New bytecode from JDK 1.5 is incompatible with previous JVMs. And that's the funniest thing.
Re:No Generics until 2006-2007 ??
on
Gosling on Computing
·
· Score: 2, Informative
No. This is just a FUD.
.NET bytecodes has a well-known "verifiable subset" which can be automatically verified to be correct just like Java bytecode. CLR can be configured to reject non-safe code in some domains and accept it in another domains. For example, you can configure CLR to behave just like an applet sandbox for downloaded applications.
I repeat, Java bytecode is just a subset of IL. So you can do in.NET everything you can do in Java.
Yes, I understand that in some cases JIT can optimize casting away (I'm somewhat familiar with JVM source code). But it's not possible in all cases, and sometimes time spent in JIT optimizer annihilates optimizations.
Besides, it takes about 5000-9000 passes of code to make HotSpot to kick in and optimize a piece of code. It might be OK for long-running server applications, but it can't be tolerated in desktop apps.
.NET is better than Java from engineer point of view, just because Java bytecode is a subset of IL (Intermediate Language). So it's possible to recompile any Java program without native methods to DotNET (Microsoft has tool which translates Java into C#).
I really like.NET, right now I'm hacking Mono sources trying to add exact garbage collector instead of conservative Boehm GC.
Take a look at C# 2.0 draft. That's the right way to implement generics.
Basicly, Java generic classes are common classes with some compile-time information. Compiler automaticaly inserts type casts when it's neccessary. But resulting byte-code is just the same code with casts from Object. So the only advantage is more typesafeness.
In C# 2.0 generic class are not real classes, they are templates (as in C++) for classes with erased type information. CLR (Common Language Runtime) instantiates parametrized class when it's neccessary (in JIT-time), during instantiation erased types are substituted with specified types, it's the same mechanism as used in C++. Advantages of this implementation: faster code (no redundant casts), smaller size, ability to parametrize templates with promitive types.
Generics are non-existent to me. They won't be anywhere around until 2006-2007. And there will be something much better than Java in that timeframe, IMHO.
DotNET platform now has all chances to bes Java replacement on desktop (OpenSource Mono will help).
I lost all my respect to Gosling after a clumsy attempt to add generics to Java.
The right way to add generics to Java is a radical modification of JVM (Java Virtual Machine), but Sun didn't want to it. So they made an attempt to add generics to Java language without touching JVM. The result of this attempt is a complex scheme of name mangling (just like C++), and some unnecessary overhead. And such implementation _still_ requires some JVM changes and is incompatible with old JVMs. So now we have an ugly generics in Java and Java 5.0 (rebranded J2SE 1.5) incompatible with previous versions.
When will PHP die and will be replaced by something more elegant and useful?
Honestly, PHP is just like BASIC for WEB. Object oriented programming in PHP is still a pain, there's no clear notion of something like packages, language itself is a mess, no standard intermediate 'bytecode', etc.
Why would anyone want to use such a clumsy tool when things like Tapestry, Echo, Struts exist?
No, this is not a troll. I'm just sick of news about PHP.
Yeah, it reminds me a curious story of "Space Race" age.
First American moon landers used microwave radars to measure distance to Moon surface and start retro-rockets just a few meters above the surface.
Russians couldn't make their radar to be precise and light enough to fit into lander so they just used a long pole to start a squib and, consequently, retro-rockets. And this approach worked fine, even during the mission to Venera:)
The moral of this story is: KISS. More of electronics = more of places-of-failure.
Lots of books are untraslatable. Sure, they can be translated to another language, but they'll be meaningless or much less interesting to people of another culture. Besides, languages are very different, some language constructs can't be translated perfectly, and sometimes it can ruin impression of reader. For example, there is no analog of Present Perfect Tense in Russian, but on the other hand, English morphology (rules of words transforming) is poor compared to Russian. That's why short English sentences sometimes are translated into long and ugly Russian ones, and lots of nuance are lost then translating from Russian into English.
Unfortunately, some of untranslatable books are the greatest ones. Even the whole layers of can be literature are untranslatable (folk-lore, some kinds of poetry, etc.).
And sometimes they might think that "The Greatest Book Of 20th Century" is written in their language. So it's pretty unfair to conduct a poll about the greatest book only between English-speaking people. And in the global worldwide poll some obscure Chinese book will probably win:)
I for one, think that Bulgakov's "The Master and Margaret" is the greatest Russian book of 20th century.
I'm a software developer and my last project was to create OpenGroupware-to-Outlook connector. So I had to become very familiar with OpenGroupware internals.
Well, OpenGroupware SUCKS. Its web-interface is hard to use and slow - you need to make 3 clicks just to mark job as "Done", contacts management is even worse.
OpenGroupware server is hard to install and use - there is no good Debian/RPM packages, so you have to read tons of HOWTOs.
From developer perspective OpenGroupware is a nightmare. XMLRPC interface is poorly documented, unoptimized and not very usable. WebDav interface is even less documented. And now the best one: you can't access all functions of OpenGroupware from these interfaces.
I don't like to admit it, but Exchange and Outlook are far superior to OpenGroupware.
We pay about 8-10 cents per megabyte. Here's price list of our local ISP: Price list. It's in Russian, of course, but you can use something like Babelfish to translate it.
Calculators and any other artificial aids should be banned from exams. In fact, in Russia and many other countries calculators (as well as mobile phones) are banned on most exams.
Exam problems should be chosen so that to solve them you shouldn't need calculators/logarithmic rulers/etc.
For example, to compare sqrt(2) and 1.4 you don't need calculator. In fact, naive calculator-solution "1,4142135623730950488016887242097 is greater that 1.4" is incorrect. Correct solution is: "2>(1.4)^2".
Slashdot Mirror
New bytecode from JDK 1.5 is incompatible with previous JVMs. And that's the funniest thing.
No. This is just a FUD.
.NET bytecodes has a well-known "verifiable subset" which can be automatically verified to be correct just like Java bytecode. CLR can be configured to reject non-safe code in some domains and accept it in another domains. For example, you can configure CLR to behave just like an applet sandbox for downloaded applications.
.NET everything you can do in Java.
I repeat, Java bytecode is just a subset of IL. So you can do in
Yes, I understand that in some cases JIT can optimize casting away (I'm somewhat familiar with JVM source code). But it's not possible in all cases, and sometimes time spent in JIT optimizer annihilates optimizations.
Besides, it takes about 5000-9000 passes of code to make HotSpot to kick in and optimize a piece of code. It might be OK for long-running server applications, but it can't be tolerated in desktop apps.
.NET is better than Java from engineer point of view, just because Java bytecode is a subset of IL (Intermediate Language). So it's possible to recompile any Java program without native methods to DotNET (Microsoft has tool which translates Java into C#).
.NET, right now I'm hacking Mono sources trying to add exact garbage collector instead of conservative Boehm GC.
I really like
It took almost 3 years to migrate from J2SE 1.3 to J2SE 1.4 for most of our customers.Even now lots of our customers run JDK 1.3
Now add 3 years to 2005...
Take a look at C# 2.0 draft. That's the right way to implement generics.
Basicly, Java generic classes are common classes with some compile-time information. Compiler automaticaly inserts type casts when it's neccessary. But resulting byte-code is just the same code with casts from Object. So the only advantage is more typesafeness.
In C# 2.0 generic class are not real classes, they are templates (as in C++) for classes with erased type information. CLR (Common Language Runtime) instantiates parametrized class when it's neccessary (in JIT-time), during instantiation erased types are substituted with specified types, it's the same mechanism as used in C++. Advantages of this implementation: faster code (no redundant casts), smaller size, ability to parametrize templates with promitive types.
Generics are non-existent to me. They won't be anywhere around until 2006-2007. And there will be something much better than Java in that timeframe, IMHO.
DotNET platform now has all chances to bes Java replacement on desktop (OpenSource Mono will help).
Java had the last chance to change, but miss
I lost all my respect to Gosling after a clumsy attempt to add generics to Java.
The right way to add generics to Java is a radical modification of JVM (Java Virtual Machine), but Sun didn't want to it. So they made an attempt to add generics to Java language without touching JVM. The result of this attempt is a complex scheme of name mangling (just like C++), and some unnecessary overhead. And such implementation _still_ requires some JVM changes and is incompatible with old JVMs. So now we have an ugly generics in Java and Java 5.0 (rebranded J2SE 1.5) incompatible with previous versions.
When will PHP die and will be replaced by something more elegant and useful?
Honestly, PHP is just like BASIC for WEB. Object oriented programming in PHP is still a pain, there's no clear notion of something like packages, language itself is a mess, no standard intermediate 'bytecode', etc.
Why would anyone want to use such a clumsy tool when things like Tapestry, Echo, Struts exist?
No, this is not a troll. I'm just sick of news about PHP.
Well, why not just make water cooling mandatory for new CPUs, just like Apple did?
Yeah, it reminds me a curious story of "Space Race" age.
:)
First American moon landers used microwave radars to measure distance to Moon surface and start retro-rockets just a few meters above the surface.
Russians couldn't make their radar to be precise and light enough to fit into lander so they just used a long pole to start a squib and, consequently, retro-rockets. And this approach worked fine, even during the mission to Venera
The moral of this story is: KISS. More of electronics = more of places-of-failure.
Lots of books are untraslatable. Sure, they can be translated to another language, but they'll be meaningless or much less interesting to people of another culture. Besides, languages are very different, some language constructs can't be translated perfectly, and sometimes it can ruin impression of reader. For example, there is no analog of Present Perfect Tense in Russian, but on the other hand, English morphology (rules of words transforming) is poor compared to Russian. That's why short English sentences sometimes are translated into long and ugly Russian ones, and lots of nuance are lost then translating from Russian into English.
Unfortunately, some of untranslatable books are the greatest ones. Even the whole layers of can be literature are untranslatable (folk-lore, some kinds of poetry, etc.).
I might add that there ARE people OUTSIDE USA.
:)
And sometimes they might think that "The Greatest Book Of 20th Century" is written in their language. So it's pretty unfair to conduct a poll about the greatest book only between English-speaking people. And in the global worldwide poll some obscure Chinese book will probably win
I for one, think that Bulgakov's "The Master and Margaret" is the greatest Russian book of 20th century.
Lets see: $599 for one license of Linux and $101 for a remote control. Good deal! Sorry, can't resist :)
1. Buy a new HD 2. Crash it after 3 years 3. Replace it under warranty 4. Profit! I distinctly remember I've still got a circuit burner somewhere....
You really think so?
Just imagine, librarian robot running Windows:
"Please, give me a book about Linux".
Thump.
May be an ape might be a better librarian? ;)
I'm a software developer and my last project was to create OpenGroupware-to-Outlook connector. So I had to become very familiar with OpenGroupware internals.
Well, OpenGroupware SUCKS. Its web-interface is hard to use and slow - you need to make 3 clicks just to mark job as "Done", contacts management is even worse.
OpenGroupware server is hard to install and use - there is no good Debian/RPM packages, so you have to read tons of HOWTOs.
From developer perspective OpenGroupware is a nightmare. XMLRPC interface is poorly documented, unoptimized and not very usable. WebDav interface is even less documented. And now the best one: you can't access all functions of OpenGroupware from these interfaces.
I don't like to admit it, but Exchange and Outlook are far superior to OpenGroupware.
No. I'm not joking.
We pay about 8-10 cents per megabyte. Here's price list of our local ISP: Price list. It's in Russian, of course, but you can use something like Babelfish to translate it.
We pay 10 cents per megabyte of Internet traffic in Russia. So crappy bandwidth-burning sites is a reason to care.
Especially then you read them every day.
Why not just change the law to make media companies to accommodate to customers needs?
Liberty is a pretty good standard, it allows federated and distributed authoring instead of Microsoft's "only we know who you are" approach.
It's a shame that everything this alliance has produced up to date is just a pile of PDF specifications. Hope it will change soon.
Don't forget: methanol is not ethanol.
Methanol is _HIGHLY_ poisonous, _and_ it can be absorbed through skin.
Is it a result of westernization of Japanese culture?
Why would you need calculator on a _math_ test?
Calculators and any other artificial aids should be banned from exams. In fact, in Russia and many other countries calculators (as well as mobile phones) are banned on most exams.
Exam problems should be chosen so that to solve them you shouldn't need calculators/logarithmic rulers/etc.
For example, to compare sqrt(2) and 1.4 you don't need calculator. In fact, naive calculator-solution "1,4142135623730950488016887242097 is greater that 1.4" is incorrect. Correct solution is: "2>(1.4)^2".