since jdk 1.4. Not exactly what I'd call "prior art", since it's more or less recent.
Castor for XML is older, and maybe (I'm not sure) even JAXB. DOM is slightly different, but may be applicable. Anyway, I'm sure there other examples. I don't see Microsoft as a pioneer in this.
Not correct, I think. IIRC, they expect 20% to fail miserabily, 30% to not give any benefit at all, 30% to give very little benefits, and 20% to compense for the full stack.
You know, like Natalie Portman naked in a beowulf cluster of P2P hottub iPods gets hacked by the next generation of Perl monkeys?
I really don't know if it makes sense, but boy, sure I would click on it to know more about the subject. And just by the words combinations it would be #1 in google for about a week, no matter what you look for.
It has one entire chapter dedicated to this concrete subject.
Long story short, they shouldn't "kill the messenger". It is also something that has an entire chapter in the book of Bill Gates. Bad news exist. Management should not take aggresive measures on that, nor kill the messenger, but _know_ about it, because if they dont, they lack feedback, and consecuences get usually worse.
That is how things should be.
Usually, upper management are a bunch of uneducated bastards that only know about business meals and profit margin.
So, even when you know that you should speak with them, and even if you knew what things they should know (which do not involve anything extremely technical), telling it will probably not make a difference.
They're comparing only the performance. It's a valid comparison. If I had a project where performance is #1 priority, I would check PHP if I thought it was faster than Java, and wouldn't mind if it's narrower in scope.
After seeing this article, I think I should stick with Java + Webwork + Velocity, given that constraint. IfY(S, JSP is just fine.
It's not that. There are TONS (and I really mean TONS) of documents explaining dozens of ways to make Java scale. They have put lots of stress in making it work well under heavy load. So, when someone argues that a framework as PHP with considerable less work on the subject scales better, he DARES.
we use PHP here for huge web applications.. we have six servers pumping out one website and it connects to a redundant database server.
The same system in java probably would not work, and if so would take up so many resources as to be no efficient.
Well, I have seen a bit of Java the last four years, and can tell you two things:
1.- To check the performance of things you have to _measure_ and not _guess_. Words from the gurus out there, not mine. If you have not seen a system like the one you describe with java (I have mounted two quite more complex at the time of this writing) you just don't know. You are guessing.
2.- I don't know PHP (beyond the basics, that is) but with the description that you give, Java can do the job quite well. And.NET. And perl. Because, basically, you are saying nothing about the stress of your system.
Sorry, but java is being used in quite big systems out there. It scales well. What you have to prove is that PHP scales BETTER.
>>>> Now, this will go through the vector, and call the appropriate version of do-foo for each time in the vector. The cool thing is that it doesn't matter what the types in the vector are, or how they're related. If they have a do-foo() defined for their class, the runtime will do the right thing automatically. In Java, to get the same effect, you'd have to use check to type of each variable, cast it to the right type, then call do-foo() for each type that could be in the vector.//// In Java we have interfaces. The fact is that the code that you propose is not encapsulated . Declaring an array of type the desired interface will achieve the same goal. And it would be encapsulated.
>>>>> Certainly, its not any more flexible than C++.///// You are only looking at the compiler, and are right. But check the runtime for BCEL-like features (capacity to generate bytecodes on the fly at runtime as if they were compiled) and introspection (capacity of finding every attribute and method of a class without knowing its type). I think we don't understand "flexible" the same way.
>>>> let vec = make(limited(, of: ));//// supported on jdk 1.5 to be released by the end of year
>>>> again, Lisp, Dylan, even C in the case of the SAFEcode project//// They don't check on COMPILED code if someone does something nasty. I mean, the goal is to get different permission levels on different classes inside the same executable, something that the OS doesn't allow me to do.
>>>>> Lots of languages are natively compiled and fully portable at the same time.//// You are telling me that you have libraries to connect to different databases, native sockets, multithreading, serial ports and, say, window handling (don't tell me seriously about using Qt, please) that works on sun, hp, win32 and alpha and don't mind if it's little or big endian?
Do you have photos?
Damn, the only reason on the end of the road to use java it's because you have thousands of libraries to get real work done. For those who have to deal with hundreds of functional points it really is worth the pain of lack of performance.
If I recall correctly GSM (the old system used in Europe) has a speed limit for the user of 135 km/h. Over that speed the switch between two GSM nodes can have some trouble and even two calls can switch (and you end up with some other phone call).
Does anybody know how to play non-DVD subtitles with an AVI? I have a japanese movie with subtitles in a file apart and only with -sub it does not work.
Ok, I have RTFM and know this is not the best place in the world to ask for it. But hey, I'm desperate:))
Sorry but I install RedHat in 30 minutes answering four questions. Lately I installed Debian Woody/Sarge (and realized this because I was in a big hurry) and it took four hours and a second machine connected to the Internet.
Most people I know thanks the ease of use of RedHat more than the completeness of Debian. Sure, it is more professional, but I wopulkd be crazy if the first distro to be installed was this.
> For Java, a smart person with a little computer background should be able to figure it out from the language definition.
Bullshit, no offence intended. In any language, pretending that you can see the syntax and infere the rest of the knowledge demonstrates lack of knowledge.
Imagine I have knowledge on C and C++, and lots of computer background. Now let me know how does that show me how to organize my application to achieve scalability, how to improve memory use (since there is no garbage collector on C), how to use reflection or which is the best Collection model to use. A quite _big_ API to try to learn "looking the language definition".
And the trails online from sun don't address but two of these topics, and partially. If you try to learn Java this way (or bash, fwiw) you won't achieve great level of knowledge
http://www.theserverside.com -> J2EE news, great level. Also J2EE patterns and book drafts. Great community. http://www.jguru.com -> great faqs (quite in-depth content) and introductory texts also. Not any news, though http://java.sun.com -> search for the blueprints, and the javaOne slides each year to follow the state-of-the-art on java technology, resumed. The community sucks, though. http://c2.com -> not java, but great pattern repository, great community, and take years to read *part* of it.
> The previous attempt failed miserably: some > people just don't want to switch, some people > honestly just don't have the mental capacity to > understand the difference between the two systems, > and relearning a new system just isn't something > that they can do
You're right: since the introduction of euro half of the population in Europe didn't learn de new system, the economy crumbled and we are again changing cows for wheat and coding html for food.
Oooops. Nope we didn't. That may be because we don't have old people nor dumbasses here. And we sure did't like our older coins, like dracmas, the older coin in the world. Well, used to be.
Sorry, I also have a 99% IQ test that proves that IQ tests aren't worth the paper they are written on. They are only orientative.
On the subject here, making cross-browser scripts costs more ONLY THE FIRST TIME. Once you learn what to use instead of document.all and other MS crap over the code, the same code will work on Mozilla and IE, period. There is 1% of different code in a standards-compliant code full application (I mean, over 100 server-generated web pages and full of pretty jscript effects).
If you're bitching about having to work twice it's because you have never tried.
I mean, almost every Linux howto I have seen on this subject (ipchains, iptables, ipforward) has been written by a man with in-depth knowledge on this matter that works for a company whose name is included in the same HOWTO.
I would look in those first. They knowledge on certain matters has been approved by the whole OS community which has seen their HOWTO and agree with it.
Slashdot Mirror
It's hard to pick the exact query that does not include other areas, but hey:
b untu
http://www.google.com/trends?q=windows+vista%2C+u
since jdk 1.4. Not exactly what I'd call "prior art", since it's more or less recent.
Castor for XML is older, and maybe (I'm not sure) even JAXB. DOM is slightly different, but may be applicable. Anyway, I'm sure there other examples. I don't see Microsoft as a pioneer in this.
Not correct, I think.
IIRC, they expect 20% to fail miserabily, 30% to not give any benefit at all, 30% to give very little benefits, and 20% to compense for the full stack.
according to wikipedia, "anywhere from 20 to 90% of the enterprises funded fail to return the invested capital"
http://en.wikipedia.org/wiki/Venture_capital
You know, like Natalie Portman naked in a beowulf cluster of P2P hottub iPods gets hacked by the next generation of Perl monkeys?
I really don't know if it makes sense, but boy, sure I would click on it to know more about the subject. And just by the words combinations it would be #1 in google for about a week, no matter what you look for.
It has one entire chapter dedicated to this concrete subject.
Long story short, they shouldn't "kill the messenger". It is also something that has an entire chapter in the book of Bill Gates. Bad news exist. Management should not take aggresive measures on that, nor kill the messenger, but _know_ about it, because if they dont, they lack feedback, and consecuences get usually worse.
That is how things should be.
Usually, upper management are a bunch of uneducated bastards that only know about business meals and profit margin.
So, even when you know that you should speak with them, and even if you knew what things they should know (which do not involve anything extremely technical), telling it will probably not make a difference.
Actually, it does. In China and Japan, at least.
1. Get Cuban nationality
2. Send proof of me guilty to M$
3. Profit!
They're comparing only the performance. It's a valid comparison. If I had a project where performance is #1 priority, I would check PHP if I thought it was faster than Java, and wouldn't mind if it's narrower in scope.
After seeing this article, I think I should stick with Java + Webwork + Velocity, given that constraint. IfY(S, JSP is just fine.
It's not that. There are TONS (and I really mean TONS) of documents explaining dozens of ways to make Java scale. They have put lots of stress in making it work well under heavy load. So, when someone argues that a framework as PHP with considerable less work on the subject scales better, he DARES.
I know of a PIII 600Mhz running with JBoss 3 (started with 2.4) entity beans with cache activated and postgres that works like a charm.
we use PHP here for huge web applications.. we have six servers pumping out one website and it connects to a redundant database server.
The same system in java probably would not work, and if so would take up so many resources as to be no efficient.
Well, I have seen a bit of Java the last four years, and can tell you two things:
1.- To check the performance of things you have to _measure_ and not _guess_. Words from the gurus out there, not mine. If you have not seen a system like the one you describe with java (I have mounted two quite more complex at the time of this writing) you just don't know. You are guessing.
2.- I don't know PHP (beyond the basics, that is) but with the description that you give, Java can do the job quite well. And
Sorry, but java is being used in quite big systems out there. It scales well. What you have to prove is that PHP scales BETTER.
Basically they say that Resin isn't that fast. I think this has started with Tomcat 4.
What I really want to know is, what the hell does yukkuri-hanashite-kudasai mean?
We really need 20% more explanations...
Man, you didn't say you accepted it!
Duh!
////
/////
////
////
////
>>>>
Now, this will go through the vector, and call the appropriate version of do-foo for each time in the vector. The cool thing is that it doesn't matter what the types in the vector are, or how they're related. If they have a do-foo() defined for their class, the runtime will do the right thing automatically. In Java, to get the same effect, you'd have to use check to type of each variable, cast it to the right type, then call do-foo() for each type that could be in the vector.
In Java we have interfaces. The fact is that the code that you propose is not encapsulated . Declaring an array of type the desired interface will achieve the same goal. And it would be encapsulated.
>>>>>
Certainly, its not any more flexible than C++.
You are only looking at the compiler, and are right. But check the runtime for BCEL-like features (capacity to generate bytecodes on the fly at runtime as if they were compiled) and introspection (capacity of finding every attribute and method of a class without knowing its type). I think we don't understand "flexible" the same way.
>>>>
let vec = make(limited(, of: ));
supported on jdk 1.5 to be released by the end of year
>>>>
again, Lisp, Dylan, even C in the case of the SAFEcode project
They don't check on COMPILED code if someone does something nasty. I mean, the goal is to get different permission levels on different classes inside the same executable, something that the OS doesn't allow me to do.
>>>>>
Lots of languages are natively compiled and fully portable at the same time.
You are telling me that you have libraries to connect to different databases, native sockets, multithreading, serial ports and, say, window handling (don't tell me seriously about using Qt, please) that works on sun, hp, win32 and alpha and don't mind if it's little or big endian?
Do you have photos?
Damn, the only reason on the end of the road to use java it's because you have thousands of libraries to get real work done. For those who have to deal with hundreds of functional points it really is worth the pain of lack of performance.
The rest imho is fud.
If I recall correctly GSM (the old system used in Europe) has a speed limit for the user of 135 km/h. Over that speed the switch between two GSM nodes can have some trouble and even two calls can switch (and you end up with some other phone call).
Highways have the same problem too.
THANKS thanks THX thx zanks man.
:)
Will try it. THanks for the advice
Does anybody know how to play non-DVD subtitles with an AVI? I have a japanese movie with subtitles in a file apart and only with -sub it does not work.
:))
Ok, I have RTFM and know this is not the best place in the world to ask for it. But hey, I'm desperate
Sorry but I install RedHat in 30 minutes answering four questions. Lately I installed Debian Woody/Sarge (and realized this because I was in a big hurry) and it took four hours and a second machine connected to the Internet.
Most people I know thanks the ease of use of RedHat more than the completeness of Debian. Sure, it is more professional, but I wopulkd be crazy if the first distro to be installed was this.
> For Java, a smart person with a little computer background should be able to figure it out from the language definition.
Bullshit, no offence intended. In any language, pretending that you can see the syntax and infere the rest of the knowledge demonstrates lack of knowledge.
Imagine I have knowledge on C and C++, and lots of computer background. Now let me know how does that show me how to organize my application to achieve scalability, how to improve memory use (since there is no garbage collector on C), how to use reflection or which is the best Collection model to use. A quite _big_ API to try to learn "looking the language definition".
And the trails online from sun don't address but two of these topics, and partially. If you try to learn Java this way (or bash, fwiw) you won't achieve great level of knowledge
Since I cannot see them, my .02:
http://www.theserverside.com -> J2EE news, great level. Also J2EE patterns and book drafts. Great community.
http://www.jguru.com -> great faqs (quite in-depth content) and introductory texts also. Not any news, though
http://java.sun.com -> search for the blueprints, and the javaOne slides each year to follow the state-of-the-art on java technology, resumed. The community sucks, though.
http://c2.com -> not java, but great pattern repository, great community, and take years to read *part* of it.
> The previous attempt failed miserably: some
> people just don't want to switch, some people
> honestly just don't have the mental capacity to
> understand the difference between the two systems,
> and relearning a new system just isn't something
> that they can do
You're right: since the introduction of euro half of the population in Europe didn't learn de new system, the economy crumbled and we are again changing cows for wheat and coding html for food.
Oooops. Nope we didn't. That may be because we don't have old people nor dumbasses here. And we sure did't like our older coins, like dracmas, the older coin in the world. Well, used to be.
Right now they're slashdotted down :)
Sorry, I also have a 99% IQ test that proves that IQ tests aren't worth the paper they are written on. They are only orientative.
On the subject here, making cross-browser scripts costs more ONLY THE FIRST TIME. Once you learn what to use instead of document.all and other MS crap over the code, the same code will work on Mozilla and IE, period. There is 1% of different code in a standards-compliant code full application (I mean, over 100 server-generated web pages and full of pretty jscript effects).
If you're bitching about having to work twice it's because you have never tried.
I mean, almost every Linux howto I have seen on this subject (ipchains, iptables, ipforward) has been written by a man with in-depth knowledge on this matter that works for a company whose name is included in the same HOWTO.
I would look in those first. They knowledge on certain matters has been approved by the whole OS community which has seen their HOWTO and agree with it.