Any statistical study with political conclusion should include the leanings of those conducting the study.
We all know how easy it is for biased people to cook the books and dress up the numbers. Just look at wall street this week.
In my own personal experience, workaholics tend to be conservative. Most liberals I know sit around collecting disability checks complaining about big evil corporations all day. Hey, maybe I can produce a biased study "proving" my preconceptions, too.
Fat chance. The nice guy theories are excuses for losers that need something to cling to to feel better about themselves.
Oddly enough; many cultures already attempted to compensate for these inherent flaws with traditions like social classes, arranged marriages, contingent inheritence and selective suffrage.
The problem is that several liberal ideals have effectively trumped all forms of reason when discussing social and family issues. And the odd part is western scientists and philosophical thinkers have for decades bought in as accomplices to this no-holds barred, girls-gone-wild, every girl a slut, every child fatherless products of teen pregnancy pacts we are devolving into.
The odd paradox is that as technology advances, it also burgeons a new geek culture to maintain itself. But the traits promoted by a successful geek career leave them less capable of playing the role of the narcissitic bad boy necessary for reproductive success. And the ranks of female geeks is so negligable. Even if they do produce offspring; they are hard pressed to compete against rock and movie stars having sex with 5000 women over a career fathering dozens of children with strings of women across the country. The end result is a world full of technology noone understands eventually. Kind of scary.
I envision something similar to the dark terminator future in the movies 8-) -- j/k.
I don't see how you can say the US "forced" China to ally with the USSR without failing to understand communism and modern history.
A big part of marxism is the idea of inevitable revolution. The US, with its emphasis on freedom and traditional laissez-faire tendencies basically represents a bouguosie state in the eyes of the Marxist. In that sense, almost by definition, the governments of the United States and the West were the enemy. For the revolution to take the form that Marx predicted; bouguosie governments must fall to give way to the universal rule of one world proletariate utopia. Your assessment of America's fear (whether justified or not) of a unified communist menace is not the fault of the United States. You just have to read a few paragraphs of the communist and marxist playbooks to realize that. It wasn't any over-idealism or parranoia on our part at work in the Cold War. We simply took them at their word.
On top of this, there are the obvious similarities in tactics that party leaders in communist governments share. They all seem without question to align themselves with Stalinist tactics to maintain control. This totalitarian mindset makes it naturally easier for these governments (Cuba, North Korea, China, etc...) to relate to each other. I think Mao would have a lot more to discuss with Stalin than he would with Roosevelt and Eisenhour simply from a philosophical perspective. For that reason alone, they were natural allies.
Besides, the US historically invested in a free China before and during World War II. We had Flying Tiger squadrons fighting the Japanese in China for Chang Kai Shek throughout the war. What were we supposed to do, abandon these people? For what?
Not to mention the American blood spilled in the Korean conflict when the Chinese crossed the river into Korea much to the surprise of General Macarthur.
Mao was not the innocent victim of the United States that you depict. He was responsible for 10's of millions of mass murders during the Cultural Revolution. He learned quite well from Stalin's butchery. The US should be MORE selective about the people we jump into bed with, not less.
Why? You must be confusing XML with DOM. In fact, XML!=DOM.
Expat works just fine for streams. While the folks at Sun/apache found an infinite number of ways to wrap Expat so that it no longer works with streams; it doesn't mean YOU have to be a lemming and just use whatever crap is most available. Simply wrap Expat using the JNI. You get the speed of C in your comfortable little pointerless java womb.
Hey I love linux. I deploy telecommunications applications that run on the platform. In many ways, it is better suited for my needs than Windows. It really is ground zero for application, shell and web scripting development. But come on, do we really need all eleventy billion packages? How many do you actually use on a regular basis? This ever growing package collection strikes me of jack-of-all-trades, master on none.
Probably the most powerful thing about linux is the shell script (BASH, shell, PERL, etc..) The bottom line is that these things (and largely Linux) are exclusively for GEEKS; and there simply are far fewer geeks buying and using computers and software than grandmothers and sales managers. And geeks don't generally don't like to spend ANY money on software so they will NEVER make you rich. Linux cannot escape this reality.
Besides, Mac and Windows have always done better at marketing and selling software to the masses. Mac has always been the user-friendly king. Linux just can't shine a candle to Mac in this regard. And windows is ground zero for cutting edge hardware development by virtue of its huge install base and widespread adoption by the business community (largely because of its OEM relationship with IBM back in the day). Because of this, windows will ALWAYS have the gaming industry in its back-pocket. Let's face it, Linux has a backseat with low-level development. Drivers get ported from windows to linux as a second priority often by the second-string development team.
So linux is left the utility knife of operating systems. Its ever growing amorphous blob of packages reducing it every worsening mediocrity.
I wouldn't dismiss this one quite so easily. Our number system is simple enough to explain to any young child; but its applications are limitless.
This event served to give credence to the concept of autonomous systems; possibly contributing to the expanding acceptance of the US military for the development of unmanned drones. The success of these programs of late is already expanding into the army as we speak. Chess isn't the only zero-sum game that can be fed to a computer as a simple model; and intuition can be counterbalanced by fear, anxiety and doubt - all of which being completely absent in a machine.
Do we need to chase after the holy grail of "True AI"? This is an interesting academic problem; but I don't believe this is truly necessary to producing viable autonomous solutions.
To be honest, the real threat posed by information technology is the softening of the population at large as a result of offloading problem solving and critical thinking over to machines. The internet has spawned an epidemic of cheating. You can probably get a high school degree today without making a single cogent arguement of your own. Calculators are allowed in the earliest math classes. Machining is now entirely automated. Mechanical knowledge is no longer a prerequisite for working in most factories. Finding a mechanic that is capable of trouble shooting problems (rather than just plugging your car into a computer and replacing a part) is getting more and more difficult. Navigation is universally offloaded to GPS satellites. Purchases are made almost entirely electronically using credit cards. Try giving your next register operator change enough to get a full dollar back; and you are likely to get an entirely random result. There is even talk of developing autonomous cars to eliminate human error in driving.
Sure, alot of people I email use gmail. But they don't receive my bank statements, invoices, newsgroup replies and political newletters.
So google gets to mine my bloviating about stories from "back in the day", comments about bands, and ideas for the next road trip. Nothing I'm worried about really.
Plus, if I went to webmail, I wouldn't be able to email customers to discuss private issues using public-key encryption. Would you give a webserver access to your private key just to read a message? I wouldn't.
How about support for public-key encryption and message signing?
I have to support systems that process private information. My customers can't even give me details about the problem without encryption. Who wants to give a web server access to your private key so you can read a message?
Would you use a phone service if you knew all of your phone calls could exist on an internet connected computer indefinitely for some geek hacker to browse through and maybe post on YouTube?
Even worse, email (which takes up considerably less space and can be compressed to single digit percentages) is prime for third parties to resell to marketing and collection companies. They can mine the data to figure out what books you've ordered from amazon and barnes and noble. Determine which political internet sites or newsgroups you subscribe to. Analyze your buying habits. Mine for personal information to resell to identity thieves for a profit. It may make you feel good to trust that you abide by the law and have nothing to hide; but not everyone does the same. Seemingly innocuous information can be used for evil purposes like identity theft or political descrimination.
Databases, like every technology devised by man, can be utilized for good or ill. Your right to privacy is a valuable part of your ability to persue happiness undeterred. Don't let big corporations or the government take that away without a fight.
Please don't take this personally; but that excuse strikes me as cowardice.
How many things does Microsoft invest in at a loss. MSN, XBox, Zune, WinCE, the list goes on. Now, the Xbox has a chance of breaking out against Sony. You've got to make bold moves in this industry. Making MacOS available on the open PC market would be a great move strategically. Maybe the numbers don't make sense now; but there are other benefits.
1) Politically, from an investors perspective, this move would be bigger than the IPhone. It would really shake things up. Mac OS X would become a direct competitor to windows. It is not today. That is a significant shift not to be overlooked. The buzz generated by this move (along with past momentum over the past couple of years) should flood AAPL's coffers with investment money.
2) Whenever you crunch numbers in a projection, you make certain assumptions. I think the assumptions are too conservative. I think they fail to take into account the custom PC market. There is more than one class of users in PCs. You have the basic user; uninterested in experimenation who uses the computer like an appliance. Right now, this is mainly the market of AAPL to date. Making this move would certainly appeal to my class of user. The one who experiments with Linux and Windows and is looking to find the most efficient and stable platform to develop and prototype solutions for their customers. It is not the run of the mill user that makes architecture decisions in automated business computing enviroments. It is the solution providers they contract with. These providers spec out servers with complex peripheral hardware like IP Trunking and SIP/RTP telephony boards. You have to remember the roots of PCs. In the 70's (before my time), PC users were alot like HAM radio enthusiasts. These people build game systems today. They are becoming increasing frustrated with windows. They desperately need a third alternative. An operating environment that leverages the flexibility of Linux with a robust and probably superior GUI is an outright dream for these people. They would actually want to participate in building and porting drivers for powerful peripherals like telephony and audio/video signal processing hardware if it could lead to systems that outperform what is presently in the field. Once you convince these guys; the market share problem will fix itself.
Projections are only as good as the assumptions that they make. I think AAPL would be well served to reinvestigate the possiblity of opening up the OS X to support broader open hardware manufacturers. I stand by my perspective on this; and envision a computing world unshackled by the status quo.
Spot on. Web 2.0 is a joke that only San Fran technocrats take seriously. This guy is right for the wrong reasons. And I don't get how someone can redefine "Dead". Suddently "dead" means non-threatening. This strikes of Clinton redefining "Is". This is obviously a deliberate mischoice of words to attract eyeballs.
Lets put things in perspective here. To find Microsoft's next competitor, you have to look at market caps a full 1/2 the size of M$. And Google (valued at 146 billion), is trading at 50 times earnings compared with microsoft's 26. More importantly, it isn't even in the same subsector. Google's stuff runs mostly on windows. They have no operatings system. Microsoft is still for all intents and purposes a monopoly. If you look at operting systems alone, Apple is less than 1/3 the size of Microsoft and Redhat was recently delisted.
On another note, in my opinion, Google is way way overvalued. The economy is slowing-down. Advertising is the first thing companies cut. You can't maintain 100% growth rate forever.
Microsoft is dead because they stop making meaningful innovations (just after the release of windows 2000 I surmise). And OS X is irrelevant looking at the numbers. If they stopped playing IBM's old closed architecture game, and leveraged the manufacturing power of the near east by allowing vendors like HP, Dell, Toshiba, and Sony build machines that run OS X; maybe we could have a duopoly in the Operating System market. Then either microsoft will start innovating for its very survival or apple will become the new industry leader. Either way, the consumer wins. Until then, its status quo for the forseeable future.
I think if Apple can abandon their closed architecture model; they can really eat into the Microsoft monopoly.
Vista is a pig. I've seen it take 30-40 seconds to open up the Save-As dialog from every application. The Visual Studio 2005 IDE was a big flop in terms of Vista compatibility. There is no better time than now.
Could you imagine if Toshiba, HP, Dell, and others could offer laptops and workstations running OS X. We'd actually have a real duopoly in PC Operating Systems. The game would be afoot in competitive terms. It would be AMD and Intel all over again. Microsoft would actually be forced to start producing again. Just think of the innovations they would have to come up with to stay competitive. That's what the industry needs.
Of course, the software lag as far as cross-platform development has always been a major problem. But you cannot begin to tackle this one as a closed architecture.
The truth is that M$ has pretty muched sucked since 2000. It took them a whopping 5 years for them to get XP working properly.
And Vista/Visual Studio 2005 is pretty much a train wreck for C developers. We used to be able to rely on the development environment. In fact, that area was always a significant innovation for these guys. No more. Fire Steve Embalmer before it is too late.
I was going to respond to the inanely academic response about pointer reuse; but the post I am responding to hits the nail on the head.
People invent hundreds and thousands of non-trivial mechanisms when writing large complex software systems every year. Some of these projects are classified. This patent is meaningless. You shouldn't be able to patent an algorithm. It is only slightly more nonsensical than patenting a line of code or a function. Stop waisting taxpayer money.
Only systems should be patentable.
Yeah, its pretty obvious that next recession; that 60 times multiple is gauranteed to disintegrate 90's style. We are already at the tail end of one of the longest continuous bull markets in history. Global tensions are heating up. Oil prices are through the roof. 13% and growing of companies are missing expectations. It is only a matter of time.
When the economy goes sour; cyclical stuff like advertising is the first to go out the window. If you are smart and you own google now; I'd wait out this rally until the next time it pulls back. Then take your profits and buy a nice safe CD. You can can get 5-6% APR guaranteed thanks to Bernanke's rate raising.
Wait things out and buy back in when everything blows over in a few years.
ALPHA software is freshly implemented, in the process of being Unit and Integration tested in the lab by the developers and QA.
BETA simply means the software is pre-released to the general public. No test set can entirely simulate production in most cases; so every complex software system benefits from being exposed to courageous testers outside of the lab.
The problem is that the fast food management styles coming out of our business schools are incompatible with software QA. Testing gets a back seat to basically everything on the management agenda. Stuff gets designed without a test plans. A short half-hearted alpha testing phase is rushed over. It then goes immediatly to BETA and stays there for the life of the product. Hence your aversion to the term BETA, which I believe is quite justified.
As far as I am concerned, smart pointers have become a decisive breakthrough in software development. It basically removes the garbage collection advantage of VIMs.
Intelligent application of smart pointers, including extensions of the basic std::auto_ptr to include smart arrays and reference counting gives C++ programmers all the benefits of garbage collection without dedicated processes wasting MIPS monitoring your code.
All of this VM background bookkeeping is not free; and is above all not perfect. I have witnessed first hand cases of the Microsoft VM taking hours to cleanup wrapped resources for DLLs I've developed. Just because the language prevents you from worrying about freeing heap blocks and handles doesn't mean it gets done right. And worse, if there is a bug in the VM; one's only recourse is to guess at a possible work around and hope it triggers the GC to do its job.
-James
This anti-Java poster is right on the money. This language tower of babel is destroying software.
Here is the source (Compiled in MSVC6):
#ifdef WIN32 #define HYPER __int64 #else #define HYPER long long #endif
HYPER mxy( long x, long y ) {
long shift = 1;
char iShift = 0;
char bNeg = 0x0;
if( x==0 || y==0 )
return 0;
else if( x0 )
{
++bNeg;
x = -x;
}
for(;(shift1)x;shift=shift1 )
++iShift;
if( bNeg )
return -( mxy( x-shift, y ) + ((HYPER)yiShift) );
else
return mxy( x-shift, y ) + ((HYPER)yiShift); }
This problem is actually pretty cool (at least I think so). As far as relevance, I have seen at least one DSP processor without an integer divide. Not certain about multiply.
My point is that java programmers look down on a problem like this as beneath them somehow. While C and assembly programmers love the challenge (or at least I do). This foolish quote about Comp Sci not being about computers is self serving. It was obviously written by some over-hyped professor whose job hinges on writing books and lectures rather than effective software.
Fact is I'd rather travel to Mars on a spacecraft whose software was written by C and assembly programmers than a bunch of academic java guys too smart to understand the underlying hardware.
Slashdot Mirror
Any statistical study with political conclusion should include the leanings of those conducting the study. We all know how easy it is for biased people to cook the books and dress up the numbers. Just look at wall street this week. In my own personal experience, workaholics tend to be conservative. Most liberals I know sit around collecting disability checks complaining about big evil corporations all day. Hey, maybe I can produce a biased study "proving" my preconceptions, too.
Fat chance. The nice guy theories are excuses for losers that need something to cling to to feel better about themselves. Oddly enough; many cultures already attempted to compensate for these inherent flaws with traditions like social classes, arranged marriages, contingent inheritence and selective suffrage. The problem is that several liberal ideals have effectively trumped all forms of reason when discussing social and family issues. And the odd part is western scientists and philosophical thinkers have for decades bought in as accomplices to this no-holds barred, girls-gone-wild, every girl a slut, every child fatherless products of teen pregnancy pacts we are devolving into. The odd paradox is that as technology advances, it also burgeons a new geek culture to maintain itself. But the traits promoted by a successful geek career leave them less capable of playing the role of the narcissitic bad boy necessary for reproductive success. And the ranks of female geeks is so negligable. Even if they do produce offspring; they are hard pressed to compete against rock and movie stars having sex with 5000 women over a career fathering dozens of children with strings of women across the country. The end result is a world full of technology noone understands eventually. Kind of scary. I envision something similar to the dark terminator future in the movies 8-) -- j/k.
I don't see how you can say the US "forced" China to ally with the USSR without failing to understand communism and modern history.
A big part of marxism is the idea of inevitable revolution. The US, with its emphasis on freedom and traditional laissez-faire tendencies basically represents a bouguosie state in the eyes of the Marxist. In that sense, almost by definition, the governments of the United States and the West were the enemy. For the revolution to take the form that Marx predicted; bouguosie governments must fall to give way to the universal rule of one world proletariate utopia. Your assessment of America's fear (whether justified or not) of a unified communist menace is not the fault of the United States. You just have to read a few paragraphs of the communist and marxist playbooks to realize that. It wasn't any over-idealism or parranoia on our part at work in the Cold War. We simply took them at their word.
On top of this, there are the obvious similarities in tactics that party leaders in communist governments share. They all seem without question to align themselves with Stalinist tactics to maintain control. This totalitarian mindset makes it naturally easier for these governments (Cuba, North Korea, China, etc...) to relate to each other. I think Mao would have a lot more to discuss with Stalin than he would with Roosevelt and Eisenhour simply from a philosophical perspective. For that reason alone, they were natural allies.
Besides, the US historically invested in a free China before and during World War II. We had Flying Tiger squadrons fighting the Japanese in China for Chang Kai Shek throughout the war. What were we supposed to do, abandon these people? For what?
Not to mention the American blood spilled in the Korean conflict when the Chinese crossed the river into Korea much to the surprise of General Macarthur.
Mao was not the innocent victim of the United States that you depict. He was responsible for 10's of millions of mass murders during the Cultural Revolution. He learned quite well from Stalin's butchery. The US should be MORE selective about the people we jump into bed with, not less.
XML not good for unbounded streams?
Why? You must be confusing XML with DOM. In fact, XML!=DOM.
Expat works just fine for streams. While the folks at Sun/apache found an infinite number of ways to wrap Expat so that it no longer works with streams; it doesn't mean YOU have to be a lemming and just use whatever crap is most available. Simply wrap Expat using the JNI. You get the speed of C in your comfortable little pointerless java womb.
James
Beverly, MA
Hey I love linux. I deploy telecommunications applications that run on the platform. In many ways, it is better suited for my needs than Windows. It really is ground zero for application, shell and web scripting development. But come on, do we really need all eleventy billion packages? How many do you actually use on a regular basis? This ever growing package collection strikes me of jack-of-all-trades, master on none.
Probably the most powerful thing about linux is the shell script (BASH, shell, PERL, etc..) The bottom line is that these things (and largely Linux) are exclusively for GEEKS; and there simply are far fewer geeks buying and using computers and software than grandmothers and sales managers. And geeks don't generally don't like to spend ANY money on software so they will NEVER make you rich. Linux cannot escape this reality.
Besides, Mac and Windows have always done better at marketing and selling software to the masses. Mac has always been the user-friendly king. Linux just can't shine a candle to Mac in this regard. And windows is ground zero for cutting edge hardware development by virtue of its huge install base and widespread adoption by the business community (largely because of its OEM relationship with IBM back in the day). Because of this, windows will ALWAYS have the gaming industry in its back-pocket. Let's face it, Linux has a backseat with low-level development. Drivers get ported from windows to linux as a second priority often by the second-string development team.
So linux is left the utility knife of operating systems. Its ever growing amorphous blob of packages reducing it every worsening mediocrity.
James
Beverly, MA
I wouldn't dismiss this one quite so easily. Our number system is simple enough to explain to any young child; but its applications are limitless.
This event served to give credence to the concept of autonomous systems; possibly contributing to the expanding acceptance of the US military for the development of unmanned drones. The success of these programs of late is already expanding into the army as we speak. Chess isn't the only zero-sum game that can be fed to a computer as a simple model; and intuition can be counterbalanced by fear, anxiety and doubt - all of which being completely absent in a machine.
Do we need to chase after the holy grail of "True AI"? This is an interesting academic problem; but I don't believe this is truly necessary to producing viable autonomous solutions.
To be honest, the real threat posed by information technology is the softening of the population at large as a result of offloading problem solving and critical thinking over to machines. The internet has spawned an epidemic of cheating. You can probably get a high school degree today without making a single cogent arguement of your own. Calculators are allowed in the earliest math classes. Machining is now entirely automated. Mechanical knowledge is no longer a prerequisite for working in most factories. Finding a mechanic that is capable of trouble shooting problems (rather than just plugging your car into a computer and replacing a part) is getting more and more difficult. Navigation is universally offloaded to GPS satellites. Purchases are made almost entirely electronically using credit cards. Try giving your next register operator change enough to get a full dollar back; and you are likely to get an entirely random result. There is even talk of developing autonomous cars to eliminate human error in driving.
I disagree.
Sure, alot of people I email use gmail. But they don't receive my bank statements, invoices, newsgroup replies and political newletters.
So google gets to mine my bloviating about stories from "back in the day", comments about bands, and ideas for the next road trip. Nothing I'm worried about really.
Plus, if I went to webmail, I wouldn't be able to email customers to discuss private issues using public-key encryption. Would you give a webserver access to your private key just to read a message? I wouldn't.
Cheers
.
One thing noone mentioned...
How about support for public-key encryption and message signing?
I have to support systems that process private information. My customers can't even give me details about the problem without encryption. Who wants to give a web server access to your private key so you can read a message?
That's just ludicrous.
Would you use a phone service if you knew all of your phone calls could exist on an internet connected computer indefinitely for some geek hacker to browse through and maybe post on YouTube?
Even worse, email (which takes up considerably less space and can be compressed to single digit percentages) is prime for third parties to resell to marketing and collection companies. They can mine the data to figure out what books you've ordered from amazon and barnes and noble. Determine which political internet sites or newsgroups you subscribe to. Analyze your buying habits. Mine for personal information to resell to identity thieves for a profit. It may make you feel good to trust that you abide by the law and have nothing to hide; but not everyone does the same. Seemingly innocuous information can be used for evil purposes like identity theft or political descrimination.
Databases, like every technology devised by man, can be utilized for good or ill. Your right to privacy is a valuable part of your ability to persue happiness undeterred. Don't let big corporations or the government take that away without a fight.
jrentona
Beverly, MA
Please don't take this personally; but that excuse strikes me as cowardice.
How many things does Microsoft invest in at a loss. MSN, XBox, Zune, WinCE, the list goes on. Now, the Xbox has a chance of breaking out against Sony. You've got to make bold moves in this industry. Making MacOS available on the open PC market would be a great move strategically. Maybe the numbers don't make sense now; but there are other benefits.
1) Politically, from an investors perspective, this move would be bigger than the IPhone. It would really shake things up. Mac OS X would become a direct competitor to windows. It is not today. That is a significant shift not to be overlooked. The buzz generated by this move (along with past momentum over the past couple of years) should flood AAPL's coffers with investment money.
2) Whenever you crunch numbers in a projection, you make certain assumptions. I think the assumptions are too conservative. I think they fail to take into account the custom PC market. There is more than one class of users in PCs. You have the basic user; uninterested in experimenation who uses the computer like an appliance. Right now, this is mainly the market of AAPL to date. Making this move would certainly appeal to my class of user. The one who experiments with Linux and Windows and is looking to find the most efficient and stable platform to develop and prototype solutions for their customers. It is not the run of the mill user that makes architecture decisions in automated business computing enviroments. It is the solution providers they contract with. These providers spec out servers with complex peripheral hardware like IP Trunking and SIP/RTP telephony boards. You have to remember the roots of PCs. In the 70's (before my time), PC users were alot like HAM radio enthusiasts. These people build game systems today. They are becoming increasing frustrated with windows. They desperately need a third alternative. An operating environment that leverages the flexibility of Linux with a robust and probably superior GUI is an outright dream for these people. They would actually want to participate in building and porting drivers for powerful peripherals like telephony and audio/video signal processing hardware if it could lead to systems that outperform what is presently in the field. Once you convince these guys; the market share problem will fix itself.
Projections are only as good as the assumptions that they make. I think AAPL would be well served to reinvestigate the possiblity of opening up the OS X to support broader open hardware manufacturers. I stand by my perspective on this; and envision a computing world unshackled by the status quo.
jrentona
Beverly, MA
Spot on. Web 2.0 is a joke that only San Fran technocrats take seriously. This guy is right for the wrong reasons. And I don't get how someone can redefine "Dead". Suddently "dead" means non-threatening. This strikes of Clinton redefining "Is". This is obviously a deliberate mischoice of words to attract eyeballs.
o int_naught_answers/
For a little sanity check on web 2.0, I love this story out of the UK (though a little old):
http://www.theregister.co.uk/2005/11/11/web_two_p
Lets put things in perspective here. To find Microsoft's next competitor, you have to look at market caps a full 1/2 the size of M$. And Google (valued at 146 billion), is trading at 50 times earnings compared with microsoft's 26. More importantly, it isn't even in the same subsector. Google's stuff runs mostly on windows. They have no operatings system. Microsoft is still for all intents and purposes a monopoly. If you look at operting systems alone, Apple is less than 1/3 the size of Microsoft and Redhat was recently delisted.
On another note, in my opinion, Google is way way overvalued. The economy is slowing-down. Advertising is the first thing companies cut. You can't maintain 100% growth rate forever.
Microsoft is dead because they stop making meaningful innovations (just after the release of windows 2000 I surmise). And OS X is irrelevant looking at the numbers. If they stopped playing IBM's old closed architecture game, and leveraged the manufacturing power of the near east by allowing vendors like HP, Dell, Toshiba, and Sony build machines that run OS X; maybe we could have a duopoly in the Operating System market. Then either microsoft will start innovating for its very survival or apple will become the new industry leader. Either way, the consumer wins. Until then, its status quo for the forseeable future.
jrentona
Beverly, MA
I think if Apple can abandon their closed architecture model; they can really eat into the Microsoft monopoly.
Vista is a pig. I've seen it take 30-40 seconds to open up the Save-As dialog from every application. The Visual Studio 2005 IDE was a big flop in terms of Vista compatibility. There is no better time than now.
Could you imagine if Toshiba, HP, Dell, and others could offer laptops and workstations running OS X. We'd actually have a real duopoly in PC Operating Systems. The game would be afoot in competitive terms. It would be AMD and Intel all over again. Microsoft would actually be forced to start producing again. Just think of the innovations they would have to come up with to stay competitive. That's what the industry needs.
Of course, the software lag as far as cross-platform development has always been a major problem. But you cannot begin to tackle this one as a closed architecture.
.
It is always unsettling when 100% of the comments to a post agree. Reminds me of the dotcom bubble or enron or the subprime market in 2006.
I don't know; maybe I just like being a contrarian too much.
The truth is that M$ has pretty muched sucked since 2000. It took them a whopping 5 years for them to get XP working properly.
s -with-visual-studio-2005/
And Vista/Visual Studio 2005 is pretty much a train wreck for C developers. We used to be able to rely on the development environment. In fact, that area was always a significant innovation for these guys. No more. Fire Steve Embalmer before it is too late.
And the evidence just keeps rolling in:
http://www.microsoftweblog.com/2005/11/05/problem
I was going to respond to the inanely academic response about pointer reuse; but the post I am responding to hits the nail on the head. People invent hundreds and thousands of non-trivial mechanisms when writing large complex software systems every year. Some of these projects are classified. This patent is meaningless. You shouldn't be able to patent an algorithm. It is only slightly more nonsensical than patenting a line of code or a function. Stop waisting taxpayer money. Only systems should be patentable.
Yeah, its pretty obvious that next recession; that 60 times multiple is gauranteed to disintegrate 90's style. We are already at the tail end of one of the longest continuous bull markets in history. Global tensions are heating up. Oil prices are through the roof. 13% and growing of companies are missing expectations. It is only a matter of time.
When the economy goes sour; cyclical stuff like advertising is the first to go out the window. If you are smart and you own google now; I'd wait out this rally until the next time it pulls back. Then take your profits and buy a nice safe CD. You can can get 5-6% APR guaranteed thanks to Bernanke's rate raising.
Wait things out and buy back in when everything blows over in a few years.
James
Beverly, MA
ALPHA software is freshly implemented, in the process of being Unit and Integration tested in the lab by the developers and QA.
BETA simply means the software is pre-released to the general public. No test set can entirely simulate production in most cases; so every complex software system benefits from being exposed to courageous testers outside of the lab.
The problem is that the fast food management styles coming out of our business schools are incompatible with software QA. Testing gets a back seat to basically everything on the management agenda. Stuff gets designed without a test plans. A short half-hearted alpha testing phase is rushed over. It then goes immediatly to BETA and stays there for the life of the product. Hence your aversion to the term BETA, which I believe is quite justified.
James
Beverly, MA
As far as I am concerned, smart pointers have become a decisive breakthrough in software development. It basically removes the garbage collection advantage of VIMs. Intelligent application of smart pointers, including extensions of the basic std::auto_ptr to include smart arrays and reference counting gives C++ programmers all the benefits of garbage collection without dedicated processes wasting MIPS monitoring your code. All of this VM background bookkeeping is not free; and is above all not perfect. I have witnessed first hand cases of the Microsoft VM taking hours to cleanup wrapped resources for DLLs I've developed. Just because the language prevents you from worrying about freeing heap blocks and handles doesn't mean it gets done right. And worse, if there is a bug in the VM; one's only recourse is to guess at a possible work around and hope it triggers the GC to do its job. -James
The comment submission process filtered the shift operator << from the code. I selected "Code" format this time.
;(shift<<1)<x;shift=shift<<1 )
Revised below:
HYPER mxy( long x, long y )
{
long shift = 1;
char iShift = 0;
char bNeg = 0x0;
if( x==0 || y==0 )
return 0;
else if( x<0 )
{
++bNeg;
x = -x;
}
for(
++iShift;
if( bNeg )
return -( mxy( x-shift, y ) + ((HYPER)y<<iShift) );
else
return mxy( x-shift, y ) + ((HYPER)y<<iShift);
}
-James
This anti-Java poster is right on the money. This language tower of babel is destroying software.
;(shift1)x;shift=shift1 )
Here is the source (Compiled in MSVC6):
#ifdef WIN32
#define HYPER __int64
#else
#define HYPER long long
#endif
HYPER mxy( long x, long y )
{
long shift = 1;
char iShift = 0;
char bNeg = 0x0;
if( x==0 || y==0 )
return 0;
else if( x0 )
{
++bNeg;
x = -x;
}
for(
++iShift;
if( bNeg )
return -( mxy( x-shift, y ) + ((HYPER)yiShift) );
else
return mxy( x-shift, y ) + ((HYPER)yiShift);
}
This problem is actually pretty cool (at least I think so). As far as relevance, I have seen at least one DSP processor without an integer divide. Not certain about multiply.
My point is that java programmers look down on a problem like this as beneath them somehow. While C and assembly programmers love the challenge (or at least I do). This foolish quote about Comp Sci not being about computers is self serving. It was obviously written by some over-hyped professor whose job hinges on writing books and lectures rather than effective software.
Fact is I'd rather travel to Mars on a spacecraft whose software was written by C and assembly programmers than a bunch of academic java guys too smart to understand the underlying hardware.
James A. Renton
Beverly, MA.