Proper memory management through design will also keep the process from dragging down the machine. It'll also make it perform well when you've reached the design limits. When you reach the limits of the JVM you get all sorts of trouble.
I work on large scale server processes in Java as well, and I find it quite annoying. Have you done the same work in C or C++? I've written a multicast file delivery system, implemented both in Java and C++. It serves over 700 locations around the US. The C++ version outperforms Java 2:1 and does so with much less performance impact.
Java is simply unsuited for some things by virtue of its JVM. It eliminates many of the power pieces that your OS may provide.
I have a wonderful example that would be best served with a shared memory segment and a couple of semaphores, routine things I've done in C or C++ that simply can't be done routinely in Java.
There's also the JVM memory limits. My web app is also routinely running out of memory at the default JVM settings. We've had to push the memory settings up and up to make sure the thing stays running. I simply hate the idea that when there's memory out there I can't simply use it. I've got to stop and reconfigure the app server and start all over again.
Also, for apps that routinely go through large pieces of dynamic memory allocation, the JVM's garbage collector simply doesn't keep up. It's nice to not have to worrry about freeing allocated memory, but dammit, when I want that memory freed I want to do it and know it's available again. Someone needs to give me a JVM with a free method as an extension.
For these reasons, and probably a few more, Java IMHO is not suited for large scale server processes.
Since HP-UX 11 came out I believe/bin/sh is the POSIX shell. Not really Bourne exactly.
And I think most of us are bright enough to test our code a bit on those platforms for which we are responsible. Whining about shell discrepancies is a bit lame. What happened to the news?
Re:My crutch doesn't exist because I need it
on
God's Debris
·
· Score: 1
I don't care to say who's right or wrong, only to
point out the flaws in one side of the argument.
Ben was a famous Deist, believing in a supreme 'being' of which little could ever be known. I happen to follow the same line of thought.
Re:My crutch doesn't exist because I need it
on
God's Debris
·
· Score: 1
First, there are not various equal, dissimilar, and arbitrary lines of science from which I am asked to choose. True science is measurable and repeatable. That doesn't mean provable, but it's a far cry more than you can say for religion.
Your argument about bouyancy is circular, and you've missed the point. If I told you that Phil in accounting walked on water you'd probably not bet your soul on it, even if you don't know Phil. But for some reason a book 2000 years old that's been translated a dozen times by religious fanatics is supposed to convince you? That's a tremendous double standard. That's the point of the "don't trust medicine" idea.
"Don't take it literally" is a standard argument that comes back from a bible thumper after an argument. I don't take it literally. I don't take it at all. But if I can't take it literally what do I have but reason to make sense of it? Why should the "correct" interpretation be contrary to reason and experience? We even have a prominent and named theory on this called Occam's Razor. Don't introduce things contrary or unrelated to the experience. We don't sit still for that in science or business or the rest of our daily lives. Why do people accept it with religion?
Re:My crutch doesn't exist because I need it
on
God's Debris
·
· Score: 1
God must be a poor doctor. He's been healing for 2000 years and hasn't gotten it right yet.
None of this is a matter of proof. I think we both know it's impossible. It's a matter of credibility. What about Christianity is any more credible than the teachings of Islam, Hinduism, Judaism, or Buddhism? Why do you believe Christianity? It's what was given to you.
I don't suppose that either creationism or the Big Bang is correct. I don't think I need to know how it started. I've enough knowledge of the world around me to gauge truth and reality. What good would it do for me to know when and how it started? I still have to live by the natural order of the Universe. That is the only truth. There's no escaping it.
But to go down the line of proof, I think there are probably plenty of chemists and physicists that can prove that no one 2000 years ago had the ability to turn water into wine, or even alcohol for that matter. And recently there has been research done on Carribean islands proving distinct but similar evolution of anoles between neighboring islands. DNA comparison shows similar ancestry for various types on one island, but not between similar types between islands. And I'm sure that bacteriology has plenty of examples of evolved resistence to antibiotics.
Should you say that none of these are proof, then stop going to the doctor, because the same methods are used in research for making medicines and other medical treatment, and you should discredit it just the same.
My biggest argument is that what you believe should be consistent. How can you claim to believe two things that are disparate? I assure you that there is no reason to believe that any man can walk on water. There is nothing about a man or about water that can ever cause you to think it possible. Either you believe that what you and I and everyone else has experienced about water and people, or you believe that Jesus walked on water. How can it be both? Your sole account of the single instance of this comes from a book of collaborators assembled 400 years after the event. Is that more credible than centuries of consistent experience, as well as your own experience, that says it couldn't have happened?
You don't have to answer because your religious crutch says you don't. You have faith. That supercedes any need for reason or accountability.
Re:My crutch doesn't exist because I need it
on
God's Debris
·
· Score: 2, Interesting
The point of a crutch is to help one along until he can function at full capacity, without hindrance or help. Existence of the crutch isn't an issue. Completeness and truth of the condition are the issue. While there is need for a crutch the condition cannot be in its true and complete form.
Cultivated need for a crutch fosters a false reliance upon it, as well as a comfort from having it that goes beyond its utility.
The habituation to that comfort leads one to further disassociate the crutch with reality. What evidence in your daily life makes you believe that God, existence granted, needs anything? Any notion of the nature of God is a construct of human emotion, and religion is the mythology built of those constructs.
If the Christian God created the universe, why do so many of the 'truths' of the Christian religion clash so horribly with the truths we experience in our daily lives? People don't rise from the dead. Water doesn't turn into wine, and bread does not turn into flesh. Alchemy fell out of favor hundreds of years ago. I think we know pretty well that one little boat doesn't hold enough animals to populate the Earth. And I think we know that Adam and Eve are as real as Mickey and Minney. It just doesn't fit with the records of humanity that we have found.
(By the way, try to build a case for Christianity without the Adam and Eve story.)
Blue collar versus scientist is a poor example of stereotyping. Wisdom has nothing to do with occupation. Education is, however, a way to gain insight that leads to wisdom.
Personally, I don't care if you use a crutch, but don't ever expect me to take you seriously in any conversation regarding truth or reality.
Pile on the memory for whichever machine you put VAJ on. All those features make it a hog at the memory trough. I agree though, that VAJ has been the best I've worked with. I've spent significant time with VAJ, JBuilder, and Visual Cafe. If you need debugging, VAJ rocks.
Actually, I think they're trying to overcome the ignorance of the customer base who have made Mhz the incorrect measurement of a CPU's performance. They're not trying to give them false information. They're trying to keep the customer from making invalid conclusions based on the information they've given. Big difference.
So you make a dynamic page and write a cacher into your app server. You check if the page is in the cache. If it is, check so see if it's updated. If not, throw the cached page. Otherwise rebuild and throw the new page, and then cache it.
Why buy a tool to do it when you can write it in an hour or two? And you don't need a webmaster to manage it.
Anne McCaffrey has been a mainstay of Sci-Fi for the last 20 years or so. The Dragon Riders of Pern was a masterful series. It got played out way too far, but the original three were wonderful.
Tolkien knew when to quit. He quit when the story was done. Jordan is going to quit when his publisher stops paying. The story died a long time (and several books) ago.
No, no bad taste at all. In fact, I wonder what you're about. What on Earth are you working for if it's not to have Linux and other open source efforts become a valuable piece of the IT industry? Open Source software isn't a flag to be waved to say, "I'm on board!" Software is a commodity required for business, and that's its value. If the OS community never meets the needs of business, never becomes part of the busines MO, it's got to die.
Tilting at windmills is exciting and all, but if you're not going to put something where the windmill stood, all you've got left is a pile of rubble and a fool who made it. Everyone else is just going to stand confused, and probably angry.
My feeling is that plain old text files, along with a descriptive file of how the text file is laid out, is overall the best solution for most data interchanges between businesses.
I think your assessment is short sighted. To suggest that there is an overall solution is silly.
I do agree, however, in the essence of your comment. XML is nothing special. It's no better or worse than any of the dozens of formats we've all come up with.
What's significant is that it's been agreed upon and is being implemented widely. I don't have to come up with a scheme when I want one, and i dont' have to code a parser or translator when I need one. I just go get it. That is the only thing significant about XML.
There's this German textile management system . ..
You mean like a bureau or drawers? I store my linens in a closet, my socks in the dresser. Most of my pants I have on hangers. I don't think any of them are German though. And I don't think there's an XML standard for pants.
No irony at all. The AUTHOR put them in, with the links he provided. What's the relation between that and what M$ wants to do? Why should that be ironic?
is the punctuation! How can you write anything about that ridiculous piece of M$ trash and make it readable! You never know where one sentence ends and another starts!
When you're writing in layers you have to design the interface between layers. The frequent benefit to designing software in layers is to provide abstraction, allowing the substitution of one layer with an upgrade or alternative. If you've not designed good layer interfaces that indeed keep the layers separate, you've likely failed to achieve that objective.
You can have any number of interface mechanisms. One application I wrote had a message queue that one layer wrote to and another read from. It had a secondary preemtive queue into which high priority messages were written that would control the reading from the other queue.
If your layers are actual processes you can use a pipe and any number of messaging options. If you're writing in Unix land with C++ and your layers are threads, you can write the signal handler to cause specific actions in particular threads.
In Java you can make your own event notification listeners. Classes implement an EventListener interface that you design yourself. They register with a static EventDispatcher thread. The lower layer can implement an EventProducer interface, which also registers with the event dispatcher. When it has an error you can post the event. All listeners for that event type take the appropriate action.
The key is defining the interface, and making sure that the interface is a clean break between the layers.
First, if you can't transition between one word processor and another, your challenges in the workplace are going to be far greater than you imagine. Knowing M$ Office is not a profession. It's one instance of a widely used skill called writing. It's the same skill that is required to use a pencil. Office keeps you from needing good penmanship and arithmetic skills.
Secondly, I hate to break it to you, but I and thousands of others in my field, the professional consulting field, make a wonderful living, making far greater money than your typical MSCE, without ever writing a stitch of Windows code. In my tenth year in the business, I've never seen an enterprise billing, payroll, reporting system written using SQLServer and VB. That's real world.
Actually, having done a bit more research, X was indeed made for X-windows. But in the big scheme of things, it's still an event messaging system, and I do know that it can be used for more than graphics.
In college we played with an 8-way with CPUs set up as a cube. Each was connected to three others. In fact, it was called the HyperCube. It was a very effective setup.
I'm absolutely guessing, but I'd think that having anything but a symmetrical setup for any processor is going to cause engineering headaches, not to mention logical complexity.
Re:Not a recipe for improving games
on
Ultima Revived
·
· Score: 1
Most people simply don't have the resources and/or skills to develop something new. What you're saying is like saying, "The Mona Lisa is old and I'm tired of seeing it. I'll just sit down and paint a new one." It's much easier to make a copy and give Lisa a new hairdo, some jewelry, and a new frock. Or simply to put it in a new frame. If it makes people happy, what's the cost?
Taking up something that's been left standing is a great option. The insight or technology that has come in the duration since the game's heyday might allow the new developer to overcome those very shortcomings that made it obsolete. Those people who are revisiting these old games are the ones who probably spent hundreds of hours mastering it and the ones who know what would have made it a better experience for them.
Slashdot Mirror
At least there's an intelligent key. Maybe it'll keep some dumb-ass from taking something like this seriously (and into traffic!)
Proper memory management through design will also keep the process from dragging down the machine. It'll also make it perform well when you've reached the design limits. When you reach the limits of the JVM you get all sorts of trouble.
I work on large scale server processes in Java as well, and I find it quite annoying. Have you done the same work in C or C++? I've written a multicast file delivery system, implemented both in Java and C++. It serves over 700 locations around the US. The C++ version outperforms Java 2:1 and does so with much less performance impact.
Java is simply unsuited for some things by virtue of its JVM. It eliminates many of the power pieces that your OS may provide.
I have a wonderful example that would be best served with a shared memory segment and a couple of semaphores, routine things I've done in C or C++ that simply can't be done routinely in Java.
There's also the JVM memory limits. My web app is also routinely running out of memory at the default JVM settings. We've had to push the memory settings up and up to make sure the thing stays running. I simply hate the idea that when there's memory out there I can't simply use it. I've got to stop and reconfigure the app server and start all over again.
Also, for apps that routinely go through large pieces of dynamic memory allocation, the JVM's garbage collector simply doesn't keep up. It's nice to not have to worrry about freeing allocated memory, but dammit, when I want that memory freed I want to do it and know it's available again. Someone needs to give me a JVM with a free method as an extension.
For these reasons, and probably a few more, Java IMHO is not suited for large scale server processes.
Since HP-UX 11 came out I believe /bin/sh is the POSIX shell. Not really Bourne exactly.
And I think most of us are bright enough to test our code a bit on those platforms for which we are responsible. Whining about shell discrepancies is a bit lame. What happened to the news?
I don't care to say who's right or wrong, only to
point out the flaws in one side of the argument.
An inconsistent mind can only be called confused.
Ben was a famous Deist, believing in a supreme 'being' of which little could ever be known. I happen to follow the same line of thought.
First, there are not various equal, dissimilar, and arbitrary lines of science from which I am asked to choose. True science is measurable and repeatable. That doesn't mean provable, but it's a far cry more than you can say for religion.
Your argument about bouyancy is circular, and you've missed the point. If I told you that Phil in accounting walked on water you'd probably not bet your soul on it, even if you don't know Phil. But for some reason a book 2000 years old that's been translated a dozen times by religious fanatics is supposed to convince you? That's a tremendous double standard. That's the point of the "don't trust medicine" idea.
"Don't take it literally" is a standard argument that comes back from a bible thumper after an argument. I don't take it literally. I don't take it at all. But if I can't take it literally what do I have but reason to make sense of it? Why should the "correct" interpretation be contrary to reason and experience? We even have a prominent and named theory on this called Occam's Razor. Don't introduce things contrary or unrelated to the experience. We don't sit still for that in science or business or the rest of our daily lives. Why do people accept it with religion?
God must be a poor doctor. He's been healing for 2000 years and hasn't gotten it right yet.
None of this is a matter of proof. I think we both know it's impossible. It's a matter of credibility. What about Christianity is any more credible than the teachings of Islam, Hinduism, Judaism, or Buddhism? Why do you believe Christianity? It's what was given to you.
I don't suppose that either creationism or the Big Bang is correct. I don't think I need to know how it started. I've enough knowledge of the world around me to gauge truth and reality. What good would it do for me to know when and how it started? I still have to live by the natural order of the Universe. That is the only truth. There's no escaping it.
But to go down the line of proof, I think there are probably plenty of chemists and physicists that can prove that no one 2000 years ago had the ability to turn water into wine, or even alcohol for that matter. And recently there has been research done on Carribean islands proving distinct but similar evolution of anoles between neighboring islands. DNA comparison shows similar ancestry for various types on one island, but not between similar types between islands. And I'm sure that bacteriology has plenty of examples of evolved resistence to antibiotics.
Should you say that none of these are proof, then stop going to the doctor, because the same methods are used in research for making medicines and other medical treatment, and you should discredit it just the same.
My biggest argument is that what you believe should be consistent. How can you claim to believe two things that are disparate? I assure you that there is no reason to believe that any man can walk on water. There is nothing about a man or about water that can ever cause you to think it possible. Either you believe that what you and I and everyone else has experienced about water and people, or you believe that Jesus walked on water. How can it be both? Your sole account of the single instance of this comes from a book of collaborators assembled 400 years after the event. Is that more credible than centuries of consistent experience, as well as your own experience, that says it couldn't have happened?
You don't have to answer because your religious crutch says you don't. You have faith. That supercedes any need for reason or accountability.
The point of a crutch is to help one along until he can function at full capacity, without hindrance or help. Existence of the crutch isn't an issue. Completeness and truth of the condition are the issue. While there is need for a crutch the condition cannot be in its true and complete form.
Cultivated need for a crutch fosters a false reliance upon it, as well as a comfort from having it that goes beyond its utility.
The habituation to that comfort leads one to further disassociate the crutch with reality. What evidence in your daily life makes you believe that God, existence granted, needs anything? Any notion of the nature of God is a construct of human emotion, and religion is the mythology built of those constructs.
If the Christian God created the universe, why do so many of the 'truths' of the Christian religion clash so horribly with the truths we experience in our daily lives? People don't rise from the dead. Water doesn't turn into wine, and bread does not turn into flesh. Alchemy fell out of favor hundreds of years ago. I think we know pretty well that one little boat doesn't hold enough animals to populate the Earth. And I think we know that Adam and Eve are as real as Mickey and Minney. It just doesn't fit with the records of humanity that we have found.
(By the way, try to build a case for Christianity without the Adam and Eve story.)
Blue collar versus scientist is a poor example of stereotyping. Wisdom has nothing to do with occupation. Education is, however, a way to gain insight that leads to wisdom.
Personally, I don't care if you use a crutch, but don't ever expect me to take you seriously in any conversation regarding truth or reality.
Pile on the memory for whichever machine you put VAJ on. All those features make it a hog at the memory trough. I agree though, that VAJ has been the best I've worked with. I've spent significant time with VAJ, JBuilder, and Visual Cafe. If you need debugging, VAJ rocks.
Oh, that's meaningful. Looking for how many Linux users have Logitech mice hooked up should shed just as much insight into this question.
Actually, I think they're trying to overcome the ignorance of the customer base who have made Mhz the incorrect measurement of a CPU's performance. They're not trying to give them false information. They're trying to keep the customer from making invalid conclusions based on the information they've given. Big difference.
So you make a dynamic page and write a cacher into your app server. You check if the page is in the cache. If it is, check so see if it's updated. If not, throw the cached page. Otherwise rebuild and throw the new page, and then cache it.
Why buy a tool to do it when you can write it in an hour or two? And you don't need a webmaster to manage it.
Anne McCaffrey has been a mainstay of Sci-Fi for the last 20 years or so. The Dragon Riders of Pern was a masterful series. It got played out way too far, but the original three were wonderful.
Tolkien knew when to quit. He quit when the story was done. Jordan is going to quit when his publisher stops paying. The story died a long time (and several books) ago.
No, no bad taste at all. In fact, I wonder what you're about. What on Earth are you working for if it's not to have Linux and other open source efforts become a valuable piece of the IT industry? Open Source software isn't a flag to be waved to say, "I'm on board!" Software is a commodity required for business, and that's its value. If the OS community never meets the needs of business, never becomes part of the busines MO, it's got to die.
Tilting at windmills is exciting and all, but if you're not going to put something where the windmill stood, all you've got left is a pile of rubble and a fool who made it. Everyone else is just going to stand confused, and probably angry.
My feeling is that plain old text files, along with a descriptive file of how the text file is laid out, is overall the best solution for most data interchanges between businesses.
I think your assessment is short sighted. To suggest that there is an overall solution is silly.
I do agree, however, in the essence of your comment. XML is nothing special. It's no better or worse than any of the dozens of formats we've all come up with.
What's significant is that it's been agreed upon and is being implemented widely. I don't have to come up with a scheme when I want one, and i dont' have to code a parser or translator when I need one. I just go get it. That is the only thing significant about XML.
There's this German textile management system . . .
You mean like a bureau or drawers? I store my linens in a closet, my socks in the dresser. Most of my pants I have on hangers. I don't think any of them are German though. And I don't think there's an XML standard for pants.
No irony at all. The AUTHOR put them in, with the links he provided. What's the relation between that and what M$ wants to do? Why should that be ironic?
is the punctuation! How can you write anything about that ridiculous piece of M$ trash and make it readable! You never know where one sentence ends and another starts!
When you're writing in layers you have to design the interface between layers. The frequent benefit to designing software in layers is to provide abstraction, allowing the substitution of one layer with an upgrade or alternative. If you've not designed good layer interfaces that indeed keep the layers separate, you've likely failed to achieve that objective.
You can have any number of interface mechanisms. One application I wrote had a message queue that one layer wrote to and another read from. It had a secondary preemtive queue into which high priority messages were written that would control the reading from the other queue.
If your layers are actual processes you can use a pipe and any number of messaging options. If you're writing in Unix land with C++ and your layers are threads, you can write the signal handler to cause specific actions in particular threads.
In Java you can make your own event notification listeners. Classes implement an EventListener interface that you design yourself. They register with a static EventDispatcher thread. The lower layer can implement an EventProducer interface, which also registers with the event dispatcher. When it has an error you can post the event. All listeners for that event type take the appropriate action.
The key is defining the interface, and making sure that the interface is a clean break between the layers.
First, if you can't transition between one word processor and another, your challenges in the workplace are going to be far greater than you imagine. Knowing M$ Office is not a profession. It's one instance of a widely used skill called writing. It's the same skill that is required to use a pencil. Office keeps you from needing good penmanship and arithmetic skills.
Secondly, I hate to break it to you, but I and thousands of others in my field, the professional consulting field, make a wonderful living, making far greater money than your typical MSCE, without ever writing a stitch of Windows code. In my tenth year in the business, I've never seen an enterprise billing, payroll, reporting system written using SQLServer and VB. That's real world.
Actually, having done a bit more research, X was indeed made for X-windows. But in the big scheme of things, it's still an event messaging system, and I do know that it can be used for more than graphics.
In college we played with an 8-way with CPUs set up as a cube. Each was connected to three others. In fact, it was called the HyperCube. It was a very effective setup.
I'm absolutely guessing, but I'd think that having anything but a symmetrical setup for any processor is going to cause engineering headaches, not to mention logical complexity.
Most people simply don't have the resources and/or skills to develop something new. What you're saying is like saying, "The Mona Lisa is old and I'm tired of seeing it. I'll just sit down and paint a new one." It's much easier to make a copy and give Lisa a new hairdo, some jewelry, and a new frock. Or simply to put it in a new frame. If it makes people happy, what's the cost?
Taking up something that's been left standing is a great option. The insight or technology that has come in the duration since the game's heyday might allow the new developer to overcome those very shortcomings that made it obsolete. Those people who are revisiting these old games are the ones who probably spent hundreds of hours mastering it and the ones who know what would have made it a better experience for them.