"We find we need to re-load Windows XP often because of its vulnerabilities and instability."
I am not sure to whom your "we" refers. I know I am not in that set as the last time I had to reinstall Windows XP because of its vulnerabilities and instability was...well, never. Reboot, why yes I have had to do that countless times when a patch was pushed out. Hopefully that kind of architecture will be out the door with Vista. Until then I can live with reboots due to certain patches. This box has been running the XP SP2 install since, well, I installed SP2. I use it heavily every day to develop code, test, install and reinstall applications, and do my daily software development work. It could be that my company has a competent IT department, but I am sure a lot of it has to do with me not running as Administrator and not installing suspicious software or browsing suspicious websites when I am. My point being that with proper care and feeding an XP system does not need to be reinstalled often.
Out of simple laws can occur complex phenomena. Pointing out that complex phenomena occur does not mean that it must have come about by a force other than simple laws.
(This comment is not directed at you, but I found your piece inspiring and wanted to jump off to my tirade from your comment about religion being a matter of heart.)
I personally do not take issue with intelligent design per se. One may believe what they wish. You called it a matter of heart. I find that apt, because that's exactly the reason the state should not be teaching my children about an issue which is internal, personal, and dogmatic. Of course I sympathize with those who believe their children are being taught heresy when in science class learning about evolution, but how do the ID proponents feel when a Buddhist, Hindu, Asatru, or other child is exposed to the dogma he finds in conflict with everything he's learned from his parents?
Basically I model the ID debate as this oversimplified argument: evolutionists do not want their children taught what they see as religious gobly-gook; intelligent design proponents do not want their children taught what they see as scientific gobly-gook. The difference is that evolution has impartial--by which I mean independently verifiable--evidence whereas intelligent design does not. Hence, the state can vouch for evolution pretty cleanly with little resort to indoctrination. On the other hand, if this really gets out of hand, the state should not permit either to be taught, as to not offend anyone, and leave it up to the parents to educate their children in an inoffensive manner. Of course, then as a nation, we're pretty screwed, but that's the road we've chosen.
As for the counter argument that evolutionists want to have their cake and eat it too by keeping schools teaching their view but not permitting the intelligent design proponents to teach theirs, I still feel it comes down to the really big difference between the two. If intelligent design does not have independently verifiable evidence in support, and the state requires everyone to be taught this doctrine, then we are supporting a state which does just that--indoctrinate our children. I find that repulsive and essentially saying "God is great, we should convince everyone to accept his view as fact based solely on faith; oh, and no thinking for yourself, that's the work of the devil."
Well, that's not always true. There are places where both employees and employers are treated equally. If you worked at a company for less than six months, between six months and two years, more than two years determines how much transition time you get: two weeks, four weeks, six weeks, respectively. If you want to leave, you must give them that amount of time in notice. If they fire you, they must give you that amount of time in notice. To counteract the gimp employee who got fired and knows he's on the way out, the company can opt to pay you for the time up front and terminate your relationship immediately. Either way, it seems fair to me, and this is legally enforced by the government. Can anyone name this country?
Thank you for posting this, I was going to do it but decided to check if someone else beat me to it. Of course, you are only partially correct. When speaking of dictionaries, there are two types: descriptive and prescriptive. The OED is descriptive, it only describes the current state of the language as observed by readers. The other type, prescriptive, tries to dictate correct usage, although I'm not aware of any examples off hand. Not all dictionaries report on current usage, but the ones scholars use most do.
While I agree with you that social IM is out of hand and pointless a lot of the time, there are legitimate uses of IM which actually increase productivity. At work, we all have private offices. It's really nice to be able to be able to open up Office Communicator, see that your coworker is in a meeting or marked busy, and not bother him at the moment. Or to see they are available and logged into their machine, so you can send them a quick "Hey, I have the repro for that nasty crash bug on my test box, want to come down and look or should I send you a remote debug session?" without having to call them or walk down to their office. On an average day I send probably one or two messages, on a heavy day 10. I find that it actually helps me get my work done more efficiently. The windows don't pop over anything else I'm doing when I have an incoming message, the status indicator is very useful, and the fact that our IM ties in with the phone makes it so I can click a button and call the person without looking up their extension first.
The negatives you cite are valid reasons to abhor IM, but there might be some positives you are overlooking when you say "[h]onestly, I can't think of one good aspect of IM."
If indeed that is their plan. If their plan is to offer higher resolution video in the future but not offer a trade up program then it's in their best interests to *not* tell you that. They wouldn't want to canabilize current sales because people will be waiting for resolutions worth-while.
No, the state asked for *relevent* source code and programmers. I do not believe they meant that to include off the shelf commercial software. Diebold does not want to show their code to the state, so they came up with a pretty poorly masked excuse that they don't have access to all the code running on the box. But, they certainly do have access to the *relevent* code which runs on top of the irrelevent parts.
Now, you can argue that the underlying OS code *is* relevent, but that's a different issue. At issue here is how Diebold wrote their code since they do not provide a paper trail and it would be pretty trivial for them to hide malicious outcome changing code. A flaw in the underlying OS could pose a threat to the election but the first step is to look at the actual software running on top of it. I believe that's what the state wants to do.
Diebold apparently does not want to show their code, so they concocted a pretty pathetic excuse, and want to take their ball and go home. Well, that might just be better for the state. Hopefully though the state will not stop there and persue access to code which was used in prior elections, regardless of Diebold's desire to no longer provide their equipment for whatever their reason. Their equipment was used to decide the outcome of public elections and the public should be able to verify they were not defrauded. Let's hope the state continues the fight!
I wasn't talking about syntatical sugar. It was speaking to the fact that C is turing complete. While it's true that languages differ in the paradigms they enable and the paradigms they restrict, saying that because you can not have a particular structure available in another language does not mean that the equivalent cannot be accomplished. I do not see what having written a garbage collector for C has to do with any of this.
Have you *used* Microsoft accessibility features? Relying on commercial companies to ensure all applications are accessible *has* done the trick. Every application MS produces is accessible, as is the OS. More so than many other pieces of software commercial or otherwise. Why? Because one of their customers is the US government, and to sell to the government, you need to implement the government's standards of accessiblity.
Are you saying that instead of OOo implementing accessibility features properly people write a hack to do background conversion between OOo and MS formats. So when someone needing the accessibility features of say, Word, needs to edit a document in OOo format it gets converted to Word format, edited, and then converted back to OOo? Have converters ever been good enough to allow such conversion back and forth without introducing issues? What if two people need to edit a shared document over the course of a month and they make changes to it about 2-3 times per day. Do you think that such a heavily edited document would come out at the end of the month after 60 back and forth conversions with any sane structure? Or do you propose that every few conversions someone sit down and fix all the issues introduced?
I take it you have not been writing much C. Everything you describe above is possible. C does not do anything to discourage one from testing for invalid states. Did you perhaps mean that--unlike certain other languages--C does nothing to enforce or encourage one to validate the state?
I agree with your assesment of the pipeline "bug".
The buffer overflow was a bug, a very classic one. It was not the intention of the designers that an incoming packet could inject an executable payload.
I'm not talking about whether your conclusion is valid or not, but your argument is flawed. You make the assumption that once a trait has been achieved in a codebase it will remain without effort. I find that assumption incorrect. If I were to write a dependable piece of software today, I would still focus on dependability for the next release. Saying that I am does not imply that I feel I failed in the past.
They can be in multiple locations at the same time. So can we, in theory, although the probability is *very* tiny. Classical physics doesn't mention it because of the mass of the objects classical physics talks about. They're so massive that the chance of the phenomena reduces to zero and never factors into the equations. That's why classical equations can describe what you see so accurately on the large scale but fall apart on the small scale...there are terms we're missing. Also, the forces at work on that scale are different than gravity and E&M.
There is far more evidence that the universe as we know it self-organized than there is that it was created by some designer. If you want to read books about these topics, I suggest:
I admit you are correct about theology being different than literalism. I was just trying to express the concern that a lot of evidence in theology is non-empirical, apocryphal, and in general not based on verifiable facts.
As your concern about quantum physics and classical physics being at odds due to electron tunneling through potential wells and the like, it's not really all that at odds. The probability that an electron is in a location where the potential would not permit it under classical physics is inversley proportional to the mass of the particle. The more massive, the less likely the state forbidden by classical physics occurs. That's why its a phenomenon seen with electrons in atoms for example, and not people and walls.
Order comes from disorder is many different ways. There have been hundreds of experiments to show that this is the case. Go read a book about chaos theory...no not that one, a good in-depth one. Go ahead, I'll wait...
Okay, done? Good, now that we see that disorder can self-organize into order, we're half way there. Now we have to get you away from the idea that sometime improbable is impossible, because it isn't. Yes, if you do something that has a very small chance of producing an imporobable result a *lot* of times (say, on the order of the universe), it could happen. The probability is non-zero even if infintessimal.
I do not mean to at all attack your theology. I do have a bone to pick about your statement that 1) theology is a science and 2) science is not that different from religion.
The -ology suffix means the study of but it does not indicate the method of inquery. Just because biology and theology have the same suffix does not make their methods of inquery identical, or even similar.
Science uses empirical data to present theories about the world around us, theology does not. When geology and other sciences place the age of the earth as much older than the Bible says, theology debunks it by saying "you're wrong, the bible says right here that the world was created X thousand years ago, so clearly you're measurements are off". That's not the same method of supporting a theory as having some evidence for it.
Once again, I'm not arguing for or against a particular theology, just making a distinction between empirical evidence and gobblygook. I for one subscribe both to science and a religion-like philosophy which are not mutually exclusive and allows a spiritual and logical world to exist which does not warp my mind.
Do you have a better reference to where this has happened before? The article to which you linked mentions nothing about the government claiming the patent. It describes a case where the inventor was working with Lucent for them to use it on a government contract. After consulting with them for a year they refused to license the invention, so he filed a few lawsuits. Then after unsuccessful settlements, his partners went ahead with the suit. The government then stepped in and invoked their state secrets power to quash the attempts to subpeona documentation of the project citing national security concerns, effectively killing the suit. It mentioned nothing about the government claiming the patent. If it really has already happened, perhaps you could provide a reference to documentation suggesting that?
P.S. It's not that I agree with the events of that case, since many of the documents that were protected with the state secrets power were unclassified and publicly published already, so it seems that it was a favor done for a government contractor. The government could have invoked the power on only the classified documents or suggested closed-door hearings rather than an outright blanket quash like they did. The point here is not the details of how I feel about that case, but that it's not an example for what you implied it was.
XXXX |||\ command or meta-command digit 0 ||\- command or meta-command digit 1 |\-- command or meta-command digit 2 \--- 0: command; 1: meta-command
Commands XXX 000 Idle 001 Build road at nearest friendly location (that can support a road) 010 Build mine at nearest friendly location (that can support a mine) 011 Destroy nearest road ... etc
Meta-Commands XXX 000 End orders, return unit to unordered status 001 Repeat last command 010 Mark this location 011 Return to last marked location ... etc
Now you can encode you're secenario like this 1010000100111011 \__/\__/\__/\__/ | | | | | | | \-- Meta-command to return to marked command | | \------ Command to destroy a road | \---------- Command to build a road \-------------- Meta-command for mark
So it took all of 16-bits to encode those instructions, and they leave room for eight types of commands and eight types of meta-commands. Of course the encodings could be even more efficient, this was just an academic example. Also, adding another bit doubles the number of meanings you can give to each command.
So you see, many types of infinite data can be encoded with different methods, this being one example. Periodic functions are another. You are correct that there are certain infinite patterns that can not be encoded finitely without proposing a finite symbol that refers to the infinite quantity, such a Pi.
Alas, I think this was all moot since I think the english statement used to imply the command list was infinate was to not be taken literaly, but to be interpreted as meaning that there is no *arbitrary* limit to the length of the command list.
"See fanboys, it is okay to enjoy a varitey of games. You don't have to insult fpss' every chance you get because you would rather play a sucky dungeon crawler."
See fanboy, it is okay to ejoy a variety of games. You don't have to insult dungeon crawlers because you would rather play a FPS.
If you are right about Google releasing its own OS, then your argument about not being able to catch up must also apply unless we're talking about a double standard. Microsoft has had an OS that has such a firm grip in the market in terms of the sheer number of and number of mission critical applications developed for their platform that its too late for Google's OS to catch up.
I don't necessarily agree with my above statement. I'm just applying the same measuring stick you used for MS chasing Google in search to the Google chasing MS in the OS side of the competition.
In any case, it'll be interesting to see what happens in the next decade or so.
Slashdot Mirror
"We find we need to re-load Windows XP often because of its vulnerabilities and instability."
I am not sure to whom your "we" refers. I know I am not in that set as the last time I had to reinstall Windows XP because of its vulnerabilities and instability was...well, never. Reboot, why yes I have had to do that countless times when a patch was pushed out. Hopefully that kind of architecture will be out the door with Vista. Until then I can live with reboots due to certain patches. This box has been running the XP SP2 install since, well, I installed SP2. I use it heavily every day to develop code, test, install and reinstall applications, and do my daily software development work. It could be that my company has a competent IT department, but I am sure a lot of it has to do with me not running as Administrator and not installing suspicious software or browsing suspicious websites when I am. My point being that with proper care and feeding an XP system does not need to be reinstalled often.
Out of simple laws can occur complex phenomena. Pointing out that complex phenomena occur does not mean that it must have come about by a force other than simple laws.
(This comment is not directed at you, but I found your piece inspiring and wanted to jump off to my tirade from your comment about religion being a matter of heart.)
I personally do not take issue with intelligent design per se. One may believe what they wish. You called it a matter of heart. I find that apt, because that's exactly the reason the state should not be teaching my children about an issue which is internal, personal, and dogmatic. Of course I sympathize with those who believe their children are being taught heresy when in science class learning about evolution, but how do the ID proponents feel when a Buddhist, Hindu, Asatru, or other child is exposed to the dogma he finds in conflict with everything he's learned from his parents?
Basically I model the ID debate as this oversimplified argument: evolutionists do not want their children taught what they see as religious gobly-gook; intelligent design proponents do not want their children taught what they see as scientific gobly-gook. The difference is that evolution has impartial--by which I mean independently verifiable--evidence whereas intelligent design does not. Hence, the state can vouch for evolution pretty cleanly with little resort to indoctrination. On the other hand, if this really gets out of hand, the state should not permit either to be taught, as to not offend anyone, and leave it up to the parents to educate their children in an inoffensive manner. Of course, then as a nation, we're pretty screwed, but that's the road we've chosen.
As for the counter argument that evolutionists want to have their cake and eat it too by keeping schools teaching their view but not permitting the intelligent design proponents to teach theirs, I still feel it comes down to the really big difference between the two. If intelligent design does not have independently verifiable evidence in support, and the state requires everyone to be taught this doctrine, then we are supporting a state which does just that--indoctrinate our children. I find that repulsive and essentially saying "God is great, we should convince everyone to accept his view as fact based solely on faith; oh, and no thinking for yourself, that's the work of the devil."
Well, that's not always true. There are places where both employees and employers are treated equally. If you worked at a company for less than six months, between six months and two years, more than two years determines how much transition time you get: two weeks, four weeks, six weeks, respectively. If you want to leave, you must give them that amount of time in notice. If they fire you, they must give you that amount of time in notice. To counteract the gimp employee who got fired and knows he's on the way out, the company can opt to pay you for the time up front and terminate your relationship immediately. Either way, it seems fair to me, and this is legally enforced by the government. Can anyone name this country?
Thank you for posting this, I was going to do it but decided to check if someone else beat me to it. Of course, you are only partially correct. When speaking of dictionaries, there are two types: descriptive and prescriptive. The OED is descriptive, it only describes the current state of the language as observed by readers. The other type, prescriptive, tries to dictate correct usage, although I'm not aware of any examples off hand. Not all dictionaries report on current usage, but the ones scholars use most do.
I think it's suspicious; you think it's suspicious. Somehow I don't believe that the average IM user would be so wise.
While I agree with you that social IM is out of hand and pointless a lot of the time, there are legitimate uses of IM which actually increase productivity. At work, we all have private offices. It's really nice to be able to be able to open up Office Communicator, see that your coworker is in a meeting or marked busy, and not bother him at the moment. Or to see they are available and logged into their machine, so you can send them a quick "Hey, I have the repro for that nasty crash bug on my test box, want to come down and look or should I send you a remote debug session?" without having to call them or walk down to their office. On an average day I send probably one or two messages, on a heavy day 10. I find that it actually helps me get my work done more efficiently. The windows don't pop over anything else I'm doing when I have an incoming message, the status indicator is very useful, and the fact that our IM ties in with the phone makes it so I can click a button and call the person without looking up their extension first.
The negatives you cite are valid reasons to abhor IM, but there might be some positives you are overlooking when you say "[h]onestly, I can't think of one good aspect of IM."
If indeed that is their plan. If their plan is to offer higher resolution video in the future but not offer a trade up program then it's in their best interests to *not* tell you that. They wouldn't want to canabilize current sales because people will be waiting for resolutions worth-while.
No, the state asked for *relevent* source code and programmers. I do not believe they meant that to include off the shelf commercial software. Diebold does not want to show their code to the state, so they came up with a pretty poorly masked excuse that they don't have access to all the code running on the box. But, they certainly do have access to the *relevent* code which runs on top of the irrelevent parts.
Now, you can argue that the underlying OS code *is* relevent, but that's a different issue. At issue here is how Diebold wrote their code since they do not provide a paper trail and it would be pretty trivial for them to hide malicious outcome changing code. A flaw in the underlying OS could pose a threat to the election but the first step is to look at the actual software running on top of it. I believe that's what the state wants to do.
Diebold apparently does not want to show their code, so they concocted a pretty pathetic excuse, and want to take their ball and go home. Well, that might just be better for the state. Hopefully though the state will not stop there and persue access to code which was used in prior elections, regardless of Diebold's desire to no longer provide their equipment for whatever their reason. Their equipment was used to decide the outcome of public elections and the public should be able to verify they were not defrauded. Let's hope the state continues the fight!
I wasn't talking about syntatical sugar. It was speaking to the fact that C is turing complete. While it's true that languages differ in the paradigms they enable and the paradigms they restrict, saying that because you can not have a particular structure available in another language does not mean that the equivalent cannot be accomplished. I do not see what having written a garbage collector for C has to do with any of this.
Have you *used* Microsoft accessibility features? Relying on commercial companies to ensure all applications are accessible *has* done the trick. Every application MS produces is accessible, as is the OS. More so than many other pieces of software commercial or otherwise. Why? Because one of their customers is the US government, and to sell to the government, you need to implement the government's standards of accessiblity.
Are you saying that instead of OOo implementing accessibility features properly people write a hack to do background conversion between OOo and MS formats. So when someone needing the accessibility features of say, Word, needs to edit a document in OOo format it gets converted to Word format, edited, and then converted back to OOo? Have converters ever been good enough to allow such conversion back and forth without introducing issues? What if two people need to edit a shared document over the course of a month and they make changes to it about 2-3 times per day. Do you think that such a heavily edited document would come out at the end of the month after 60 back and forth conversions with any sane structure? Or do you propose that every few conversions someone sit down and fix all the issues introduced?
I take it you have not been writing much C. Everything you describe above is possible. C does not do anything to discourage one from testing for invalid states. Did you perhaps mean that--unlike certain other languages--C does nothing to enforce or encourage one to validate the state?
I agree with your assesment of the pipeline "bug".
The buffer overflow was a bug, a very classic one. It was not the intention of the designers that an incoming packet could inject an executable payload.
I'm not talking about whether your conclusion is valid or not, but your argument is flawed. You make the assumption that once a trait has been achieved in a codebase it will remain without effort. I find that assumption incorrect. If I were to write a dependable piece of software today, I would still focus on dependability for the next release. Saying that I am does not imply that I feel I failed in the past.
When it is not announce as product, it's not really vaporware. It's not even a ware. Do you even know what Microsoft Research is and what they do?
They can be in multiple locations at the same time. So can we, in theory, although the probability is *very* tiny. Classical physics doesn't mention it because of the mass of the objects classical physics talks about. They're so massive that the chance of the phenomena reduces to zero and never factors into the equations. That's why classical equations can describe what you see so accurately on the large scale but fall apart on the small scale...there are terms we're missing. Also, the forces at work on that scale are different than gravity and E&M.
There is far more evidence that the universe as we know it self-organized than there is that it was created by some designer. If you want to read books about these topics, I suggest:
e rse
Self-Organization: http://en.wikipedia.org/wiki/Self-organization, especially the bibliography.
For how the raw materials came to be, you should read up on the big bang and cosmology in general.
Start here: http://en.wikipedia.org/wiki/Timeline_of_the_Univ
Then read about how the heavier elements got created, but I don't have any good sources right now.
I admit you are correct about theology being different than literalism. I was just trying to express the concern that a lot of evidence in theology is non-empirical, apocryphal, and in general not based on verifiable facts.
As your concern about quantum physics and classical physics being at odds due to electron tunneling through potential wells and the like, it's not really all that at odds. The probability that an electron is in a location where the potential would not permit it under classical physics is inversley proportional to the mass of the particle. The more massive, the less likely the state forbidden by classical physics occurs. That's why its a phenomenon seen with electrons in atoms for example, and not people and walls.
Order comes from disorder is many different ways. There have been hundreds of experiments to show that this is the case. Go read a book about chaos theory...no not that one, a good in-depth one. Go ahead, I'll wait...
Okay, done? Good, now that we see that disorder can self-organize into order, we're half way there. Now we have to get you away from the idea that sometime improbable is impossible, because it isn't. Yes, if you do something that has a very small chance of producing an imporobable result a *lot* of times (say, on the order of the universe), it could happen. The probability is non-zero even if infintessimal.
I do not mean to at all attack your theology. I do have a bone to pick about your statement that 1) theology is a science and 2) science is not that different from religion.
The -ology suffix means the study of but it does not indicate the method of inquery. Just because biology and theology have the same suffix does not make their methods of inquery identical, or even similar.
Science uses empirical data to present theories about the world around us, theology does not. When geology and other sciences place the age of the earth as much older than the Bible says, theology debunks it by saying "you're wrong, the bible says right here that the world was created X thousand years ago, so clearly you're measurements are off". That's not the same method of supporting a theory as having some evidence for it.
Once again, I'm not arguing for or against a particular theology, just making a distinction between empirical evidence and gobblygook. I for one subscribe both to science and a religion-like philosophy which are not mutually exclusive and allows a spiritual and logical world to exist which does not warp my mind.
Do you have a better reference to where this has happened before? The article to which you linked mentions nothing about the government claiming the patent. It describes a case where the inventor was working with Lucent for them to use it on a government contract. After consulting with them for a year they refused to license the invention, so he filed a few lawsuits. Then after unsuccessful settlements, his partners went ahead with the suit. The government then stepped in and invoked their state secrets power to quash the attempts to subpeona documentation of the project citing national security concerns, effectively killing the suit. It mentioned nothing about the government claiming the patent. If it really has already happened, perhaps you could provide a reference to documentation suggesting that?
P.S. It's not that I agree with the events of that case, since many of the documents that were protected with the state secrets power were unclassified and publicly published already, so it seems that it was a favor done for a government contractor. The government could have invoked the power on only the classified documents or suggested closed-door hearings rather than an outright blanket quash like they did. The point here is not the details of how I feel about that case, but that it's not an example for what you implied it was.
There's a more clever way to encode this.
XXXX
|||\ command or meta-command digit 0
||\- command or meta-command digit 1
|\-- command or meta-command digit 2
\--- 0: command; 1: meta-command
Commands
XXX
000 Idle
001 Build road at nearest friendly location (that can support a road)
010 Build mine at nearest friendly location (that can support a mine)
011 Destroy nearest road
... etc
Meta-Commands
XXX
000 End orders, return unit to unordered status
001 Repeat last command
010 Mark this location
011 Return to last marked location
... etc
Now you can encode you're secenario like this
1010000100111011
\__/\__/\__/\__/
| | | |
| | | \-- Meta-command to return to marked command
| | \------ Command to destroy a road
| \---------- Command to build a road
\-------------- Meta-command for mark
So it took all of 16-bits to encode those instructions, and they leave room for eight types of commands and eight types of meta-commands. Of course the encodings could be even more efficient, this was just an academic example. Also, adding another bit doubles the number of meanings you can give to each command.
So you see, many types of infinite data can be encoded with different methods, this being one example. Periodic functions are another. You are correct that there are certain infinite patterns that can not be encoded finitely without proposing a finite symbol that refers to the infinite quantity, such a Pi.
Alas, I think this was all moot since I think the english statement used to imply the command list was infinate was to not be taken literaly, but to be interpreted as meaning that there is no *arbitrary* limit to the length of the command list.
"See fanboys, it is okay to enjoy a varitey of games. You don't have to insult fpss' every chance you get because you would rather play a sucky dungeon crawler."
See fanboy, it is okay to ejoy a variety of games. You don't have to insult dungeon crawlers because you would rather play a FPS.
If you are right about Google releasing its own OS, then your argument about not being able to catch up must also apply unless we're talking about a double standard. Microsoft has had an OS that has such a firm grip in the market in terms of the sheer number of and number of mission critical applications developed for their platform that its too late for Google's OS to catch up.
I don't necessarily agree with my above statement. I'm just applying the same measuring stick you used for MS chasing Google in search to the Google chasing MS in the OS side of the competition.
In any case, it'll be interesting to see what happens in the next decade or so.