Mouse - two computers, one mouse? madness. Same for keyboards.
Actually I use synergy to do this all the time. (Between Windows & Linux no less.) It's useful when you have a laptop and a desktop workstation, like I do in my lab at school.
Sometimes I wonder, and this "multiswitch" idea just brings home the point, why we have USB and Firewire when it seems like it would be just as effective, and more standardized, for every device to just have built-in gigabit ethernet chips that can communicate using UDP or something.
Protocol? Why not USB over ethernet? Or use OpenSoundControl! _anything_ standardized... Think how much easier that would make it to write drivers. The point is that we can easily separate the protocol from the physical layer, or even from the transport layer. And yet we still have very specific protocols for USB and Firewire technology that are tied to the hardware they run on. It makes little sense to me.
It just seems silly to have all these communication standards that are basically just reinventing the IP protocol. IP has been "plug and play" for like a decade before USB was invented. At the time, of course, it was necessary to have something that could transfer data at certain rates that were unachievable otherwise, but now that most new computers have on-board gigabit ethernet, maybe it's time we took advantage of it. The nice thing about sticking to STANDARDS is that the next time they upgrade the ethernet hardware (10 Gb onboard, for example), device communication would automatically be upgraded with it. As a bonus, backwards compatibility would be easily assured.
Meanwhile, let's improve those damn ethernet connectors already. Goddamn tabs always breaking off... I know the plastic tabs are a cost-effective solution, but I think we could do better, I honestly do.
using or simulating Autodesk's TrustedDWG technology
If it's just a matter of some kind of crypto-enabled signature technology to prove the file was created by Autodesk software, couldn't the signature simply be ignored? If that block of the file is simply skipped over, and other techniques are used to determine if its a valid file (like checking the file structure), would that still count as "simulating TrustedDWG"? Probably not, I think. I mean, if we're reverse engineering the format, who cares if it's a signed file or not. Just get the data from the relevent parts and call it a day.
Or is it impossible to access the data without verifying the signature? In that case it would be an encrypted file, not just a signed file -- a different case altogether.
Harrison says that the current PS3 game lineup is using less than half of the machines power, adding that 'nobody will ever use 100 percent of its capacity.' Is he right?
As long as they don't release drivers for the rest of the hardware (graphics etc) to the Linux community, then maybe he's right. Make the secrets public, and I'm sure someone can come up with a use for all this power. Until then, the PS3 supposedly isn't much better than a cheap computer. There IS power hidden in the PS3, but they aren't making it easy to access.
If they are trying to sell this thing as a console, fine, but if they are trying to sell it as "more than a console" (i.e., a computer), then provide the required drivers and people will use it to its full potential, I promise.
The problem I can see is that they aren't being consistant with their marketing efforts. I feel like they keep flipflopping back and forth between claiming that this thing is a console, or a computer. (Is it something in between? Do we need a new word? Conputer? Compusole?)
This is a great step for Google! One thing I haven't seen mentioned yet is that that open-sourcing the toolkit guarantees its longevity. I had looked at GWT before and thought it was really great, but I'm always a little wary of starting a project based on a company's "free" offering, no matter how nice it looks, because the last thing I want is for the rug to be pulled out from under me at the last minute. Now that it has been open-sourced under a decent license, there is an implied guarantee that I can now depend on this product to be around as long as there is a public interest in it. That makes the difference, for me, to whether I'll use it or not.
What do you mean "rather than". In a communist system each person is considered worth equally much, and the wealth is distributed accordingly, ie. equally.
You're absolutely right. Thanks for the insightful correction.
This non-concept of "economy of sharing" goes like: I have something, which you want, and I am morally obligated to give it to you, by virtue of the fact that I have it.
You're wrong. You are describing a communist system, where wealth is distributed evenly, rather than according to how much each person is worth.
That's not open-source. To me, there is a huge difference with open-source: It is specifically about acknowledging how much something is worth, giving credit where it's due, and respecting the wishes of the authors. Thus, if you build something on top of what I have built, and I have shared it, all I ask is that you share it too. There is nothing in open-source that says that if you build something from scratch, you absolutely must open-source it. Only if you use parts of what other people did. Frankly, I think that's a reasonable request.
What it means is that it's more efficient than traditional innovation, because it means not having to re-invent the wheel. All we ask is that you open your code, too. You're perfectly free to not use what someone else did, but it would be re-doing a lot of work, so I don't recommend it.
Where is it in my interest to do so, if I do not accept your premise that I am somehow inherently obligated to?
You're only obligated if you are using something someone else did. Again, how is this not reasonable? If you're going to go and sell some code you wrote, but it includes a bunch of code I wrote, and I stated originally that I'd prefer you to share your code if you use it, then you're not inherently obliged to, you're obliged to according to the license agreement that you chose to comply with.
The "one laptop per child" mentality is great at giving people the information that they need in order to succeed, but it will not make them succeed. It will ensure that everyone starts the race at the same point, but it will not make everyone a winner.
Absolutely. However, the hope is that it will, in total, create more winners. Or at least even out the distribution of winners over the globe. Right now there is a serious imbalance in the world that is making it a very unhealthy place to live. We can't just keep giving money to developing countries, hoping that they'll invest it properly and fix all their economic problems. Instead, this is an attempt to help them help themselves, a much better approach IMHO. Anyways, notice that the OLPC project isn't exactly a charity. It is an effort to create a machine that is useful, but made in such a way that the target demographic can actually afford it. This is perfectly moral from a capitalist perspective. (Yes it is a non-profit organization, but as far as I'm concerned that doesn't change anything. They are still selling the machines, not giving them away.)
I agree. The CPU issue only becomes important when talking about servers accepting thousands of connections at a time, where all data being passed around is gzipped XML. Believe it or not, I have read that there are situations where CPU time can exceed the I/O bottleneck. Not to mention the power consumption associated with it. It all depends on what kind of scale you are talking about. Certainly, for most applications, the CPU thing is a non-issue.
However, the BitTorrent protocol enables content distributors like Azureus to share large files using much less bandwidth.
Content distributors? Since when is Azureus a content distributor? Is that some sort of entity? I thought Azureus was just a software program implementing a specific internet protocol.
I would think the "content distributors" would be the websites hosting the trackers, etc.
Oh well, maybe it's just a semantic misunderstanding on my part.
You could certainly make XML vastly more compact if you had some table of tags mapped to 2-byte codes.
I used to entertain this idea as well. But then it occurred to me, that it is almost the same thing if you simply zip the XML code, using a compression program such as zip, gzip, or bzip2. If you read up on how these lossless compression algorithms work, essentially they go through the input and built up an "index" of repeated strings. That is, any sequence of characters found to be repeated is placed in the index, and the instances of this sequence are replaced by the index number. (Essentially, as you probably know, this is the Lempel-Ziv algorithm.)
So basically, using standard compression techniques, you can build a "binary-encoded XML" file by simply compressing the XML. Then you have standard programs and libraries that can decompress it, and parse it with a standard XML parser. Magic!
I've since concluded that this way you can get the advantages of a binary encoding AND the advantages of a simple and standard markup at the same time. (Note however that it is definitely not as CPU-friendly as a well-defined binary format like RIFF.)
As an aside, is an excellent illustration of the value of a real computing science degree (assuming you take it at an institution where they actually expect you to learn math, numerical methods, etc).
Sadly, I have one.;-)
Don't get me wrong, I learned a hell of a lot and wouldn't exchange it for anything in the world, but frankly I wasn't the best in th e math classes, I'll admit.
However, I don't remember covering this particular algorithm. (I never took the advanced algorithms class though.)
It's one of the few instructions GCC isn't smart enough to use automatically. (another is rlwimi, which is without a doubt, the coolest instruction ever)
Wouldn't the optimization be in the PPC-specific libc code for the sqrt() function, rather than being something that GCC does?
Because in C's own weird way, that's the only way of refering to a float as an int without changing the bits.
If you do this: int i = (int)3.0f;
You get i=3, like what you'd get from the floor() function.
If you do this: float f = 3.0f; int i = *(int*)
Then i contains a bit-for-bit copy of the IEEE floating-point representation of 3.0.
It's because C knows how to cast a float to an int by applying the floor function. However, if you do it the second way, you aren't casting a float to an int, you are casting a pointer-to-float to a pointer-to-int and then dereferencing it.
By the way, I just wanted to say... this is one of the most interesting things I've read on Slashdot in a while. Wow. That function is just amazing. I only wish I understood how it worked. I know nothing about what a "Newton-Raphson iteration" is.
I completely agree. Hell, even when they decided to change the header file from "d3d.h" to "d3d9.h", it completely screwed up the source code I was supporting, I had to add #ifdef flags and inform the user how to deal with them depending on which version of DX they wanted to use.
They go and completely change API stuff all the time without any regard to how it affects their customers. The change from DX9 to DX10 is the most drastic ever, and it will be completely incompatible. Aaaaagh... *shakes head in frustration*
I understand that they implemented a new driver model and all the arguments, but fuck, do they not understand the idea of an API?? To abstract the details so we don't have to worry about changes to the back-end implementation.. and the whole point of all the complexities of COM was supposed to be so we wouldn't ever have to deal with these issues. It's enough to drive a guy crazy!
linux is not a deploy and forget solution, those costs like an MS environment are very real and VERY expensive.
Yes, but it compares to MS, I would think. MS is certainly not a deploy and forget solution. You assume I was talking about initial costs.. I was not. I absolutely think that a properly administered *nix system will generally run better and be easier to maintain, given the appropriate training.
But I'm a programmer, not a network administrator, so be my guest to take it with a grain of salt.
"'The evidence on the cost savings attributable to a switch to Linux has been mixed,' according to Chris Swenson, director of software industry analysis at research group NPD. 'There has been some evidence that companies have to spend a good deal on training and support after you deploy...'"
Oh my god am I tired of this argument... some people seem to have very little grasp over "long term" and "short term" savings.
"It's different! It's hard to learn! Therefore it can't be good for us in the long run..."
A few Nov 27 11:43:05 <Nat_> We are collaborating with Microsoft on a few different interop areas Nov 27 11:43:27 <Nat_> We'll be adding Open XML support to OpenOffice, building a virtualization shim to run SLES optimized on Veridian and Vista on Xen Nov 27 11:43:44 <Nat_> We'll also be working together on WS-Management Nov 27 11:43:46 <Nat_> All this code will be released open source Nov 27 11:43:47 * cboltz (n=cboltz@88.134.58.13) has joined #opensuse-project Nov 27 11:43:51 <Nat_> so everyone gets that, and can benefit from it Nov 27 11:44:33 <Nat_> (By the way, in that process, we don't plan to add MS-patented code to our contributions) Nov 27 11:44:42 <Nat_> (Our policy on that is unchanged -- and MS didn't give us the right to do that anyway!)
Hm, wow, I'm convinced. So what was the point of the deal then? Either you'll be contributing code that you couldn't have before, meaning no one else who doesn't have a similar MS deal can use, or you'll be contributing code that you could have easily added previously anyways. I don't get it.
it's curious... normally for small devices like these there are tons of posts about "does it run linux?" or other talk of hacking the hardware to do what it wasn't meant to.
Is it such a bad hardware design that even misappropriating it to run linux or rockbox would be a waste of time? It seems the MS-inspired nature of the device has been enough to even ward off the most dedicated of hackers.. very rare indeed. I have pretty much seen NO talk of wanting to hack this thing, which I find really curious.
I for one could see a little wireless mp3 player running custom software to be a very useful device.
Is the fact that it requires special software a HARD limitation or is it something in the firmware that could be overridden? Imagine a DRM-less wireless mp3 player / PDA / portable video game machine...
The only thing that makes me feel uneasy about this whole thing is the necessity for government intervention. It would seem that things like this, in a free market, should take care of themselves...
Other companies can't create compatible software, thus Microsoft should somehow feel that burden and suffer somehow. I guess that hasn't happened..
But when the government has to intervene in ways like this, it reminds me a little too much of Reardon being forced to hand over the recipe for his metal alloy (Atlas Shrugged, Ayn Rand). It seems wrong, somehow, that Microsoft should be forced to give up their secret sauce. Believe me, I am no Microsoft fan, but I would have thought that the market should take care of this kind of problem itself. It should have been, somehow, in MS's interests to allow other companies to interoperate with their own. For example, "the more people we can work with, the more we'll sell our operating system." Why has this not happened?
I guess that's what being a "monopoly" is then -- when it's no longer in your self-interests to "play nice" with others.
Does the analogy with Reardon Metal, or McDonalds Secret Sauce, end when you realize that software is inherently different than a physical substance? If so, why is that? How is it different?
This just goes to show what I've suspected for a while... there is a problem with the software industry in general. The goals of a software company ultimately have a contradiction with the goals of software itself: The software company must fight its own products.
This is one of the reasons that, although I am a programmer by nature and by trade, I have a really hard time with the idea of starting up some kind of software company. I'd rather other people take those risks and hire me. As I see it, there are two problems with starting a software company:
1) Your product is inherently easily copied, giving it low value no matter how good it is. In fact, the more popular it is, the more likely it will be pirated, thus the better it is the LESS value it potentially has. This is definitely counter-intuitive. 2) Once you create a product that does what it needs to do and is easy to use, what then? Software eventually always reaches a plateau, and it becomes a question of "now what?" At this point software companies start to add "features" that bloat the software bundle and aren't wanted by customers, in the hopes that they've at least acquired a dedicated customer base that will buy the new version simply because it is "the newest version".
No, as I see it, it's better to do software in your spare time, and release it for free. Not because I'm some kind of altruist, but just because I see it as being a much more viable way to focus on the "product" rather than the "profits".
To clarify -- I'm certainly not any kind of anti-profit advocate. I'm a capitalist. I just don't see software and other information-based services as fitting into a capitalist model very well. As soon as you are a software company, you must focus on getting customers to upgrade, rather than on making sure they have a good experience with your product. Any industry in which it's in a company's interests to make sure its own customers are having a bad experience is in contradiction with itself, as far as I can see. I think the same thing goes for anti-virus products -- it's in their interests to make sure there are viruses around. They have built up their flagship products on the existance of something evil. There is simply something wrong with that.
This is why I tend to trust open-source products. I know that they have no reason to exist except to "get the job done", and therefore they do what they are meant to, and nothing else.
Neal Stephenson mentions this in the Baroque Cycle. He talks about how the little eggs of steel were forged in India and hammered out to make watered steel, then sold to the asian market. I assume he is talking about the same thing? I believe he even used the word "wootz", but I can't recall.
Does anyone have a list of Microsoft patents that affect GNU/Linux? Like, what exactly are they providing indemnification for? And how many of them likely have plenty of prior art that could be used to fight in court? Are there any that we should specifically be worried about?
Additionally, another thing I don't get about this is that by making this Novell deal, they seem to be indicating that they are willing to sue customers of other distros for patent infringement. But since when do CUSTOMERS get sued for patent infringement? Last I checked it was only the vendors of infringing products that could get sued for patent infringement.
Of course, the real WTF here is that it doesn't just show up as a USB disk drive when you plug it in, requiring no special drivers or software at all. Every other damn player on the market works this way, and we have standards for a reason.
(Apparently that reason is so MS has something to point at and laugh about...)
A corporation exists to make money for its owners period
Agreed. But there is a difference between short-term money and long-term money. One doesn't guarantee the other, and it takes a good CEO to tell the difference. It also takes intelligent shareholders to avoid lynching him in the meantime. I think in a lot of cases, bad business decisions can be attributed to shareholders will very short-term vision.
Slashdot Mirror
Actually I use synergy to do this all the time. (Between Windows & Linux no less.)
It's useful when you have a laptop and a desktop workstation, like I do in my lab at school.
Sometimes I wonder, and this "multiswitch" idea just brings home the point, why we have USB and Firewire when it seems like it would be just as effective, and more standardized, for every device to just have built-in gigabit ethernet chips that can communicate using UDP or something.
Protocol? Why not USB over ethernet? Or use OpenSoundControl! _anything_ standardized... Think how much easier that would make it to write drivers. The point is that we can easily separate the protocol from the physical layer, or even from the transport layer. And yet we still have very specific protocols for USB and Firewire technology that are tied to the hardware they run on. It makes little sense to me.
It just seems silly to have all these communication standards that are basically just reinventing the IP protocol. IP has been "plug and play" for like a decade before USB was invented. At the time, of course, it was necessary to have something that could transfer data at certain rates that were unachievable otherwise, but now that most new computers have on-board gigabit ethernet, maybe it's time we took advantage of it. The nice thing about sticking to STANDARDS is that the next time they upgrade the ethernet hardware (10 Gb onboard, for example), device communication would automatically be upgraded with it. As a bonus, backwards compatibility would be easily assured.
Meanwhile, let's improve those damn ethernet connectors already. Goddamn tabs always breaking off...
I know the plastic tabs are a cost-effective solution, but I think we could do better, I honestly do.
If it's just a matter of some kind of crypto-enabled signature technology to prove the file was created by Autodesk software, couldn't the signature simply be ignored?
If that block of the file is simply skipped over, and other techniques are used to determine if its a valid file (like checking the file structure), would that still count as "simulating TrustedDWG"? Probably not, I think.
I mean, if we're reverse engineering the format, who cares if it's a signed file or not. Just get the data from the relevent parts and call it a day.
Or is it impossible to access the data without verifying the signature? In that case it would be an encrypted file, not just a signed file -- a different case altogether.
As long as they don't release drivers for the rest of the hardware (graphics etc) to the Linux community, then maybe he's right.
Make the secrets public, and I'm sure someone can come up with a use for all this power. Until then, the PS3 supposedly isn't much better than a cheap computer. There IS power hidden in the PS3, but they aren't making it easy to access.
If they are trying to sell this thing as a console, fine, but if they are trying to sell it as "more than a console" (i.e., a computer), then provide the required drivers and people will use it to its full potential, I promise.
The problem I can see is that they aren't being consistant with their marketing efforts. I feel like they keep flipflopping back and forth between claiming that this thing is a console, or a computer. (Is it something in between? Do we need a new word? Conputer? Compusole?)
This is a great step for Google! One thing I haven't seen mentioned yet is that that open-sourcing the toolkit guarantees its longevity. I had looked at GWT before and thought it was really great, but I'm always a little wary of starting a project based on a company's "free" offering, no matter how nice it looks, because the last thing I want is for the rug to be pulled out from under me at the last minute.
Now that it has been open-sourced under a decent license, there is an implied guarantee that I can now depend on this product to be around as long as there is a public interest in it. That makes the difference, for me, to whether I'll use it or not.
You're absolutely right. Thanks for the insightful correction.
You're wrong. You are describing a communist system, where wealth is distributed evenly, rather than according to how much each person is worth.
That's not open-source. To me, there is a huge difference with open-source: It is specifically about acknowledging how much something is worth, giving credit where it's due, and respecting the wishes of the authors. Thus, if you build something on top of what I have built, and I have shared it, all I ask is that you share it too. There is nothing in open-source that says that if you build something from scratch, you absolutely must open-source it. Only if you use parts of what other people did. Frankly, I think that's a reasonable request.
What it means is that it's more efficient than traditional innovation, because it means not having to re-invent the wheel. All we ask is that you open your code, too. You're perfectly free to not use what someone else did, but it would be re-doing a lot of work, so I don't recommend it.
You're only obligated if you are using something someone else did. Again, how is this not reasonable? If you're going to go and sell some code you wrote, but it includes a bunch of code I wrote, and I stated originally that I'd prefer you to share your code if you use it, then you're not inherently obliged to, you're obliged to according to the license agreement that you chose to comply with.
Absolutely. However, the hope is that it will, in total, create more winners. Or at least even out the distribution of winners over the globe. Right now there is a serious imbalance in the world that is making it a very unhealthy place to live. We can't just keep giving money to developing countries, hoping that they'll invest it properly and fix all their economic problems. Instead, this is an attempt to help them help themselves, a much better approach IMHO.
Anyways, notice that the OLPC project isn't exactly a charity. It is an effort to create a machine that is useful, but made in such a way that the target demographic can actually afford it. This is perfectly moral from a capitalist perspective. (Yes it is a non-profit organization, but as far as I'm concerned that doesn't change anything. They are still selling the machines, not giving them away.)
I agree. The CPU issue only becomes important when talking about servers accepting thousands of connections at a time, where all data being passed around is gzipped XML. Believe it or not, I have read that there are situations where CPU time can exceed the I/O bottleneck. Not to mention the power consumption associated with it. It all depends on what kind of scale you are talking about.
Certainly, for most applications, the CPU thing is a non-issue.
Content distributors? Since when is Azureus a content distributor? Is that some sort of entity? I thought Azureus was just a software program implementing a specific internet protocol.
I would think the "content distributors" would be the websites hosting the trackers, etc.
Oh well, maybe it's just a semantic misunderstanding on my part.
I used to entertain this idea as well. But then it occurred to me, that it is almost the same thing if you simply zip the XML code, using a compression program such as zip, gzip, or bzip2.
If you read up on how these lossless compression algorithms work, essentially they go through the input and built up an "index" of repeated strings. That is, any sequence of characters found to be repeated is placed in the index, and the instances of this sequence are replaced by the index number.
(Essentially, as you probably know, this is the Lempel-Ziv algorithm.)
So basically, using standard compression techniques, you can build a "binary-encoded XML" file by simply compressing the XML. Then you have standard programs and libraries that can decompress it, and parse it with a standard XML parser. Magic!
I've since concluded that this way you can get the advantages of a binary encoding AND the advantages of a simple and standard markup at the same time.
(Note however that it is definitely not as CPU-friendly as a well-defined binary format like RIFF.)
Sadly, I have one.
Don't get me wrong, I learned a hell of a lot and wouldn't exchange it for anything in the world, but frankly I wasn't the best in th e math classes, I'll admit.
However, I don't remember covering this particular algorithm. (I never took the advanced algorithms class though.)
Wouldn't the optimization be in the PPC-specific libc code for the sqrt() function, rather than being something that GCC does?
Because in C's own weird way, that's the only way of refering to a float as an int without changing the bits.
If you do this:
int i = (int)3.0f;
You get i=3, like what you'd get from the floor() function.
If you do this:
float f = 3.0f;
int i = *(int*)
Then i contains a bit-for-bit copy of the IEEE floating-point representation of 3.0.
It's because C knows how to cast a float to an int by applying the floor function. However, if you do it the second way, you aren't casting a float to an int, you are casting a pointer-to-float to a pointer-to-int and then dereferencing it.
By the way, I just wanted to say... this is one of the most interesting things I've read on Slashdot in a while. Wow. That function is just amazing. I only wish I understood how it worked. I know nothing about what a "Newton-Raphson iteration" is.
I completely agree. Hell, even when they decided to change the header file from "d3d.h" to "d3d9.h", it completely screwed up the source code I was supporting, I had to add #ifdef flags and inform the user how to deal with them depending on which version of DX they wanted to use.
They go and completely change API stuff all the time without any regard to how it affects their customers. The change from DX9 to DX10 is the most drastic ever, and it will be completely incompatible. Aaaaagh... *shakes head in frustration*
I understand that they implemented a new driver model and all the arguments, but fuck, do they not understand the idea of an API?? To abstract the details so we don't have to worry about changes to the back-end implementation.. and the whole point of all the complexities of COM was supposed to be so we wouldn't ever have to deal with these issues. It's enough to drive a guy crazy!
Yes, but it compares to MS, I would think. MS is certainly not a deploy and forget solution.
You assume I was talking about initial costs.. I was not. I absolutely think that a properly administered *nix system will generally run better and be easier to maintain, given the appropriate training.
But I'm a programmer, not a network administrator, so be my guest to take it with a grain of salt.
Oh my god am I tired of this argument... some people seem to have very little grasp over "long term" and "short term" savings.
"It's different! It's hard to learn! Therefore it can't be good for us in the long run..."
Some people have no vision.
Hm, wow, I'm convinced.
So what was the point of the deal then?
Either you'll be contributing code that you couldn't have before, meaning no one else who doesn't have a similar MS deal can use, or you'll be contributing code that you could have easily added previously anyways.
I don't get it.
it's curious... normally for small devices like these there are tons of posts about "does it run linux?" or other talk of hacking the hardware to do what it wasn't meant to.
Is it such a bad hardware design that even misappropriating it to run linux or rockbox would be a waste of time? It seems the MS-inspired nature of the device has been enough to even ward off the most dedicated of hackers.. very rare indeed. I have pretty much seen NO talk of wanting to hack this thing, which I find really curious.
I for one could see a little wireless mp3 player running custom software to be a very useful device.
Is the fact that it requires special software a HARD limitation or is it something in the firmware that could be overridden?
Imagine a DRM-less wireless mp3 player / PDA / portable video game machine...
The only thing that makes me feel uneasy about this whole thing is the necessity for government intervention.
It would seem that things like this, in a free market, should take care of themselves...
Other companies can't create compatible software, thus Microsoft should somehow feel that burden and suffer somehow. I guess that hasn't happened..
But when the government has to intervene in ways like this, it reminds me a little too much of Reardon being forced to hand over the recipe for his metal alloy (Atlas Shrugged, Ayn Rand). It seems wrong, somehow, that Microsoft should be forced to give up their secret sauce. Believe me, I am no Microsoft fan, but I would have thought that the market should take care of this kind of problem itself. It should have been, somehow, in MS's interests to allow other companies to interoperate with their own. For example, "the more people we can work with, the more we'll sell our operating system." Why has this not happened?
I guess that's what being a "monopoly" is then -- when it's no longer in your self-interests to "play nice" with others.
Does the analogy with Reardon Metal, or McDonalds Secret Sauce, end when you realize that software is inherently different than a physical substance? If so, why is that? How is it different?
This just goes to show what I've suspected for a while... there is a problem with the software industry in general. The goals of a software company ultimately have a contradiction with the goals of software itself: The software company must fight its own products.
This is one of the reasons that, although I am a programmer by nature and by trade, I have a really hard time with the idea of starting up some kind of software company. I'd rather other people take those risks and hire me. As I see it, there are two problems with starting a software company:
1) Your product is inherently easily copied, giving it low value no matter how good it is. In fact, the more popular it is, the more likely it will be pirated, thus the better it is the LESS value it potentially has. This is definitely counter-intuitive.
2) Once you create a product that does what it needs to do and is easy to use, what then? Software eventually always reaches a plateau, and it becomes a question of "now what?" At this point software companies start to add "features" that bloat the software bundle and aren't wanted by customers, in the hopes that they've at least acquired a dedicated customer base that will buy the new version simply because it is "the newest version".
No, as I see it, it's better to do software in your spare time, and release it for free. Not because I'm some kind of altruist, but just because I see it as being a much more viable way to focus on the "product" rather than the "profits".
To clarify -- I'm certainly not any kind of anti-profit advocate. I'm a capitalist. I just don't see software and other information-based services as fitting into a capitalist model very well. As soon as you are a software company, you must focus on getting customers to upgrade, rather than on making sure they have a good experience with your product. Any industry in which it's in a company's interests to make sure its own customers are having a bad experience is in contradiction with itself, as far as I can see. I think the same thing goes for anti-virus products -- it's in their interests to make sure there are viruses around. They have built up their flagship products on the existance of something evil. There is simply something wrong with that.
This is why I tend to trust open-source products. I know that they have no reason to exist except to "get the job done", and therefore they do what they are meant to, and nothing else.
Here are some direct links to the videos...
.... Chip Bag
.... CD
.... Light Bulb
.... Egg
.... Tin Foil
.... Lemon
.... Grape
.... Soap
Unfortunalely in wmv format, but playable in Mplayer.
Neal Stephenson mentions this in the Baroque Cycle. He talks about how the little eggs of steel were forged in India and hammered out to make watered steel, then sold to the asian market. I assume he is talking about the same thing? I believe he even used the word "wootz", but I can't recall.
Does anyone have a list of Microsoft patents that affect GNU/Linux?
Like, what exactly are they providing indemnification for?
And how many of them likely have plenty of prior art that could be used to fight in court?
Are there any that we should specifically be worried about?
Additionally, another thing I don't get about this is that by making this Novell deal, they seem to be indicating that they are willing to sue customers of other distros for patent infringement. But since when do CUSTOMERS get sued for patent infringement? Last I checked it was only the vendors of infringing products that could get sued for patent infringement.
Of course, the real WTF here is that it doesn't just show up as a USB disk drive when you plug it in, requiring no special drivers or software at all.
Every other damn player on the market works this way, and we have standards for a reason.
(Apparently that reason is so MS has something to point at and laugh about...)
Agreed. But there is a difference between short-term money and long-term money.
One doesn't guarantee the other, and it takes a good CEO to tell the difference.
It also takes intelligent shareholders to avoid lynching him in the meantime.
I think in a lot of cases, bad business decisions can be attributed to shareholders will very short-term vision.