You sir, are a fucking idiot who spends too much time watching fox news. Both New York and California have Republican governors. Both of them are predominately "true" republican states. None of this hogwash jesus freak crap that dominates the south and midwest. The biggest tech centers in red states are RTP NC and Austin TX. Both of which are dwarfed by NYC and silicon valley. I would also point out that those areas are driven by companies that are based in NYC and/or the West coast.
I own guns. I like guns, but if you think a place with 12 million people living in close proximity is going to have the same type of laws as places with only 5 million people in an entire state.... well then your fucking cracked. That doesnt even touch on the fact that the vast majority of the countries (70%) population lives on/near the coast. Or that most young people dont want to live in a dead end town with one employer, or no nightlife.
Orlando has one of the most bland culture scenes I have ever witnessed. Part of the draw that the bay area has is the diversity and overall "scene" not just the techies.
And you would have a hell of a time affording it. (reverse commuting) Most tech people want to work with other tech people, otherwise they tend to be treated as a step above a secretary.
If you believe that you have obviously never worked in a large environment. Places that have droves of people in a wide open area will have one slacker bring down half the department to show them a flash cartoon/youtube video.
"Religious beliefs:
After his parents' divorce, McVeigh and his siblings lived with their father, a devout Roman Catholic who often attended Daily Mass. In a recorded interview with Time Magazine[2] he professed his belief in "a God"
All information stating that he was agnostic is speculation. His belief in god however is on the record. If you really want to dig farther then some of the additional article's on him and his background also show a tendency towards christianity/catholicism.
Bullshit. They offered two different options. Niether of them was even remotely close to a smooth end user oriented version of Linux. Walmart sells what 2% of the total PC's out there ? Non-brand name crap.
Microsoft is a convicted monopoly for a reason. In the 8 years I have worked in the computer industry at varying levels I still have not met an non-technical end user who has upgraded windows. Period. Everyone seems to think that microsoft is being chosen by the end users. Its not, the sad fact of the matter is that most people dont understand or care. They just buy a "computer" and expect it to work. It could ship with friggin BeOS or Irix for all they give a damn, so long as the crap they plug in works 51% of the time, they are happy.
How many people drive Kia ? Cause thats the market for a walmart PC. Windows is dominant because they are already installed.
Advertising ? If manufacturers didnt install vista on systems by default going forward I bet it would never top 15% adoption. Its impossible to topple a monopoly who has a choke hold on the distribution channels.
One million dollars in the hands of a motivated person, or a person who knows how to make money is enough money to be financially independant for a long time if they are smart.
no. Its for "System Binaries" or "System Administration Binaries".
If you doubt me, go into/sbin and do an ldd on something (ldd $foo) and look at the libraries printed, they are very much shared.
I would buy into this whole "the user must know everythings location and purpose" crap.... if they already knew it about... whatever they are running now. OSX, Windows or $other. Users dont know squat. The 10% of people who know something still dont know much.
No, you sir just missed half of his comment. A *home* user is going to use the gui apps to administer such things. Tell me that the average windows user has any fracking clue what the registry is, or what that little section of crap in the lower right hand corner is or how to edit it. Go ahead and try. Then go talk to non-technical people shopping at the local radio shack, best buy, target, wal mart, circuit city or whatever and ask them to explain it to you. Your IQ will drop a few points.
Users dont know anything, as long as the software "works" then they dont care.
Users dont pay for anything in their eyes, windows is included on that computer they just bought. Most people dont care enough to bother switching to another os, free or not. Switch it so linux is installed by default and they would have to buy windows and install it and you'll see the exact same thing holds true 10% do something "different" and 90% just roll with it.
Depends on what your doing. In a desktop environment... I wouldnt. On a server I would use solaris in many places, except the lightweight "disposable" boxes like web servers. The real boxes like the DB's and application boxes get the OS that is better suited for dealing with high load, high availability. Which is solaris.
Linux hasnt been actively developed for the server since early 2.4
I can certainly get you some links if you wish. Before I go to that length bear in mind that even HUGE advocates of pair programming admit that its an "idealist" solution that in many circumstances has a less than ideal situation/environment. Bear in mind the following problems:
1. Location. In todays workforce we are becoming more and more spread out. I have worked with hundreds (no exageration) of people all over the world, most of whom I have never met. Most companies wont pay relocation. That leaves a shallow talent/personality pool. XP doesnt work well over distances, in those cases its essentially no different than most (well run)corporate environments.
2. Time. It assumes that whatever pair you currently have working together is going to work the same hours, take the same breaks and so on. VERY rare. Most people who get into a decent job do not expect to be doing things like factory workers. Why would a good coder subject themselves to someone elses life requirements, when they can get a job elsewhere ?
3. Personalities. Most programmers are not outgoing. Its a fact. That makes them very difficult to work with in this type of situation. A good percent of those who are outgoing are control freaks who "must do it their way". I have personally witnessed the hell that occurs when management puts "teams" together to do this sort of thing. More often than not a bussiness would have to "re-hire" everyone to make sure they are suited for this type of work. Given that a good chunk of them will not be, that leads to turnover which is a cost. Bussiness's hate costs.
4. Infrastructure. Most companies do not have adequate desk space/office space/conference rooms to do something like this. Not to mention, some people are just not comfortable working that close (physically) with someone else.
I think from an idealist or an educational standpoint its a great concept, and can probably be implemented with amazing results. However in my experience with the huge corporate environment I can say this wont work most of the time. For many reasons, not the least of which are highlighted above.
A better way of putting this would be what components/sections are writen by individuals. Sure on large projects people collaborate, but thats out of need, not nature.
Maybe to managers who dont understand the typical coder they should be an endagered species. To me I look around at corporate software and I am wholly unimpressed. I look at the things that the "overnighters" have accomplished vs the things the guys wearing ties and sitting in cubicles have accomplished and I pick the "overnighters" without any hesitation, flow charts, fancy buzz words or any fluff.
You'll notice that firms who force crap like XP (but not limited to) onto coders fail in the long term. (or at least from a technical perspective they fail) I can also guarantee that sitting two people down to code on one terminal is a quick way to cut production and ramp up cost in a big way. Even the best minds dont always work well together.
Software engineering is not the same as coding or testing and people who try to force everything into one role are in for some real shockers. You dont see engineers building the bridge, and you likely never will. The most efficient design is an architect/software engineer who designs the program/application/whatever with some input from the coders and/or testers, the coders then start to build the thing in question with oversight from the architect. Once your at a certain point you then include the testers. In the current corporate culture (which largely fails to understand the technical aspect of things) they try to replace the architect position and force the technical aspect down a level. Which is why you end up with so many project managers who use things like XP.
The differences are Google is not a monopoly and for the average person, its very hard to escape using a microsoft desktop. So its not the same. Give Mac OSX or Linux ~30% market share and then come see me.
YES !! My point exactly, the market can bear more than just one design, and they can succeed. The way some people make it sound its as if everything should be generic Intel/AMD chips doing everything.
I would say that niagra is going after fairly "common" markets for x86 but thats just my opinion.
Having not been here very long I couldnt say for sure when the change took place, but from what I have heard it was back when Gershner was in charge. The environment is very relaxed. Perhaps not all of IBM is this way, but I work at a fairly large site and would guess that its fairly common. IBM has its own silly "business" things that make no sense. The dress code (at least for technical people) is not one of them.
Well start with saying that yes I may have misunderstood you, that is entirely possible.
Alpha is indeed the best to work on, everyone who has worked with one has made that statement and it was a good performer as well. Why intel didnt take some of those parts of alpha is beyond me.
To be more correct some manager made the statement that you dont want to be doing much assembly on Cell. At some point if the things is going to run things fast there is going to need to be low level coding done. Not much on the assembler level, but some none-the-less. Yes it is a complicated design to deal with on that level, and because of its complexity it makes most people run away. I would say its better than x86 IMHO, but then again I dont do much with assembly.
The market meaning the companies, not the consumers. If x86 is so grand why is niagra doing so well ? Because for the first time in YEARS the larger companies are trying something different. That different thing is specialized processors. I think it will hurt intel far more than opteron did. Perhaps the market doesnt consider the details, but it does consider what works best for them. I know a good deal of web design and hosting (and a few content delivery) places that are switching to niagra (the t1000 or t2000) because it slaughters all others at web and (from what I have read) DB.
If I sound a bit annoyed its because I am working on this project at the moment. Granted the Cell is only on the periphery at the moment (the entire thing is supposed to take a few years, as stated in the article) but I know a good bit about what it does well, and what it doesnt in its current state.
I believe you do not know what you are talking about. The reason this cluster has so many opterons is because they are double precision and has such are better suited for doing the task at hand, the cell's are used for specialized tasks only in this implementation. The Cell architecture is not going to be limited to just one type of application, it is designed with everything from supercomputing to embeded stuff in mind, its just going to take a while to get to the point where there are chips of varrying types available for all of those markets.
This is not supposed to be an x86 replacement. Ask a real developer what its like to work on x86 assembler as opposed to sparc, power, and so on. With the desire for "general" ability comes problems. It took the market a while to figure that out, but now you are seeing the correction taking place with cell, niagra, power6 and so on.
Saying the Cell was designed just for games (which it seems your implying) is like saying Power was designed just to run Apple desktops. Sure ppc is derivative of it, and according to some (ignorant) people that was the largest success of Power, however thats not what it was designed for.
Sure this itteration of the chip was largely influenced/steered towards graphics, that doesnt mean the entire architecture was or is claiming to be a "solve all" right now.
Lastly it is important to note that developers complain about any "new" thing that they have a hard time adapting to. Cell is a bitch to code for, but it is beautiful when done correctly.
Slashdot Mirror
That would make sense ... if google and its employees didnt already own most of google. (why do you think they wont split the stock.)
You sir, are a fucking idiot who spends too much time watching fox news. Both New York and California have Republican governors. Both of them are predominately "true" republican states. None of this hogwash jesus freak crap that dominates the south and midwest. The biggest tech centers in red states are RTP NC and Austin TX. Both of which are dwarfed by NYC and silicon valley. I would also point out that those areas are driven by companies that are based in NYC and/or the West coast.
.... well then your fucking cracked. That doesnt even touch on the fact that the vast majority of the countries (70%) population lives on/near the coast. Or that most young people dont want to live in a dead end town with one employer, or no nightlife.
I own guns. I like guns, but if you think a place with 12 million people living in close proximity is going to have the same type of laws as places with only 5 million people in an entire state
Orlando has one of the most bland culture scenes I have ever witnessed. Part of the draw that the bay area has is the diversity and overall "scene" not just the techies.
And you would have a hell of a time affording it. (reverse commuting) Most tech people want to work with other tech people, otherwise they tend to be treated as a step above a secretary.
If you believe that you have obviously never worked in a large environment. Places that have droves of people in a wide open area will have one slacker bring down half the department to show them a flash cartoon/youtube video.
Actually it is quite the opposite.
"Religious beliefs:
After his parents' divorce, McVeigh and his siblings lived with their father, a devout Roman Catholic who often attended Daily Mass. In a recorded interview with Time Magazine[2] he professed his belief in "a God"
All information stating that he was agnostic is speculation. His belief in god however is on the record. If you really want to dig farther then some of the additional article's on him and his background also show a tendency towards christianity/catholicism.
Tell it to the people in oklahoma city. Christians did that. American Christians to boot.
Bullshit. They offered two different options. Niether of them was even remotely close to a smooth end user oriented version of Linux. Walmart sells what 2% of the total PC's out there ? Non-brand name crap.
Microsoft is a convicted monopoly for a reason. In the 8 years I have worked in the computer industry at varying levels I still have not met an non-technical end user who has upgraded windows. Period. Everyone seems to think that microsoft is being chosen by the end users. Its not, the sad fact of the matter is that most people dont understand or care. They just buy a "computer" and expect it to work. It could ship with friggin BeOS or Irix for all they give a damn, so long as the crap they plug in works 51% of the time, they are happy.
How many people drive Kia ? Cause thats the market for a walmart PC. Windows is dominant because they are already installed.
Advertising ? If manufacturers didnt install vista on systems by default going forward I bet it would never top 15% adoption. Its impossible to topple a monopoly who has a choke hold on the distribution channels.
One million dollars in the hands of a motivated person, or a person who knows how to make money is enough money to be financially independant for a long time if they are smart.
no. Its for "System Binaries" or "System Administration Binaries". /sbin and do an ldd on something (ldd $foo) and look at the libraries printed, they are very much shared.
.... if they already knew it about ... whatever they are running now. OSX, Windows or $other. Users dont know squat. The 10% of people who know something still dont know much.
If you doubt me, go into
I would buy into this whole "the user must know everythings location and purpose" crap
No, you sir just missed half of his comment. A *home* user is going to use the gui apps to administer such things. Tell me that the average windows user has any fracking clue what the registry is, or what that little section of crap in the lower right hand corner is or how to edit it. Go ahead and try. Then go talk to non-technical people shopping at the local radio shack, best buy, target, wal mart, circuit city or whatever and ask them to explain it to you. Your IQ will drop a few points.
Users dont know anything, as long as the software "works" then they dont care.
Users dont pay for anything in their eyes, windows is included on that computer they just bought. Most people dont care enough to bother switching to another os, free or not. Switch it so linux is installed by default and they would have to buy windows and install it and you'll see the exact same thing holds true 10% do something "different" and 90% just roll with it.
Depends on what your doing. In a desktop environment ... I wouldnt. On a server I would use solaris in many places, except the lightweight "disposable" boxes like web servers. The real boxes like the DB's and application boxes get the OS that is better suited for dealing with high load, high availability. Which is solaris.
Linux hasnt been actively developed for the server since early 2.4
I can certainly get you some links if you wish. Before I go to that length bear in mind that even HUGE advocates of pair programming admit that its an "idealist" solution that in many circumstances has a less than ideal situation/environment. Bear in mind the following problems:
1. Location. In todays workforce we are becoming more and more spread out. I have worked with hundreds (no exageration) of people all over the world, most of whom I have never met. Most companies wont pay relocation. That leaves a shallow talent/personality pool. XP doesnt work well over distances, in those cases its essentially no different than most (well run)corporate environments.
2. Time. It assumes that whatever pair you currently have working together is going to work the same hours, take the same breaks and so on. VERY rare. Most people who get into a decent job do not expect to be doing things like factory workers. Why would a good coder subject themselves to someone elses life requirements, when they can get a job elsewhere ?
3. Personalities. Most programmers are not outgoing. Its a fact. That makes them very difficult to work with in this type of situation. A good percent of those who are outgoing are control freaks who "must do it their way". I have personally witnessed the hell that occurs when management puts "teams" together to do this sort of thing. More often than not a bussiness would have to "re-hire" everyone to make sure they are suited for this type of work. Given that a good chunk of them will not be, that leads to turnover which is a cost. Bussiness's hate costs.
4. Infrastructure. Most companies do not have adequate desk space/office space/conference rooms to do something like this. Not to mention, some people are just not comfortable working that close (physically) with someone else.
I think from an idealist or an educational standpoint its a great concept, and can probably be implemented with amazing results. However in my experience with the huge corporate environment I can say this wont work most of the time. For many reasons, not the least of which are highlighted above.
A better way of putting this would be what components/sections are writen by individuals. Sure on large projects people collaborate, but thats out of need, not nature.
Maybe to managers who dont understand the typical coder they should be an endagered species. To me I look around at corporate software and I am wholly unimpressed. I look at the things that the "overnighters" have accomplished vs the things the guys wearing ties and sitting in cubicles have accomplished and I pick the "overnighters" without any hesitation, flow charts, fancy buzz words or any fluff.
You'll notice that firms who force crap like XP (but not limited to) onto coders fail in the long term. (or at least from a technical perspective they fail) I can also guarantee that sitting two people down to code on one terminal is a quick way to cut production and ramp up cost in a big way. Even the best minds dont always work well together.
Software engineering is not the same as coding or testing and people who try to force everything into one role are in for some real shockers. You dont see engineers building the bridge, and you likely never will. The most efficient design is an architect/software engineer who designs the program/application/whatever with some input from the coders and/or testers, the coders then start to build the thing in question with oversight from the architect. Once your at a certain point you then include the testers. In the current corporate culture (which largely fails to understand the technical aspect of things) they try to replace the architect position and force the technical aspect down a level. Which is why you end up with so many project managers who use things like XP.
Unless google signed them to an exclusive agreement, or required them to not do bussiness with certain other parties.
Welcome to bussiness. Google can either fight their opponents on level ground or loose and go out of bussiness.
The differences are Google is not a monopoly and for the average person, its very hard to escape using a microsoft desktop. So its not the same. Give Mac OSX or Linux ~30% market share and then come see me.
Yeah, but that doesnt mean easy is right. Or good.
YES !! My point exactly, the market can bear more than just one design, and they can succeed. The way some people make it sound its as if everything should be generic Intel/AMD chips doing everything.
I would say that niagra is going after fairly "common" markets for x86 but thats just my opinion.
Having not been here very long I couldnt say for sure when the change took place, but from what I have heard it was back when Gershner was in charge. The environment is very relaxed. Perhaps not all of IBM is this way, but I work at a fairly large site and would guess that its fairly common. IBM has its own silly "business" things that make no sense. The dress code (at least for technical people) is not one of them.
Surprisingly working at IBM, I wear shorts and sandals to work.
Working at sun I could get away with everything but the sandals
Working at a small contractor I had to wear a tie.
Working at a small start up I had to wear a collared shirt.
Its been backwards for me.
Well start with saying that yes I may have misunderstood you, that is entirely possible.
Alpha is indeed the best to work on, everyone who has worked with one has made that statement and it was a good performer as well. Why intel didnt take some of those parts of alpha is beyond me.
To be more correct some manager made the statement that you dont want to be doing much assembly on Cell. At some point if the things is going to run things fast there is going to need to be low level coding done. Not much on the assembler level, but some none-the-less. Yes it is a complicated design to deal with on that level, and because of its complexity it makes most people run away. I would say its better than x86 IMHO, but then again I dont do much with assembly.
The market meaning the companies, not the consumers. If x86 is so grand why is niagra doing so well ? Because for the first time in YEARS the larger companies are trying something different. That different thing is specialized processors. I think it will hurt intel far more than opteron did. Perhaps the market doesnt consider the details, but it does consider what works best for them. I know a good deal of web design and hosting (and a few content delivery) places that are switching to niagra (the t1000 or t2000) because it slaughters all others at web and (from what I have read) DB.
If I sound a bit annoyed its because I am working on this project at the moment. Granted the Cell is only on the periphery at the moment (the entire thing is supposed to take a few years, as stated in the article) but I know a good bit about what it does well, and what it doesnt in its current state.
I believe you do not know what you are talking about. The reason this cluster has so many opterons is because they are double precision and has such are better suited for doing the task at hand, the cell's are used for specialized tasks only in this implementation. The Cell architecture is not going to be limited to just one type of application, it is designed with everything from supercomputing to embeded stuff in mind, its just going to take a while to get to the point where there are chips of varrying types available for all of those markets.
This is not supposed to be an x86 replacement. Ask a real developer what its like to work on x86 assembler as opposed to sparc, power, and so on. With the desire for "general" ability comes problems. It took the market a while to figure that out, but now you are seeing the correction taking place with cell, niagra, power6 and so on.
Saying the Cell was designed just for games (which it seems your implying) is like saying Power was designed just to run Apple desktops. Sure ppc is derivative of it, and according to some (ignorant) people that was the largest success of Power, however thats not what it was designed for.
Sure this itteration of the chip was largely influenced/steered towards graphics, that doesnt mean the entire architecture was or is claiming to be a "solve all" right now.
Lastly it is important to note that developers complain about any "new" thing that they have a hard time adapting to. Cell is a bitch to code for, but it is beautiful when done correctly.