I didn't either until people with 600,000+ started saying like "such a low number". I don't have my original sign-up email anymore, but it clocks in somewhere around the summer of 1997.
I read the book years ago. I read it before I ever read Slashdot - and 33728 is a fairly low Slashdot ID. It is an amazing book. Regardless of what you think about his politics, lifestyle or whatever, you should read the book.
Hard drives with voicecoil head positioning (some 99.99% of them sold in the past 10 years) will autopark on poweroff. The real risk is how much latency the os has between when an application saved it's data and when the buffers are flushed.
Some applications keep files open for a long time: MySQL, gDBM-based apps, Squid. Most of those application implement their own mini-filesystems within a file optimised for task. These systems are supposed to preserve their integrity by journaling their modifications in case the underlying os doesn't.
Switching off a computer because it has a journaling filesystem is like stopping a car by driving into a something because it has seat belts.
There is a condition in psychology called "Displacement of Goals." Management, of course, requires structures and methods to manage a project. That's a given for me. The "BUT" for me, and it's a big one for me because I once worked somewhere that greviously failed to recognize it, management can't allow its staff to substitute the goal of clinging to "The Plan" to displace the goal of solving the customer's problem.
I've seen applications transform how its users see their jobs. I'm not talking about changes staffing levels, but how the system shoulders so much work that their brains can now focus on aspects that they were otherwise too busy to notice - the proverbial trees in the forest. They now understand their job in a new way and the goal shifts.
Sure, some programers have user communities that we would all kill for: The Shuttle Program. You've got areospace engineers making a spec for software engineers. It's a match made in heaven. Most of us aren't so lucky. Most users have tacit requirement that never get written down will never be understood by the developer. I realised one afternoon that a vendor, a business analyst and a business customer spent three hours working out requirement using the word "Gate" (like at an airport) and that weren't talking about the same thing.
Planning is important. But, the plan that comes out of it, must be sacrificed one line item at a time as the customer realises that will get what they asked for and then realises that the asked for the wrong stuff. A project must allow for the organic nature of people. The project also has to plan the cheese moving politic of change.
That makes lots of sense. Software costs do multiply in clustering. Zero times 100 is still zero. But, clustering has other headaches beyond money.
The usual clustering I've seen is "Hot Spare" clustering. The primary runs until it goes kaput, then the second takes over. For database clustering, the two boxes usually share the same disks. I think I've seen more outages from false takeovers by the seconday than real failures of the primary.
The other problem with clustering is that all of your software applications have to be cluster tolerant. If the user app keeps a database connection open and a rollover occurs, the connection state doesn't and can't rollover with it. To a client system, a cluster failover looks like a server reboot. Don't underestimate the difficulty of this problem. A new application has to be designed with that in mind. Retro-fitting it in later is hard - and costly, even with free platforms.
Another issue that can't be solved with clustering is application failure or application limits. You may recall the airline system failure last Christmas? Some 80% of Slashdot readers asked where was the backup? (there was) should have used Unix (they were).
The box (RS6000) and operating system (AIX) kept running just fine. A hundred computer cluster couldn't solve the the real problem: the application couldn't handle the volume of information it was required to hold and they at the mercy of a proprietary source code vendor.
What I think he is really trying to fight is what the Agile Movement calls Big Design Up Front. BDUF recognizes that customers will never see how a software application will transform their business and will invariably change their minds. Dwight D. Eisenhower summed it up as "In preparing for battle I have always found that plans are useless, but planning is indispensable."
No. Java is a perfect example of API based threading and in Java it's easy to do. Still a class has to implement Runnable and the programmer has to create a Thread and start it.
The synchronize keyword was closer to where I was going. Suppose Java had a thread modifier keyword for looping operators. You could then:
public void renderImage(Image images[]) { thread for (int i = 0; i < images.length; i++) { render(images[i]); } }
each iteration of the looping block launches as a different task running in parallel and the loop exits once all tasks are complete.
Or a easier RPC like
I certianly recognize that there are some very hard problems that would need to be solved in such a senario. Thread synchronization, mutexes, semaphores would need to be looked at. A clean way to integrate directory services and other ways of defining environmental resources is not trivial either and critical for the success of such a language.
Most compilers/interpreters support languages designed for single thread execution. Fortran, COBOL, C, C++, Ruby, Perl, PHP, Java,... Sure all these have API calls to make use of multiple threads, but the language itself isn't multi-threaded.
In my shameless search for a site to cite, I found this http://www-unix.mcs.anl.gov/dbpp/
which covers lots of problems that have to be solved.
I'd love to see a language (or language extension) cleanly define a way to let me define a code block attributes which could affect how and where it gets executed. The runtime library could then distribute that block as the environment best allows.
Just because something works doesn't mean it's not obsolescent. I don't care what it's doing, a fifteen-year-old machine is obsolete NOW.
A agree with you, or at least a nuance of your argument.
A system is more than the bolts and bytes that goes into it, it's the service it provides to those that use it.
The real obsolesence is the deterioration of the knowledge of what service that machine provides, how that machine does it and who is qualified to admin that machine.
IT management show regularly review all those factors and determine if any of those categories are lacking in the systems^H^H^H^H services they provide.
In short, if I have a 20 year old system that still runs and I can still get parts for and I still have staff who can diagnose and correct problems. It's not obsolete.
Important Slashdot safety tip, boy and girls, make sure your morning coffee has kicked in before typing! Either that or my Freudian Slip is showing.
Quite funny none the less.
To the spelling nazi's, I checked: "dyke" and "dike" are different spellings for the same word. Much like "tire" and "tyre". In the US, however, the "dike" spelling is almost never associated with lesbians.
Somewhat tangental...
Fourteen years ago, I took a tour of one of the pumping stations in New Orleans. The operator took great pride in pointing out the the Dutch asked them about pumping technology. I don't know for sure when, but I think it was recently. The Dutch, of course, have been pumping for centuries, but there was a catastrophic failure during the 1950's which lead to a major redesign and fortification of the who dyke system.
One of the theatres near me has started a two-tier theatre system. The lower tier is a typical theatre. The upper tier has assigned seating and leather seats.
While I've always been too cheap to try the upper tier, I would imagine the piceing system would discourage the casual viewer in favor of someone who has a little at stake in enjoying the film.
While waiting to get my hair cut a few days ago, I read in Time (or Newsweek or one of them) that more and more movie houses are fighting for customers. More are trying upscale theatres because they can no long depend on the films from Hollywood to automatically draw large crowds.
Do you honestly believe that your government can do it cheaper as a monopoly than competing private companies?
There are countless example where this is true. But that's usually because the private company seeks to cut costs to increase profit. But, a private company isn't interested unless there is profit. There are countless examples where governtment programs are already so lean that a private company can't complete.
Actually, we're not that far apart. I intended to say the same thing. A bar code would be an excellent example of a "dormant" technology. If I'm hauled into a hospital, a bar code is still useful.
One of my big concerns is foreign travel. I've had the good fortune to travel to ten countries and Americans stick out enough as it is. Given the recent, abrasive US policies, I fear that there are those with increased motivation to find and harm US citizens.
Something like RFID doesn't have an "off switch."
Whether implanted, in my passport, or anywhere the problem is you can't prevent someone from strolling along Le Champs-Ellesse waiting for his "American Detector" to go off.
The big thing that scares me about any kind of implant is that it requires non-physical contact to activate. Calling it an implant doesn't change that it's RFID all dressed up. It's pretty obvious has to have some means of being triggered and read from a distance. Even if it's only two feet, that's close enough for some one to detect it in a crowded area.
Assume for a moment that something like this is inevitable - which currently isn't too hard, I'd want something that is completely dormant until I chose who gets to read it.
It seems a lot of really bright people have come up with "The Reason" Apple switched. It's as if there is only one reason.
I'm sure I've missed one, but we've heard, A) Faster CPUs, B) Less power consumption, C) Cheaper CPUs (preferred vendor prices), D) Cheaper iPod CPUs (XScale is an Intel), E) Trusted Computing Platform.
I think Apple weighed several factors before switching. In short, F) All of the Above.
Except that it won't suck. Of course, it wouldn't be great, but it won't suck. In short, they will make it good enough that most people won't care.
This is plugging a "hole." If I have video card with digital outputs, what's to prevent someone from developing a recording device that otherwise mimics a digital monitor and records hi-res content?
While I don't agree Jaschan's crime was "minor," Tierney's argument that the value of Jaschan's life is reduced to a "cost-benefit analysis" is completely depraived of any morality I can accept.
Jaschan's crime exacted financial damage on the world. The convicted Enron executives exacted financial damage on the world.
Computers are tools. Even though many find them mysterious, computers are not mystical or enchanted or creations of witchcraft. They're tools.
A crime is a crime. Using a computer to commit a crime shouldn't significantly alter the punishment of the crime any more than, say, using an automobile, telephone, or sledgehammer to commit that crime. Virtual crimes shoud carry a penalty similar to the real thing.
The Islamists have the same goals as Hitler -- world domination with them in charge.
I couldn't disagree more. I'd even suggest quite the opposite. The rallying cry of terrorist leaders, the method they use to recruit, is assert that the Western world, the US in particular, is out for world domination. For evidence, they point to the US military bases in their backyards and torture in Guantanamo.
Another post quotes from Battlestart Galatica, "The opposite of war isn't always peace, sometimes is't slavery." Freedom and Slavery are on opposite ends of the spectrum. War is in the middle. The Islamist who feels oppressed or fears enslavement by Western countries knows the pathway to freedom, his freedom, is war against the oppressor.
A key front in the "war on terror" needs to include stopping the supply of new terrorists.
It's one thing (an important one at that) to go after the known terrorists; but, terrorists aren't born terrorists. Ordinary people are made into terrorists by some radicalizing event. The life of Ayman al Zawahiri is a text-book example of how an otherwise decent fellow is radicalized into a monster.
I don't have the easy answer to how to do that. Frankly, I wouldn't trust anyone who said they did. But I don't see anyone in authority even trying to work on the problem.
Slashdot Mirror
I didn't either until people with 600,000+ started saying like "such a low number". I don't have my original sign-up email anymore, but it clocks in somewhere around the summer of 1997.
I read the book years ago. I read it before I ever read Slashdot - and 33728 is a fairly low Slashdot ID. It is an amazing book. Regardless of what you think about his politics, lifestyle or whatever, you should read the book.
Some applications keep files open for a long time: MySQL, gDBM-based apps, Squid. Most of those application implement their own mini-filesystems within a file optimised for task. These systems are supposed to preserve their integrity by journaling their modifications in case the underlying os doesn't.
Switching off a computer because it has a journaling filesystem is like stopping a car by driving into a something because it has seat belts.
There is a condition in psychology called "Displacement of Goals." Management, of course, requires structures and methods to manage a project. That's a given for me. The "BUT" for me, and it's a big one for me because I once worked somewhere that greviously failed to recognize it, management can't allow its staff to substitute the goal of clinging to "The Plan" to displace the goal of solving the customer's problem.
I've seen applications transform how its users see their jobs. I'm not talking about changes staffing levels, but how the system shoulders so much work that their brains can now focus on aspects that they were otherwise too busy to notice - the proverbial trees in the forest. They now understand their job in a new way and the goal shifts.
Sure, some programers have user communities that we would all kill for: The Shuttle Program. You've got areospace engineers making a spec for software engineers. It's a match made in heaven. Most of us aren't so lucky. Most users have tacit requirement that never get written down will never be understood by the developer. I realised one afternoon that a vendor, a business analyst and a business customer spent three hours working out requirement using the word "Gate" (like at an airport) and that weren't talking about the same thing.
Planning is important. But, the plan that comes out of it, must be sacrificed one line item at a time as the customer realises that will get what they asked for and then realises that the asked for the wrong stuff. A project must allow for the organic nature of people. The project also has to plan the cheese moving politic of change.
The usual clustering I've seen is "Hot Spare" clustering. The primary runs until it goes kaput, then the second takes over. For database clustering, the two boxes usually share the same disks. I think I've seen more outages from false takeovers by the seconday than real failures of the primary.
The other problem with clustering is that all of your software applications have to be cluster tolerant. If the user app keeps a database connection open and a rollover occurs, the connection state doesn't and can't rollover with it. To a client system, a cluster failover looks like a server reboot. Don't underestimate the difficulty of this problem. A new application has to be designed with that in mind. Retro-fitting it in later is hard - and costly, even with free platforms.
Another issue that can't be solved with clustering is application failure or application limits. You may recall the airline system failure last Christmas? Some 80% of Slashdot readers asked where was the backup? (there was) should have used Unix (they were). The box (RS6000) and operating system (AIX) kept running just fine. A hundred computer cluster couldn't solve the the real problem: the application couldn't handle the volume of information it was required to hold and they at the mercy of a proprietary source code vendor.
What I think he is really trying to fight is what the Agile Movement calls Big Design Up Front. BDUF recognizes that customers will never see how a software application will transform their business and will invariably change their minds. Dwight D. Eisenhower summed it up as "In preparing for battle I have always found that plans are useless, but planning is indispensable."
Wouldn't that violate the Endangered Species Act?
The synchronize keyword was closer to where I was going. Suppose Java had a thread modifier keyword for looping operators. You could then:
each iteration of the looping block launches as a different task running in parallel and the loop exits once all tasks are complete. Or a easier RPC likeI certianly recognize that there are some very hard problems that would need to be solved in such a senario. Thread synchronization, mutexes, semaphores would need to be looked at. A clean way to integrate directory services and other ways of defining environmental resources is not trivial either and critical for the success of such a language.In my shameless search for a site to cite, I found this http://www-unix.mcs.anl.gov/dbpp/ which covers lots of problems that have to be solved.
I'd love to see a language (or language extension) cleanly define a way to let me define a code block attributes which could affect how and where it gets executed. The runtime library could then distribute that block as the environment best allows.
It's still wierd seeing Race Street and Sawyer's Point in a tech article.
Light. Lots of it. And one an articulated arm too.
Let me know the next time you need to stress test anything.
A system is more than the bolts and bytes that goes into it, it's the service it provides to those that use it. The real obsolesence is the deterioration of the knowledge of what service that machine provides, how that machine does it and who is qualified to admin that machine. IT management show regularly review all those factors and determine if any of those categories are lacking in the systems^H^H^H^H services they provide.
In short, if I have a 20 year old system that still runs and I can still get parts for and I still have staff who can diagnose and correct problems. It's not obsolete.
Important Slashdot safety tip, boy and girls, make sure your morning coffee has kicked in before typing! Either that or my Freudian Slip is showing.
Quite funny none the less.
To the spelling nazi's, I checked: "dyke" and "dike" are different spellings for the same word. Much like "tire" and "tyre". In the US, however, the "dike" spelling is almost never associated with lesbians.
Somewhat tangental ...
Fourteen years ago, I took a tour of one of the pumping stations in New Orleans. The operator took great pride in pointing out the the Dutch asked them about pumping technology. I don't know for sure when, but I think it was recently. The Dutch, of course, have been pumping for centuries, but there was a catastrophic failure during the 1950's which lead to a major redesign and fortification of the who dyke system.
While I've always been too cheap to try the upper tier, I would imagine the piceing system would discourage the casual viewer in favor of someone who has a little at stake in enjoying the film.
While waiting to get my hair cut a few days ago, I read in Time (or Newsweek or one of them) that more and more movie houses are fighting for customers. More are trying upscale theatres because they can no long depend on the films from Hollywood to automatically draw large crowds.
The RIAA most likely has Power-of-Attorney to act on behalf of each copyright owner for that purpose.
There are countless example where this is true. But that's usually because the private company seeks to cut costs to increase profit. But, a private company isn't interested unless there is profit. There are countless examples where governtment programs are already so lean that a private company can't complete.
One of my big concerns is foreign travel. I've had the good fortune to travel to ten countries and Americans stick out enough as it is. Given the recent, abrasive US policies, I fear that there are those with increased motivation to find and harm US citizens.
Something like RFID doesn't have an "off switch." Whether implanted, in my passport, or anywhere the problem is you can't prevent someone from strolling along Le Champs-Ellesse waiting for his "American Detector" to go off.
The big thing that scares me about any kind of implant is that it requires non-physical contact to activate. Calling it an implant doesn't change that it's RFID all dressed up. It's pretty obvious has to have some means of being triggered and read from a distance. Even if it's only two feet, that's close enough for some one to detect it in a crowded area.
Assume for a moment that something like this is inevitable - which currently isn't too hard, I'd want something that is completely dormant until I chose who gets to read it.
I'm sure I've missed one, but we've heard, A) Faster CPUs, B) Less power consumption, C) Cheaper CPUs (preferred vendor prices), D) Cheaper iPod CPUs (XScale is an Intel), E) Trusted Computing Platform.
I think Apple weighed several factors before switching. In short, F) All of the Above.
This is plugging a "hole." If I have video card with digital outputs, what's to prevent someone from developing a recording device that otherwise mimics a digital monitor and records hi-res content?
Especially for someone with a four digit Slashdot ID.
Jaschan's crime exacted financial damage on the world. The convicted Enron executives exacted financial damage on the world.
Computers are tools. Even though many find them mysterious, computers are not mystical or enchanted or creations of witchcraft. They're tools.
A crime is a crime. Using a computer to commit a crime shouldn't significantly alter the punishment of the crime any more than, say, using an automobile, telephone, or sledgehammer to commit that crime. Virtual crimes shoud carry a penalty similar to the real thing.
Another post quotes from Battlestart Galatica, "The opposite of war isn't always peace, sometimes is't slavery." Freedom and Slavery are on opposite ends of the spectrum. War is in the middle. The Islamist who feels oppressed or fears enslavement by Western countries knows the pathway to freedom, his freedom, is war against the oppressor.
A key front in the "war on terror" needs to include stopping the supply of new terrorists. It's one thing (an important one at that) to go after the known terrorists; but, terrorists aren't born terrorists. Ordinary people are made into terrorists by some radicalizing event. The life of Ayman al Zawahiri is a text-book example of how an otherwise decent fellow is radicalized into a monster.
I don't have the easy answer to how to do that. Frankly, I wouldn't trust anyone who said they did. But I don't see anyone in authority even trying to work on the problem.