It doesn't make their systems better, it turns them into Rube Goldberg systems. The NT kernel really is "the most modern kernel in the world" in terms of having lots of features that CS researchers ever dreamed up, and that is not an asset. Microsoft developers get paid to add features, not because they need them or because it makes sense to do so, but because they can and because it looks good on glossies. Many people go to work for Microsoft because they can finally get their ideas into a shipping system without such pesky details as having to convince customers that they actually need it enough to pay for it--all they need at Microsoft is buy-in from a few managers.
Or, to continue the dermatological paradigm, if you keep scratching itches you only imagine you have, you are going to get scratch dermatitis. Too much scratching is bad for you, as is too much fiddling around with software.
Linux should keep things simple. Some of the attention from big companies and the resources they contribute (IBM, Sun, etc.) threaten to cause Linux to make the same mistakes Microsoft is making. Fortunately, Linux is still much more modular, so all that stuff can be left out. But it will be the end of Linux when the kernel and desktop will ship with all the "commercial grade" stuff lots of vendors have contributed.
You can't prove a physical theory - you can either show that it fits experimental evidence (in which case it might be right), or that it doesn't (in which case you've disproved it).
The information you gain from an experiment is related to the change in the probabilities you assign to different hypotheses. In particular, what that means is that, unless you have a plausible alternative hypothesis, experiments that agree with your hypothesis tell you essentially nothing.
Physicists regularly disregard this simple principle; they pick one hypothesis, elevate it to dogma, and then carry out lots of experiments. And when everything is said and done, they say "hey, see, the hypothesis survived thousands of experimental tests". But none of those tests may have tested against a logically plausible (or possible) hypothesis, or the tests may not have been independent of each other.
More specifically, unless you can come up with a logically consistent alternative to GR, something that makes experimentally testable predictions, there isn't really much point in spending a lot of time and money on experiments. Shots in the dark like this one are only worth it if they are cheap (as this one was), but when the result is as expected, you didn't learn anything from it.
Yeah, like OpenOffice under X11 has any chance of remotely looking like an OS X App
Do you complain about Classic apps on OS X as well? Do you think Carbon apps are "foreign" on OS X or have trouble looking like native apps just because they use a different API?
It took some work for Apple to make Carbon apps look like OS X, but they got it done. If anything, it's easier for X11, many X11 toolkits and apps already have all the necessary theming and rebinding in place--they can look as much like Aqua apps as Apple's lawyers allow it.
Slashdot Mirror
It's common in these kinds of settlements for manufacturers to give vouchers for their own products. This is likely no different: you can probably use the vouchers to buy Microsoft software and hardware running Microsoft software. Non-redeemed vouchers will be given as 1/3 in Microsoft software and 1/3 in hardware (presumably, PCs running Microsoft software) to schools.
Even $1.1 billion in cash would be a slap on the wrist. This "settlement" is an insult: it's a marketing promotion for Microsoft and a means by which they can get more of their software into the education market.
The kicker is, however, $1bn is about as much as Toshiba had to pay for shipping supposedly defective floppy disk drives on their laptops.
I think this is absolutely evil. Even if Microsoft had to pay everything in cash, it would be peanuts. Instead, they'll be able to further contaminate schools with their proprietary software, something they have already volunteered to do as a "donation"--a tax sheltered marketing ploy.
That's two-channel audio. You are going to get many more channels (in fact, DVD audio already does, I believe).
Video isn't going to get a heck of a lot bigger than DVD-Video sizes.
Sure it is: DVD resolution is really low. You are first going to see HDTV-resolution DVDs, and later probably 3-6 MPixel video. Beyond that, there will be many simultaneous video streams: half a dozen camera angles from a live event, etc.
First of all, you can really only dual-license if there is something to sell. That means dual-licensed open source software needs to have one of the less "business-oriented" open source licenses, like the GPL. But for some software, the GPL just isn't appropriate. For example, for a GUI toolkit, if it is GPL'ed, that limits its adoption.
Also, with dual-licensed software, companies have other considerations in mind than just creating technically the best possible software or the best software for the users. They may want to try to dominate a particular market niche and exclude competition, for example.
I think in the case of Troll Tech, both of these apply. Companies like Sun and IBM have chosen Gnome because they did not want to hand the keys to commercial applications for their desktops to some small company that they have no control over. Furthermore, Troll Tech has a strong incentive to have something like Qt/Embedded take over the entire display--if people could use free or competing toolkits on something like the Sharp Zaurus, it would threaten their revenue stream.
I think dual licensing can make sense in some circumstances, but I think it's risky. If you are considering using dual-licensed software, I think you would do well to look at the license, business, and track record very carefully. Personally, I think for MySQL it works, but for Qt, it doesn't.
Another concern about TiVo is that they get detailed information about my viewing habits and can correlate that with my credit card (whether they do it or not is another matter).
I want to reduce the number of companies that have their fingers in my finances.
I think subscription-based PVR models are really a bit expensive for the consumer in the long run. Tivo may be selling their hardware under cost, but enough to justify the subscription cost? I personally don't believe so.
Trouble is: it isn't new. Smart wrist watches are an old idea, as are portable multimedia player boxes.
Microsoft wouldn't be ridiculed if they came up with something genuinely new, or if they did a really good job reimplementing some known idea. But mediocre copies of other people's ideas just invite ridicule; they'll have to deal with it. Gates can at least laugh all the way to the bank: mediocre copies do sell, after all, as many other companies also show us.
Last I looked, Apple's license on this also made it unclear whether Linux distributions would pick it up. So, widespread deployment may have to wait until GPL or BSD implementations.
Safari is a beta software anyway. Use it at your own risk.
Beta software is handed out by companies to get useful feedback from users. It's of benefit to the companies. Users download it out of curiosity, but that only goes so far. Companies that take your attitude will soon run out of beta testers (and probably users).
Beta software may perform poorly, but if anything, it should have extensive error checks compiled into it.
But beta software is deliberately released to customers for testing. This is for the benefit of the company, not the customers. Companies should go out of their way to make sure their beta software doesn't do anything seriously bad, and blaming the user for trying the software is self-defeating.
If anything, beta software should have lots of extra checking compiled into it; it might quit more often than released software and run slower, but it shouldn't crash in unexpected ways any more frequently, and it should definitely go out of its way to guard against data loss.
What we have here is two greedy organizations battling it out. If we want to avoid getting dragged into this, we really do need open video standards.
I don't have any of those certifications (thank God!), but I have had no trouble pulling together a "home media unit" from scratch.
A standard Linux install pretty much has all you need, and you have lots of choices for how to set it up--all of them pretty simple. You can set up the box as a streaming media server, or you can make it part of your own in-home P2P network, or you can handle all the music through web interfaces. You can push audio to a Linux box or have it pull it from other systems with standard commands.
Another very simple approach is to get a Macintosh--iTunes pretty much does everything you need for that out of the box.
So, we have more need for 6 disc changes than we do for 10 GB discs of hard drives on which to store mostly-illegally-obtained mp3s.
I own all the CDs for the MP3s that I have. Why store them on-line? Because a computer is much, much more convenient than two 300 CD jukeboxes or, worse, lots of jewel cases and strange looking pieces of furniture.
No one aside from the most hardcore Slashdotter would even know what to do with one of these
My parents seem to have no trouble understanding the convenience of just selecting a CD from an on-screen list, as opposed to dealing with hundreds of jewel boxes.
Of course, little of this applies to this HP device, which does sound much more complex, less functional, and proprietary than just getting a Mini-ITX system.
That would still be annoying as hell, but you wouldn't lose any functionality over what you get now; it would just become harder and harder to access things like the next generation of digital audio/video and Windows media from Linux--at least if the big studios and record companies get their way.
Overall, though, I still think that this will just flop: no hardware or software vendor really needs the hassle or additional support costs that result from this.
If you want something today, get a mini laptop from Fujitsu, Sony, or Dynamism.
For $400, you can get a Mini-ITX system or a WalMart PC and a better set of powered speakers. That also gives you decent amounts of MP3 storage. The result will be more flexible and convenient than this thing.
I'm not sure I like the restriction, but at least I can understand why a traditional publisher or librarian might want to impose it.
The Google cache shouldn't be a problem, and Citeseer shouldn't be a problem either (it doesn't try to be archival, as far as I can tell).
The route to power and better pay for the engineering profession can be (and already is, in many cases) is to work independently and hire out your services. But with that comes a level of risk and entrepreneurship that many technically inclined people don't feel comfortable with--otherwise they wouldn't be engineers in the first place.
Really gifted kids aren't stupid. They will also figure out that they likely won't get to do science for decades if they follow standard career paths. And they'll figure out that there is a good chance that they end up poorly paid and without a reasonable job in their 40s. In physics and biophysics (two scientific careers that I was considering), in many subfields, you end up being someone's underpaid lab assistant for a decade or more.
The best of the best still get good jobs, and there's still a lot of jobs at 2nd & 3rd tier universities.
Jobs in academia and science are often not awarded based on the ability to do science; they are awarded based on the ability to attract funding, students, and attention, and to get good peer reviews. That's not the same. It may be the best measure of "good science" that we have, but that doesn't make it so. The past shows us that much of the best science was not the stuff that peers thought valuable at the time. And the only way to make sure enough of that happens is to make sure there is a lot of excess science funding for stuff beyond "the best of the best", according to current wisdom.
And academic positions are not primarily about science. Even in the ideal case, they should be about teaching. And in the real world, they often are about neither.
Finally, doing science at 2nd and 3rd tier universities is hard because funding is disproportionately difficult.
Let supply and demand sort it out.
It is sorting it out: the demand for scientists is actually quite low in the US (and even lower elsewhere). That's why people choose different careers. The question is: is that a good thing?
You mean kind of like what companies are already doing when they encase electronics in epoxy or other plastics?
The Foveon images are nice, but frankly, I'd much rather have a 9 megapixel regular sensor than a 3 megapixel Foveon sensor: on most images, the 9 megapixel sensor is going to give you better resolution, and on images where it is worse, it will be only slightly worse (assuming the anti-aliasing and color processing is done correctly).
It's about probabilities, not potential. If the probability of alternative hypotheses is low, then that means that you learn very little from the outcome of the experiment if it goes as expected.
The slur against physicists is unjustified, particularly the "elevate it to dogma" line.
Oh, it's quite justified: hypothesis testing and the information derived from experiments are well understood subjects in mathematics, statistics, and information theory.
If you want to test a hypothesis, you first assume that it is correct, then try to prove the assumption wrong. If you have a better method, please share it.
There is nothing wrong with the methodology, there is something wrong with the interpretation. A diferent outcome would have given you a lot of information. As is, we learned almost nothing. Most importantly, this experiment should not strengthen our confidence in GR significantly.
You mean like Carbon on OS X? Sure, Carbon apps are annoying, but I can live with them, as can apparently most other people.
Seriously, this isn't about "skinning". X11 widget sets are often highly reconfigurable: color and font settings will follow system guidelines and keystrokes are reconfigurable. Drag-and-drop integration is something Apple could handle in the X11 server (and many Carbon and Cocoa apps, unfortunately, have serious limitations on dnd anyway).
X11 support for OS X could be as good as Carbon support, if not better. Apple just has to want it and let it happen. And that's what they seem to be realizing, common prejudices like yours notwithstanding.
The original UNIX developers limited their feature sets because that was their philosophy. Open source projects limit their feature sets because they rationally decide that the time and effort it takes to implement and maintain the feature are not warranted by its utility.
Microsoft, however, has the resources to implement every nifty feature they can think of. It doesn't make their systems better, it turns them into Rube Goldberg systems. The NT kernel really is "the most modern kernel in the world" in terms of having lots of features that CS researchers ever dreamed up, and that is not an asset. Microsoft developers get paid to add features, not because they need them or because it makes sense to do so, but because they can and because it looks good on glossies. Many people go to work for Microsoft because they can finally get their ideas into a shipping system without such pesky details as having to convince customers that they actually need it enough to pay for it--all they need at Microsoft is buy-in from a few managers.
Or, to continue the dermatological paradigm, if you keep scratching itches you only imagine you have, you are going to get scratch dermatitis. Too much scratching is bad for you, as is too much fiddling around with software.
Linux should keep things simple. Some of the attention from big companies and the resources they contribute (IBM, Sun, etc.) threaten to cause Linux to make the same mistakes Microsoft is making. Fortunately, Linux is still much more modular, so all that stuff can be left out. But it will be the end of Linux when the kernel and desktop will ship with all the "commercial grade" stuff lots of vendors have contributed.
The information you gain from an experiment is related to the change in the probabilities you assign to different hypotheses. In particular, what that means is that, unless you have a plausible alternative hypothesis, experiments that agree with your hypothesis tell you essentially nothing.
Physicists regularly disregard this simple principle; they pick one hypothesis, elevate it to dogma, and then carry out lots of experiments. And when everything is said and done, they say "hey, see, the hypothesis survived thousands of experimental tests". But none of those tests may have tested against a logically plausible (or possible) hypothesis, or the tests may not have been independent of each other.
More specifically, unless you can come up with a logically consistent alternative to GR, something that makes experimentally testable predictions, there isn't really much point in spending a lot of time and money on experiments. Shots in the dark like this one are only worth it if they are cheap (as this one was), but when the result is as expected, you didn't learn anything from it.
Do you complain about Classic apps on OS X as well? Do you think Carbon apps are "foreign" on OS X or have trouble looking like native apps just because they use a different API?
It took some work for Apple to make Carbon apps look like OS X, but they got it done. If anything, it's easier for X11, many X11 toolkits and apps already have all the necessary theming and rebinding in place--they can look as much like Aqua apps as Apple's lawyers allow it.