Mono is nearing release 1.0 and is a very attractive platform for developers.
In what way? Mono will always be playing catch-up with.Net, with significant sections (enterprise server functions) not fully implemented. It will be a sub-standard version of a platform that Microsoft intends to use to keep developers on Windows.
Java is fully implemented on Linux and supported there by companies such as IBM, Sun and HP. All enterprise features are present and up-to-date, and provided by companies who work together to ensure that developers can write portable systems.
Releasing Java open source 3 years ago would have screwed Microsoft hard, but now I'm not so sure.
It already has. Java is the main development language for server-side applications..Net is playing catch-up, and not doing that well.
Legacy app support is absolutely crucial and right now Mono can do something that not even Linux can do - support Microsoft-based legacy apps with a minimum of changes..Net has nothing to do with legacy app support. Just look at all the complaints from the VB6 developers at the considerable porting effort required to move to VB.Net
Re:Benchmark: Mono vs. Java vs. C++/C
on
Mono Beta 2 Released
·
· Score: 2, Insightful
Funny, I've never seen Java perform anywhere close to the speed of C and C++ and that benchmark is quite believable.
This would definitely have been the case 4 or 5 years ago, before hotspot compiling. Java was then typically 10%-20% of the speed of compiled C++, but there is good reason why Java should be pretty much the same speed as C++: All good Java VMs now produce not just native code at runtime, but highly optimised native code resulting from profiling analysis. Java 1.5 even caches the native code to disk, so you are running a pre-compiled binary the next time you start an application - there is no interpretation phase.
I use Java for numerical work and have found that some math routines run up to 10% faster (on the IBM VM) than the same code compiled with optimisation using GCC.
Re:Benchmark: Mono vs. Java vs. C++/C
on
Mono Beta 2 Released
·
· Score: 2, Interesting
Firstly, if you want to test Java for speed, use IBM's VM - it can be very much faster than Sun's. Secondly, I just don't believe these results. Given equivalent code Java is usually quoted as being within 10%-20% of C++ speed (and IBM's VM is faster). I am highly suspicious that exactly equivalent code was not used. Was an Integer class implemented in C++?
Also, what about memory use? by default, the 1.4 VM allocates 64MB max heap. The test is allocating a million objects. I don't know how Integer is implemented, but if it involves even a few tens of bytes, there is going to be a lot of garbage collecting going on.
I think you are missing my point, or maybe I'm misunderstanding you - I'm not talking about GUI components, I'm talking about the entire J2EE side (this is what I mean by server-side). As far as I can tell, almost all Java development is J2EE, and eclipse (unlike netbeans) has no built-in and free tools for J2EE deployment and debugging. Sure, there are some great tools, but they aren't part of the basic open source system.
I'd be really interested to know what developers like you use eclipse for.
I must do, mustn't I, as I disagree with the mainstream Slashdot view? Actually, no.
Except for Sun Studio, I believe you can freely download the other items you mentioned. I don't know why you need a whole Linux distribution for it anyway.
But that is the major complaint from so many Linux geeks.... Java is not pre-installed! You don't want to have to download and install systems on hundreds of corporate desktops.
As for a while Linux distribution... why not? Its free.
It seems to me that the primary benefactor of the JDS is Sun.
Wow! A company sells something and its designed to benefit them!
They want to collect site licensing fees.
Shame on them.
I'm not saying it's a bad idea, it's just not that amazing once you look past the marketing.
That is not the point being discussed. The point was that it was 'nothing to do with Java'.
There was recently an article on slashdot about how awful this desktop was.
Yeah.... what a surprise.
No surprise, SUN is not a front end company.
They pioneered the Unix Workstation, and innovative GUI systems like NeWS and the OpenWindows Window Manager. They helped port Gnome to Solaris. You must have some strange definition of 'not a front end company' that does not match our Earth usage.
The OSS community could kill the SUN desktop before it gets to that point by making GPL remote network admin software.
I thought OSS was about choice, not killing competition.
If the OSS community fails to do this, do it well, and do it in a timely way they may get their wish of linux desktops in the mainstream......but it will be the SUN desktop.
So the sun desktop is inferior, and awful and unusable.... but if we don't watch out it will take over the mainstream?
There are no viable third party solutions that can be drop in replacements for what business need. Businesses, want a real solution, they want it NOW and they want it from the same vendor.
That's not what the majority of corporate desktops are for. Been there, seen them, supported them, know what I'm talking about. Its web browsing, e-mail client, word processing, spreadsheets, and a front-end to corporate applications such as SAP.
As for same vendor - nonsense - most corporations are mixed vendor environments, and most corporate app servers are not Microsoft.
1. Because they are neither open source or Microsoft. 2. Because they haven't totally given up all proprietary UNIX, abandoned decades of experience in OS and kernel design, and immediately adopted Linux for everything. 3. Because they haven't sacked McNeally and appointed ESR as lifetime president. 4. Because they aren't allowing everyone to fiddle with Java (it's not like Java has been a raving success, after all: How can it possibly be called 'successful' until its a Debian package?) 5. Because you can't overclock their processors and post pictures of neat cooling systems to Slashdot.
What is holding JDS back isn't necessarily Linux but the fact that SUN has done nothing about dire shortage of third party software vendors for the desktop.
Eh? its Linux and there is an up-to-date Java VM included. It runs KDE and Gnome apps. What shortage of third party software is there?
I'm puzzled by the huge uptake of eclipse. As a basic IDE its truly awesome, but it has no features to handle user interfaces, be they client-side or web. Netbeans may lack a lot of eclipse features (for now), but it has a very powerful Swing GUI designer and lots of vital J2EE features such as servlet and JSP debugging pre-installed. As far as I know you can only get those features for eclipse with commercial or shareware plug-ins.
Eclipse is a great free GUI for software that doesn't either have a user interface or run on a server.
Who says it has nothing to do with Java? It has an up-to-date pre-installed Java VM, Java tools for adding extensions to Star Office, it now includes Sun Studio for Java development. The JVM is linked with Mozilla, so user can run serious Applets (making Java Desktop ideal for corporate environments and intranets).
Its a corporate linux desktop which includes substantial Java tools to allow integration with, and development for, server-side J2EE installations.
I can see it happening on a couple of legacy systems spread around a company, but to have an entire company on it?
Why not? Windows 98 is only 6 years old. Something that young should not be defined as legacy. For almost all uses it is functionally equivalent to XP, so why bother to upgrade?
AFAIK we've never actually seen a star collapse and a black hole appear... that wouldn't even be proof but it certainly would be the least of what we'd need to see before claiming that IS how they are formed. Rather, it's just our best guess of what could create this thing we call black hole.
Its pretty much the only way. Most of what is 'out there' is gas. When areas of this gas collapses it gets dense and hot. When it gets dense enough and hot enough it starts to undergo nuclear fusion, forming a star. When the fusion runs out after millions or billions of years the gas can finish collapsing, and if there is enough it will collapse into a black hole (well, some of it - a lot will experience shock waves and neutrinos and explode outwards). Stars and black holes are just different stages in what you get if you let a large amount of hydrogen gas collapse.
So you see damn near anything is theoretically possible. And if it's theoretically possible it can be proven via computer simulation, we control the initial variables.
Heh. Have you ever done computer simulation? Its nothing like that at all. Most things are impossible to simulate, and you get all kinds of emergent behaviour. In simulations we (at least try to) set constraints and just see what happens. If we are lucky we get repeatable behaviour and can statistically analyse the results.
this is all just to outline the real question. would i every reach the event horizon, befor the univers came to an end?
Yes. You are only frozen at the event horizon from the point of view of someone distant from the black hole. Also, you are only frozen there for a short while in practice. The light by which they could see you would be red shifted by gravity until pretty soon you are invisible.
From your point of view, you fall in in finite time.
Remember, relativity does not guarantee synchronicity. A black hole produces the ultimate split in synchroncity: From the point of view of an outsider, you don't fall in. From your point of view, you do. The paradox is resolved because even for the outsider, you become invisible and undetectable except as a mass increase in the black hole.
If you look at the behaviour of some stars you can see that they are orbiting something massive. If its really massive and you can't see it, its probably a black hole. If an object is massive enough, and its not keeping itself spread out because of heat (like a star that has run out of fuel) it will inevitably collapse into a black hole. You can measure the size of some objects by how rapidly they flicker. If things change in a matter of hours, then the effect can be no larger that that number of light-hours across. If you also know the mass of the object (by how fast things orbit round it) you can calculate a minimum density. In many cases this works out at black hole density.
Basic: blackhole% Fortran: BLACKHOL Pascal: BlackHole C: black_hole Java: blackHole Hungarian Notation: lpzBlackHole (a long pointer which terminates in null)
Look for publications by Peter Sheehan of the Milwaukee Public Museum.
From an interview: "After the three summers of studying the K-T data Sheehan reached one of the most important conclusions of the mass extinction debate: communities were changing little throughout the Hell Creek formation. Therefore, the extinction of the dinosaurs was sudden, as Alvarez and Smit suggested before him."
There is a LOT of fossil evidence showing that Chicxulub did NOT wipe out the dinos. There is also fossil evidence that dinos were already in decline BEFORE Chicxulub hit.
Well, actually, no, there isn't. A recent review of the fossils has resulted in a review of the dates, and suggests that there was no decline before the meteor. The dinosaurs were thriving just before it hit.
Slashdot Mirror
Mono is nearing release 1.0 and is a very attractive platform for developers.
.Net, with significant sections (enterprise server functions) not fully implemented. It will be a sub-standard version of a platform that Microsoft intends to use to keep developers on Windows.
.Net is playing catch-up, and not doing that well.
In what way? Mono will always be playing catch-up with
Java is fully implemented on Linux and supported there by companies such as IBM, Sun and HP. All enterprise features are present and up-to-date, and provided by companies who work together to ensure that developers can write portable systems.
Releasing Java open source 3 years ago would have screwed Microsoft hard, but now I'm not so sure.
It already has. Java is the main development language for server-side applications.
Surely it should be
RPN Calculators Future of the
Legacy app support is absolutely crucial and right now Mono can do something that not even Linux can do - support Microsoft-based legacy apps with a minimum of changes. .Net has nothing to do with legacy app support. Just look at all the complaints from the VB6 developers at the considerable porting effort required to move to VB.Net
Funny, I've never seen Java perform anywhere close to the speed of C and C++ and that benchmark is quite believable.
This would definitely have been the case 4 or 5 years ago, before hotspot compiling. Java was then typically 10%-20% of the speed of compiled C++, but there is good reason why Java should be pretty much the same speed as C++: All good Java VMs now produce not just native code at runtime, but highly optimised native code resulting from profiling analysis. Java 1.5 even caches the native code to disk, so you are running a pre-compiled binary the next time you start an application - there is no interpretation phase.
I use Java for numerical work and have found that some math routines run up to 10% faster (on the IBM VM) than the same code compiled with optimisation using GCC.
Firstly, if you want to test Java for speed, use IBM's VM - it can be very much faster than Sun's.
Secondly, I just don't believe these results. Given equivalent code Java is usually quoted as being within 10%-20% of C++ speed (and IBM's VM is faster). I am highly suspicious that exactly equivalent code was not used. Was an Integer class implemented in C++?
Also, what about memory use? by default, the 1.4 VM allocates 64MB max heap. The test is allocating a million objects. I don't know how Integer is implemented, but if it involves even a few tens of bytes, there is going to be a lot of garbage collecting going on.
I think you are missing my point, or maybe I'm misunderstanding you - I'm not talking about GUI components, I'm talking about the entire J2EE side (this is what I mean by server-side). As far as I can tell, almost all Java development is J2EE, and eclipse (unlike netbeans) has no built-in and free tools for J2EE deployment and debugging. Sure, there are some great tools, but they aren't part of the basic open source system.
I'd be really interested to know what developers like you use eclipse for.
So, SuSE has Java security tools built in? network management built in? Java pre-integrated with Open Office?
SuSE includes Sun Studio, and Java 1.4 pre-installed? Woo! Show me the URL.
Do you work for Sun or something?
I must do, mustn't I, as I disagree with the mainstream Slashdot view? Actually, no.
Except for Sun Studio, I believe you can freely download the other items you mentioned. I don't know why you need a whole Linux distribution for it anyway.
But that is the major complaint from so many Linux geeks.... Java is not pre-installed! You don't want to have to download and install systems on hundreds of corporate desktops.
As for a while Linux distribution... why not? Its free.
It seems to me that the primary benefactor of the JDS is Sun.
Wow! A company sells something and its designed to benefit them!
They want to collect site licensing fees.
Shame on them.
I'm not saying it's a bad idea, it's just not that amazing once you look past the marketing.
That is not the point being discussed. The point was that it was 'nothing to do with Java'.
There was recently an article on slashdot about how awful this desktop was.
Yeah.... what a surprise.
No surprise, SUN is not a front end company.
They pioneered the Unix Workstation, and innovative GUI systems like NeWS and the OpenWindows Window Manager. They helped port Gnome to Solaris. You must have some strange definition of 'not a front end company' that does not match our Earth usage.
The OSS community could kill the SUN desktop before it gets to that point by making GPL remote network admin software.
I thought OSS was about choice, not killing competition.
If the OSS community fails to do this, do it well, and do it in a timely way they may get their wish of linux desktops in the mainstream......but it will be the SUN desktop.
So the sun desktop is inferior, and awful and unusable....
but if we don't watch out it will take over the mainstream?
Make your mind up!
Wake up sunshine.
<Jar-Jar>How Wude</Jar-Jar>
There are no viable third party solutions that can be drop in replacements for what business need. Businesses, want a real solution, they want it NOW and they want it from the same vendor.
That's not what the majority of corporate desktops are for. Been there, seen them, supported them, know what I'm talking about. Its web browsing, e-mail client, word processing, spreadsheets, and a front-end to corporate applications such as SAP.
As for same vendor - nonsense - most corporations are mixed vendor environments, and most corporate app servers are not Microsoft.
i dont know why so many people talk down on sun.
1. Because they are neither open source or Microsoft.
2. Because they haven't totally given up all proprietary UNIX, abandoned decades of experience in OS and kernel design, and immediately adopted Linux for everything.
3. Because they haven't sacked McNeally and appointed ESR as lifetime president.
4. Because they aren't allowing everyone to fiddle with Java (it's not like Java has been a raving success, after all: How can it possibly be called 'successful' until its a Debian package?)
5. Because you can't overclock their processors and post pictures of neat cooling systems to Slashdot.
What is holding JDS back isn't necessarily Linux but the fact that SUN has done nothing about dire shortage of third party software vendors for the desktop.
Eh? its Linux and there is an up-to-date Java VM included. It runs KDE and Gnome apps. What shortage of third party software is there?
I'm puzzled by the huge uptake of eclipse. As a basic IDE its truly awesome, but it has no features to handle user interfaces, be they client-side or web. Netbeans may lack a lot of eclipse features (for now), but it has a very powerful Swing GUI designer and lots of vital J2EE features such as servlet and JSP debugging pre-installed. As far as I know you can only get those features for eclipse with commercial or shareware plug-ins.
Eclipse is a great free GUI for software that doesn't either have a user interface or run on a server.
Who says it has nothing to do with Java? It has an up-to-date pre-installed Java VM, Java tools for adding extensions to Star Office, it now includes Sun Studio for Java development. The JVM is linked with Mozilla, so user can run serious Applets (making Java Desktop ideal for corporate environments and intranets).
Its a corporate linux desktop which includes substantial Java tools to allow integration with, and development for, server-side J2EE installations.
I can see it happening on a couple of legacy systems spread around a company, but to have an entire company on it?
Why not? Windows 98 is only 6 years old. Something that young should not be defined as legacy. For almost all uses it is functionally equivalent to XP, so why bother to upgrade?
AFAIK we've never actually seen a star collapse and a black hole appear... that wouldn't even be proof but it certainly would be the least of what we'd need to see before claiming that IS how they are formed. Rather, it's just our best guess of what could create this thing we call black hole.
Its pretty much the only way. Most of what is 'out there' is gas. When areas of this gas collapses it gets dense and hot. When it gets dense enough and hot enough it starts to undergo nuclear fusion, forming a star. When the fusion runs out after millions or billions of years the gas can finish collapsing, and if there is enough it will collapse into a black hole (well, some of it - a lot will experience shock waves and neutrinos and explode outwards). Stars and black holes are just different stages in what you get if you let a large amount of hydrogen gas collapse.
So you see damn near anything is theoretically possible. And if it's theoretically possible it can be proven via computer simulation, we control the initial variables.
Heh. Have you ever done computer simulation? Its nothing like that at all. Most things are impossible to simulate, and you get all kinds of emergent behaviour. In simulations we (at least try to) set constraints and just see what happens. If we are lucky we get repeatable behaviour and can statistically analyse the results.
this is all just to outline
the real question. would i every reach the event
horizon, befor the univers came to an end?
Yes. You are only frozen at the event horizon from the point of view of someone distant from the black hole. Also, you are only frozen there for a short while in practice. The light by which they could see you would be red shifted by gravity until pretty soon you are invisible.
From your point of view, you fall in in finite time.
Remember, relativity does not guarantee synchronicity. A black hole produces the ultimate split in synchroncity: From the point of view of an outsider, you don't fall in. From your point of view, you do. The paradox is resolved because even for the outsider, you become invisible and undetectable except as a mass increase in the black hole.
If you look at the behaviour of some stars you can see that they are orbiting something massive. If its really massive and you can't see it, its probably a black hole. If an object is massive enough, and its not keeping itself spread out because of heat (like a star that has run out of fuel) it will inevitably collapse into a black hole. You can measure the size of some objects by how rapidly they flicker. If things change in a matter of hours, then the effect can be no larger that that number of light-hours across. If you also know the mass of the object (by how fast things orbit round it) you can calculate a minimum density. In many cases this works out at black hole density.
Depends on the language:
Basic: blackhole%
Fortran: BLACKHOL
Pascal: BlackHole
C: black_hole
Java: blackHole
Hungarian Notation: lpzBlackHole (a long pointer which terminates in null)
Look for publications by Peter Sheehan of the Milwaukee Public Museum.
From an interview:
"After the three summers of studying the K-T data Sheehan reached one of the most important conclusions of the mass extinction debate: communities were changing little throughout the Hell Creek formation. Therefore, the extinction of the dinosaurs was sudden, as Alvarez and Smit suggested before him."
We don't have any small dinosaurs running around today
Yes we do, millions of them. They are therapod dinosaurs, otherwise known as birds.
Just hypothetically, what could possibly disprove evolution without involving time travel?
Finding a human fossil 100 million years old.
There is a LOT of fossil evidence showing that Chicxulub did NOT wipe out the dinos. There is also fossil evidence that dinos were already in decline BEFORE Chicxulub hit.
Well, actually, no, there isn't. A recent review of the fossils has resulted in a review of the dates, and suggests that there was no decline before the meteor. The dinosaurs were thriving just before it hit.
It seems only a few years ago I was using Psion Organisers with 16 kilobyte memory packs.
OK, so it was nearly 20 years ago.
In 20 years time, if technology continues at the same pace, what will we be doing with petabyte drives?