I think you'll find that Rivest's (not Shamir's) three ballot system is flawed too. In fact, it's less understandable than the article's system, and it's still susceptible to vote buying, as described by Appel: http://www.cs.princeton.edu/~appel/papers/Defeatin gThreeBallot.pdf.
I agree with you except for one point: 6 weeks is much too long. The correct time to wait is zero days. Here's why:
1. Vendors need an incentive to write bug free code. If vendors know that they can get away with sell-then-patch, they will do just that. But if bugs mean public exploits, angry users, and bad press, they will spend more money on security.
2. Black hats often have the security hole before you. So you're not doing the vendors much of a favor by giving them six weeks; you're just shielding them from bad press.
3. It is unfair to the user. Given #2, those six weeks are six more weeks that a user is susceptible to damage without any chance at turning off the service, reconfiguring it, putting up a firewall, changing passwords, and otherwise mitigating any potential problems. Remember, the bug is the vendor's fault, but it's the user that ends up hurting the most.
Full, immediate, public disclosure is only rational option.
The entire service manual is available online in various places (e.g. http://www.repairyourmac.com/macbook-pro.pdf). Why is Apple freaking out about a small excerpt in a forum? The simplest explanation is that they are embarrassed and are trying to cover it up.
That's a good point. However, it's probably good it turned out the way it did:
To borrow from your example, suppose I want to go to the Linux search site, com.search.linux in your example. Now I have to ask myself, is it http:/com.search.linux/, or is it http:/com.search/linux/? Did they make a subdomain, or is it a subfolder? You could have both, of course, but this would not be implemented reliably across the net.
The nice thing about having the order reversed is that there is a clear mental separation of the two hierarchies, domain and path.
Bullshit. It's quite clear what Slashdot is trying here. Just look at the posts! There aren't people trying to confirm or deny it, there is the presumption of guilt until proven innocent. This is absolutely obnoxious -- Slashdot readers have neither the resources nor the credibility to determine the validity of the claim. When your only source is the Wonkette, you shut up until the story is confirmed by a credible source.
But hey, if Slashdot can play that game, so can I. I heard that RobLimo eats babies. Now, I don't know if this is true, guys! I'm just the messenger! But it would be similar to an experience I had once, just so you know. Discuss.
No, it's not an executable. If you will read the article, it's considered a data file for the Terminal application. This is because it does not contain the usual #!/bin/sh header, so the system does not recognize it as a shell script, but Terminal still passes it to bash, which interprets it.
If I may take an unpopular viewpoint, I think this is precisely one of the reasons that some people do not trust research about today's popular issues (global warming, for example). As you and other posters have pointed out, if you want grants, you have to be in the trendy research of the day. What better way for your research to remain trendy than to market your results as important to the future of the world? (I am not claiming that any wrongdoing is going on, only that there is an incentive for wrongdoing.)
I hope this does not turn into a global warming debate; that is not the objective of this post. I just want to point out that, while perhaps it is necessary, this need to market science is hurting its reputation, correctly or incorrectly. When scientists become salesmen and politicians, they are no longer trustworthy. Unfortunately, I am afraid that as long as research funding is provided by the government, politics will be an inexorable part of science.
Your unjustified speculation is cute and everything, but Microsoft has offered multiple versions of Office for as long as I can remember. Some packages include programs such as Access that not everyone needs. They have even offered Word + Works Suite for low-end PCs. In none of these cases have they prevented you from reading or writing data files created with the more expensive suite (given that your suite the program in question, e.g. Access).
But hey, why have a reasonable discussion when you can just bash Microsoft for something it hasn't done?
The problem with provably correct software is not that the specifications are too complicated. It's that it's currently too difficult to prove desireable properties about code. The academics still have not produced any workable system for real-world code. However, it is still a hot area of research, with groups at Princeton, CMU, Yale, and other schools working on it.
The specification needn't be complex to reap many of the benefits of verified code; in some programs, you might just want a proof that the program cannot seg fault. In more widely used or more important programs, you might want a proof that the program cannot throw exceptions. In only the most vital code will you see very specific specifications, and this will be in areas where detailed specifications are already the norm (medical, military, etc.).
Yes, software bugs are a human problem. Maybe they're even a social problem. But we cannot just throw our hands into the air and claim that since we haven't solved these problems with technical means in 40 years, we never will. We need more time to develop the tools and ideas that will allow the computer to assist in finding bugs.
It's nice that Google has enabled s2s support, but I don't really see this as a big move. The lack of Google Talk support is (was) not holding back federated XMPP. The lack of users is. And frankly, there is simply no motivation for most users to switch to a Jabber-based system right now.
For Jabber to become the dominant IM service, it would require a critical mass of users. The best way for that to happen today would be AIM or MSN to support XMPP with s2s since it would not require the millions of existing users to change their behavior.
This is unlikely to occur. We have heard that Google is planning on somehow providing AIM support. However, based on the few articles out there that discuss this, it sounds like you will have to have an AIM account in addition to your XMPP account. This absolutely defeats the purpose of "combining" the networks, since you will still have two handles: your Jabber address, and your AIM screenname. I could already do that with Jabber transports and/or Gaim/Trillian multi-IM support.
Here's hoping that the news reports got it wrong and that AIM users will be able to communicate natively with XMPP users. If not, Google Talk will be a failed experiment.
Do you expect Microsoft to write two HTML rendering libraries, one for Outlook and one for Internet Explorer? HTML has become ubiquitous, so it makes sense that HTML rendering would be included with the operating system; now individual applications do not have to include (bundle?) their own rendering engines.
Do you have any references regarding undocumented hooks for Microsoft Office? I find that hard to believe, given that Wine can run Office.
Microsoft did make certain choices that have led to today's security-related mayhem. However, hindsight is 20/20; most of those choices were reasonable at the time they were made. Consider WMF, for example; today, it seems idiotic to allow data files to contain executable code, but when Windows 3.0 was released, it may have been a useful feature. Decisions such as these get buried in large code bases or made permanent by backwards compatibility requirements.
"For the first time in history, a Mac is 99% identical to a PC except for the BIOS and a few hardware/software tweaks."
This is true, if you count "an entirely different operating system and user experience" as a mere software tweak. People who subscribe to the Megahertz Myth may think Win86's and Mac86's can be compared directly, people who actually interact with a computer to determine its value will have other criteria for performance and value.
You have missed the point entirely. Take Dell computer A and Apple computer B with similar hardware specifications. We can install Windows on both computers and run benchmarks that demonstrate that both computers have comparable hardware. We can also look at trends to see which hardware is more reliable.
The problem for Apple is that they will no longer be able to make claims that Apple computers run faster; if trends continue, it will always be possible to buy a (hardware) comparable Dell for less money. Therefore, they will have to convince you that you should pay the Apple premium for the shiny case and the operating system (er, user experience), even though the $120 OS can technically run on any modern x86 machine with the proper video card.
Are you willing to pay hundreds of dollars extra to get Mac OS X? Do you think the average user will be?
"For a man that ran an entire's country's Internet"
Kieren, you do not seem to understand the difference between running a ccTLD and "running the Internet." You make this mistake several times within the first page of your article, referring to ownership of "Iraq's internet [sic]" and "Kazakhstan's internet [sic] ownership". I am not sure whether you are trying to exaggerate the impact of ccTLD control, or if you are simply blathering about concepts that you do not understand; in either case, I could not be bothered to read beyond the first page of your article.
If you want to be taken seriously in the future, please do a little research on Internet infrastructure before making such sweeping claims and touting conspiracy theories. At the very least, please use proper terminology to avoid embarrassing yourself.
This is a well-founded assumption. Much of Google's internal development is done in Python. Thus, it is important that Python development continue quickly and continue in a direction that benefits Google. They cannot risk that Guido find himself unable to afford to continue as Python project leader.
No, they are most certainly hiring Guido to continue Python development. It would be a disaster for Python, and thus for Google, if they diverted his talent toward some random Google project.
I don't understand why you use the term "given the power" in reference to Microsoft's ability to bundle various applications into their operating system platform. Windows was created by developers under Microsoft's employ. Microsoft retains distribution rights to the Windows operating system. Why would Microsoft need any additional rights or powers to include various functionalities such as web browsing or instant messenging?
I have chosen to download and use the Firefox web browser. It is more than adequate for my needs. I generally choose not to visit sites that require ActiveX plugins, although I make exceptions when the need arises. It's nice, however, to have a web browser to use to download Firefox. Bootstrapping problem, and all that.
By the way, you should try to get over your insistence on standards. AJAX, for example, is not a standard: the XMLHttpRequest object was created by Microsoft as part of Internet Explorer. It was later added, in modified form, by most of the other popular web browsers. Macromedia Flash is incredibly useful in certain problem domains; it is unlikely that your prized standards, Javascript and SVG, will ever catch up in quality and consistency. Nothing innovative ever comes out of standards organizations; they simply codify existing practice. Thus, as a developer, if I restrict myself to standards alone, I will miss out on new technologies.
And finally, I see no reason for Microsoft to work to bundle third-party applications. This is the role of the distributor of Windows, which is, for most users, Dell. And you should note that Dell does often bundle other third-party media players and similar applications with their PCs. I think the whole confusion is really due to the fact that nobody sells a modified Windows bundle without a PC; I suspect such an idea would not sell well, as the majority of users do not buy operating systems retail.
I have stated that I do not want to wait even a few seconds for an application to load. It's obnoxious to me and, I would venture, most other users. I don't understand why you are supporting the idea that latency is good. You may only open NetBeans once or twice a day, so it doesn't matter for you. But imagine if all your applications took a couple seconds to open. Even notepads or terminals or calculators or system monitors. It would change the way you used your computer. You would demand applications with more integrated features so that you would not have to suffer the short delays dozens of times throughout your day. Every time I have to way 3 seconds to open a calculator, I lose my train of thought. Better include a calculator in my IDE. No thanks, I like the idea of small, fast, quick loading applications.
Small, orthogonal tools. That's the UNIX philosophy, right?
Also, please note that the bulk of Microsoft Word's libraries are NOT preloaded with the operating system. Your statement's only grain of truth is that IE's rendering libraries are indeed preloaded (they are used within the Explorer shell). It is also, of course, true that the native Windows support libraries like GDI, etc. are "preloaded" i.e. already in memory, just like they are for almost any graphical system. However, this only saves on I/O time. The JVM startup seems to be slow for reasons other than I/O.
My distaste for Java the language has nothing to do with my distaste for the Java VM. They are completely separate concerns. One does not have to write JVM-targeted applications in Java. One does not have to compile Java to the JVM. So we really have four options: Java -> JVM, Non-Java -> JVM, Java -> native, Non-Java -> native. Because of the JVM's failings, I dislike anything -> JVM. Because of Java's failings, I dislike Java -> anything. Does this make sense?
Why should I have to wait 7 seconds, or 2.5 seconds, or 5 seconds, or whatever? That's absurd! Once I have made a decision to do something on my computer, I want that action to be carried out as quickly as possible. That's why we buy faster and faster computers. If you plan to open applications just once and leave them open all day, maybe slow startup times are fine, but that's not always desireable.
Besides, I have already identified Microsoft applications that have very short load times, so we know it's possible to create efficiently loading applications. Whether or not GNOME's programs load quickly is irrelevant (note, also, that there have been articles recently on improving GNOME's startup time and general performance). If GNOME performance doesn't improve too, we should abandon it for something better. Also, the fact that Java runs on low-memory systems is a red herring; what matters is the behavior of real applications under the real desktop JRE. And in those, we see slow startup time and high memory use.
Look, I have programmed server-side stuff in Java professionally. I am familiar with Java's strengths. However, Java's load-time performance cannot be excused just because you have a hard on for cross-platform binary compatibility. (I happen to think that Java is an awful language, too, but this is orthogonal to discussions of its runtime environment).
Even 2.5 seconds is an unreasonable amount of time for a simple application such as the notepad example, especially on a modern machine like yours. I should note that Microsoft's Wordpad opens in less than a second loading from disk. Subsequent (cached I/O) loads occur effectively instantaneously. Amazingly, Microsoft Word behaves identically. This is a Pentium M 1.4GhZ laptop with 768MB RAM. I believe these applications are quite a bit more powerful than the Notepad example, and they load much faster (well, presumably -- unfortunately, I do not have any Java GUI applications on this specific machine to provide comparison test results.)
There really is no excuse for Java's kind of bloat in a desktop application; a language with just as much abstraction, expressive power, and (source) portability as Java could be natively compiled without much trouble (see academic languages, such as Lisp, ML, etc. for examples). The result would be applications whose startup times are measured in milliseconds, not seconds. With proper design, Java can achieve decent response times *once the application is loaded*. However, load times are critical when you are running a multitasking GUI.
Unfortunately, nobody seems interested in native code compilers for high level languages these days. VMs and JITs are all the rage. They seem well suited to long-running server processes, but until the load time issues are solved, perhaps with a JIT code cache (doesn't.NET do this?), Java has no place on the desktop.
That's pretty cute, the use of "obstructing" in the summary. Usually when I hear the word obstructing it is in phrases like "obstructing justice." Obstructing is usually something the criminals do. The word has picked up a pretty negative connotation.
But here, we have HBO obstructing the downloading of their copyrighted material. HBO is obstructing copyright violation. Would you say that a lock obstructs breaking and entering? Or that self defense obstructs assault? Perhaps good server administration obstructs the stealing of private data. Of course you wouldn't say that. It sounds silly. So why is HBO obstructing downloads?
Slashdot Mirror
I think you'll find that Rivest's (not Shamir's) three ballot system is flawed too. In fact, it's less understandable than the article's system, and it's still susceptible to vote buying, as described by Appel: http://www.cs.princeton.edu/~appel/papers/Defeatin gThreeBallot.pdf.
I'll post without reading the article, reading the FAQ, or viewing the sample video; you can even verify it by my stupid comment.
RTFA.
I agree with you except for one point: 6 weeks is much too long. The correct time to wait is zero days. Here's why:
1. Vendors need an incentive to write bug free code. If vendors know that they can get away with sell-then-patch, they will do just that. But if bugs mean public exploits, angry users, and bad press, they will spend more money on security.
2. Black hats often have the security hole before you. So you're not doing the vendors much of a favor by giving them six weeks; you're just shielding them from bad press.
3. It is unfair to the user. Given #2, those six weeks are six more weeks that a user is susceptible to damage without any chance at turning off the service, reconfiguring it, putting up a firewall, changing passwords, and otherwise mitigating any potential problems. Remember, the bug is the vendor's fault, but it's the user that ends up hurting the most.
Full, immediate, public disclosure is only rational option.
Infected is right. These iPods are Apple's smallpox blankets to Microsoft's American Indians.
The entire service manual is available online in various places (e.g. http://www.repairyourmac.com/macbook-pro.pdf). Why is Apple freaking out about a small excerpt in a forum? The simplest explanation is that they are embarrassed and are trying to cover it up.
That's a good point. However, it's probably good it turned out the way it did:
To borrow from your example, suppose I want to go to the Linux search site, com.search.linux in your example. Now I have to ask myself, is it http:/com.search.linux/, or is it http:/com.search/linux/? Did they make a subdomain, or is it a subfolder? You could have both, of course, but this would not be implemented reliably across the net.
The nice thing about having the order reversed is that there is a clear mental separation of the two hierarchies, domain and path.
Wow, the Daily Kos parroted the same unsubstantiated rumor as Slashdot. Shocking!
Wake me up when we hear a report from someone who can tell a fact from a hole in the ground.
Bullshit. It's quite clear what Slashdot is trying here. Just look at the posts! There aren't people trying to confirm or deny it, there is the presumption of guilt until proven innocent. This is absolutely obnoxious -- Slashdot readers have neither the resources nor the credibility to determine the validity of the claim. When your only source is the Wonkette, you shut up until the story is confirmed by a credible source.
But hey, if Slashdot can play that game, so can I. I heard that RobLimo eats babies. Now, I don't know if this is true, guys! I'm just the messenger! But it would be similar to an experience I had once, just so you know. Discuss.
No, it's not an executable. If you will read the article, it's considered a data file for the Terminal application. This is because it does not contain the usual #!/bin/sh header, so the system does not recognize it as a shell script, but Terminal still passes it to bash, which interprets it.
If I may take an unpopular viewpoint, I think this is precisely one of the reasons that some people do not trust research about today's popular issues (global warming, for example). As you and other posters have pointed out, if you want grants, you have to be in the trendy research of the day. What better way for your research to remain trendy than to market your results as important to the future of the world? (I am not claiming that any wrongdoing is going on, only that there is an incentive for wrongdoing.)
I hope this does not turn into a global warming debate; that is not the objective of this post. I just want to point out that, while perhaps it is necessary, this need to market science is hurting its reputation, correctly or incorrectly. When scientists become salesmen and politicians, they are no longer trustworthy. Unfortunately, I am afraid that as long as research funding is provided by the government, politics will be an inexorable part of science.
Your unjustified speculation is cute and everything, but Microsoft has offered multiple versions of Office for as long as I can remember. Some packages include programs such as Access that not everyone needs. They have even offered Word + Works Suite for low-end PCs. In none of these cases have they prevented you from reading or writing data files created with the more expensive suite (given that your suite the program in question, e.g. Access).
But hey, why have a reasonable discussion when you can just bash Microsoft for something it hasn't done?
The problem with provably correct software is not that the specifications are too complicated. It's that it's currently too difficult to prove desireable properties about code. The academics still have not produced any workable system for real-world code. However, it is still a hot area of research, with groups at Princeton, CMU, Yale, and other schools working on it.
The specification needn't be complex to reap many of the benefits of verified code; in some programs, you might just want a proof that the program cannot seg fault. In more widely used or more important programs, you might want a proof that the program cannot throw exceptions. In only the most vital code will you see very specific specifications, and this will be in areas where detailed specifications are already the norm (medical, military, etc.).
Yes, software bugs are a human problem. Maybe they're even a social problem. But we cannot just throw our hands into the air and claim that since we haven't solved these problems with technical means in 40 years, we never will. We need more time to develop the tools and ideas that will allow the computer to assist in finding bugs.
It's nice that Google has enabled s2s support, but I don't really see this as a big move. The lack of Google Talk support is (was) not holding back federated XMPP. The lack of users is. And frankly, there is simply no motivation for most users to switch to a Jabber-based system right now.
For Jabber to become the dominant IM service, it would require a critical mass of users. The best way for that to happen today would be AIM or MSN to support XMPP with s2s since it would not require the millions of existing users to change their behavior.
This is unlikely to occur. We have heard that Google is planning on somehow providing AIM support. However, based on the few articles out there that discuss this, it sounds like you will have to have an AIM account in addition to your XMPP account. This absolutely defeats the purpose of "combining" the networks, since you will still have two handles: your Jabber address, and your AIM screenname. I could already do that with Jabber transports and/or Gaim/Trillian multi-IM support.
Here's hoping that the news reports got it wrong and that AIM users will be able to communicate natively with XMPP users. If not, Google Talk will be a failed experiment.
Do you expect Microsoft to write two HTML rendering libraries, one for Outlook and one for Internet Explorer? HTML has become ubiquitous, so it makes sense that HTML rendering would be included with the operating system; now individual applications do not have to include (bundle?) their own rendering engines.
Do you have any references regarding undocumented hooks for Microsoft Office? I find that hard to believe, given that Wine can run Office.
Microsoft did make certain choices that have led to today's security-related mayhem. However, hindsight is 20/20; most of those choices were reasonable at the time they were made. Consider WMF, for example; today, it seems idiotic to allow data files to contain executable code, but when Windows 3.0 was released, it may have been a useful feature. Decisions such as these get buried in large code bases or made permanent by backwards compatibility requirements.
"For the first time in history, a Mac is 99% identical to a PC except for the BIOS and a few hardware/software tweaks."
This is true, if you count "an entirely different operating system and user experience" as a mere software tweak. People who subscribe to the Megahertz Myth may think Win86's and Mac86's can be compared directly, people who actually interact with a computer to determine its value will have other criteria for performance and value.
You have missed the point entirely. Take Dell computer A and Apple computer B with similar hardware specifications. We can install Windows on both computers and run benchmarks that demonstrate that both computers have comparable hardware. We can also look at trends to see which hardware is more reliable.
The problem for Apple is that they will no longer be able to make claims that Apple computers run faster; if trends continue, it will always be possible to buy a (hardware) comparable Dell for less money. Therefore, they will have to convince you that you should pay the Apple premium for the shiny case and the operating system (er, user experience), even though the $120 OS can technically run on any modern x86 machine with the proper video card.
Are you willing to pay hundreds of dollars extra to get Mac OS X? Do you think the average user will be?
"For a man that ran an entire's country's Internet"
Kieren, you do not seem to understand the difference between running a ccTLD and "running the Internet." You make this mistake several times within the first page of your article, referring to ownership of "Iraq's internet [sic]" and "Kazakhstan's internet [sic] ownership". I am not sure whether you are trying to exaggerate the impact of ccTLD control, or if you are simply blathering about concepts that you do not understand; in either case, I could not be bothered to read beyond the first page of your article.
If you want to be taken seriously in the future, please do a little research on Internet infrastructure before making such sweeping claims and touting conspiracy theories. At the very least, please use proper terminology to avoid embarrassing yourself.
This is a well-founded assumption. Much of Google's internal development is done in Python. Thus, it is important that Python development continue quickly and continue in a direction that benefits Google. They cannot risk that Guido find himself unable to afford to continue as Python project leader.
No, they are most certainly hiring Guido to continue Python development. It would be a disaster for Python, and thus for Google, if they diverted his talent toward some random Google project.
I don't understand why you use the term "given the power" in reference to Microsoft's ability to bundle various applications into their operating system platform. Windows was created by developers under Microsoft's employ. Microsoft retains distribution rights to the Windows operating system. Why would Microsoft need any additional rights or powers to include various functionalities such as web browsing or instant messenging?
I have chosen to download and use the Firefox web browser. It is more than adequate for my needs. I generally choose not to visit sites that require ActiveX plugins, although I make exceptions when the need arises. It's nice, however, to have a web browser to use to download Firefox. Bootstrapping problem, and all that.
By the way, you should try to get over your insistence on standards. AJAX, for example, is not a standard: the XMLHttpRequest object was created by Microsoft as part of Internet Explorer. It was later added, in modified form, by most of the other popular web browsers. Macromedia Flash is incredibly useful in certain problem domains; it is unlikely that your prized standards, Javascript and SVG, will ever catch up in quality and consistency. Nothing innovative ever comes out of standards organizations; they simply codify existing practice. Thus, as a developer, if I restrict myself to standards alone, I will miss out on new technologies.
And finally, I see no reason for Microsoft to work to bundle third-party applications. This is the role of the distributor of Windows, which is, for most users, Dell. And you should note that Dell does often bundle other third-party media players and similar applications with their PCs. I think the whole confusion is really due to the fact that nobody sells a modified Windows bundle without a PC; I suspect such an idea would not sell well, as the majority of users do not buy operating systems retail.
I have stated that I do not want to wait even a few seconds for an application to load. It's obnoxious to me and, I would venture, most other users. I don't understand why you are supporting the idea that latency is good. You may only open NetBeans once or twice a day, so it doesn't matter for you. But imagine if all your applications took a couple seconds to open. Even notepads or terminals or calculators or system monitors. It would change the way you used your computer. You would demand applications with more integrated features so that you would not have to suffer the short delays dozens of times throughout your day. Every time I have to way 3 seconds to open a calculator, I lose my train of thought. Better include a calculator in my IDE. No thanks, I like the idea of small, fast, quick loading applications.
Small, orthogonal tools. That's the UNIX philosophy, right?
Also, please note that the bulk of Microsoft Word's libraries are NOT preloaded with the operating system. Your statement's only grain of truth is that IE's rendering libraries are indeed preloaded (they are used within the Explorer shell). It is also, of course, true that the native Windows support libraries like GDI, etc. are "preloaded" i.e. already in memory, just like they are for almost any graphical system. However, this only saves on I/O time. The JVM startup seems to be slow for reasons other than I/O.
My distaste for Java the language has nothing to do with my distaste for the Java VM. They are completely separate concerns. One does not have to write JVM-targeted applications in Java. One does not have to compile Java to the JVM. So we really have four options: Java -> JVM, Non-Java -> JVM, Java -> native, Non-Java -> native. Because of the JVM's failings, I dislike anything -> JVM. Because of Java's failings, I dislike Java -> anything. Does this make sense?
Why should I have to wait 7 seconds, or 2.5 seconds, or 5 seconds, or whatever? That's absurd! Once I have made a decision to do something on my computer, I want that action to be carried out as quickly as possible. That's why we buy faster and faster computers. If you plan to open applications just once and leave them open all day, maybe slow startup times are fine, but that's not always desireable.
Besides, I have already identified Microsoft applications that have very short load times, so we know it's possible to create efficiently loading applications. Whether or not GNOME's programs load quickly is irrelevant (note, also, that there have been articles recently on improving GNOME's startup time and general performance). If GNOME performance doesn't improve too, we should abandon it for something better. Also, the fact that Java runs on low-memory systems is a red herring; what matters is the behavior of real applications under the real desktop JRE. And in those, we see slow startup time and high memory use.
Look, I have programmed server-side stuff in Java professionally. I am familiar with Java's strengths. However, Java's load-time performance cannot be excused just because you have a hard on for cross-platform binary compatibility. (I happen to think that Java is an awful language, too, but this is orthogonal to discussions of its runtime environment).
Even 2.5 seconds is an unreasonable amount of time for a simple application such as the notepad example, especially on a modern machine like yours. I should note that Microsoft's Wordpad opens in less than a second loading from disk. Subsequent (cached I/O) loads occur effectively instantaneously. Amazingly, Microsoft Word behaves identically. This is a Pentium M 1.4GhZ laptop with 768MB RAM. I believe these applications are quite a bit more powerful than the Notepad example, and they load much faster (well, presumably -- unfortunately, I do not have any Java GUI applications on this specific machine to provide comparison test results.)
.NET do this?), Java has no place on the desktop.
There really is no excuse for Java's kind of bloat in a desktop application; a language with just as much abstraction, expressive power, and (source) portability as Java could be natively compiled without much trouble (see academic languages, such as Lisp, ML, etc. for examples). The result would be applications whose startup times are measured in milliseconds, not seconds. With proper design, Java can achieve decent response times *once the application is loaded*. However, load times are critical when you are running a multitasking GUI.
Unfortunately, nobody seems interested in native code compilers for high level languages these days. VMs and JITs are all the rage. They seem well suited to long-running server processes, but until the load time issues are solved, perhaps with a JIT code cache (doesn't
That's pretty cute, the use of "obstructing" in the summary. Usually when I hear the word obstructing it is in phrases like "obstructing justice." Obstructing is usually something the criminals do. The word has picked up a pretty negative connotation.
But here, we have HBO obstructing the downloading of their copyrighted material. HBO is obstructing copyright violation. Would you say that a lock obstructs breaking and entering? Or that self defense obstructs assault? Perhaps good server administration obstructs the stealing of private data. Of course you wouldn't say that. It sounds silly. So why is HBO obstructing downloads?
Uh, Nessus is still out there. They cannot take the old versions away.
"herassment"
What does her ass have anything to do with it?
Hmmm, are you lying, trolling, joking, or just don't feel like providing links? I mean, jeez, what the hell are you talking about?