Well, for a start that JITing takes time itself. The end result might have C++ speed, but while the JVM is doing that optimisation work it's taking up cycles. Anyway, my point is that for me trying to actually use it, it's slow, people who don't think it's slow tend to tell me to upgrade. But Java is the only thing that doesn't give decent performance on this system.
Yes, it may well be swing (but it's not an old version, I'm using the 1.5.0.04 and it's still slow) but that's the only java gui you can use without losing the whole point of using java. So the effect is that java programs are slow.
Nope. Source under the GPL is defined as the preferred form for making modifications, if a tarball isn't what you work with then you can't just dump that.
I don't have all the answers, but the fact is that if it's a bot google shouldn't be charging them for it, and if the people getting the ads think they can show in a court of law that it was a bot they have every right to sue google.
There are, they're just not popular. And since the community is small, the discussions aren't as good, so less people like them. Slashdot's value isn't in its content but in its communities. You can get better actual stories (better written, and often more interesting) at osnews or ars technica or for the less directly technical ones plastic and K5 any day of the week.
My provider for the last year was among the cheapest available at the time and yet included a free static IP. If they can do that, why the hell would I want IPv6?
Huh? You can't explicitly delete objects as far as I can see, you have to get rid of all references and then leave your fate in the hands of the JVM to see when they get GCed. If it happens at an inconvenient time, what exactly can you do?
The fact that MS tried doesn't mean it isn't a problem. It is a problem, the language has no free implementation, I can't recall anyone ever trying to exert as much control over a whole language as Sun does with Java.
Why not? For a start, using a scripting language would have probably resulted in a factor of 10 reduction in the number of lines needed. And perl or python run on as many platforms, possibly more (java didn't work on beos last time I checked). Java may have open specs but it isn't open, wheras almost every other language has a truly free implementation available. And why do you claim linux owes its success to Java? There's no J in LAMP.
Comparisons to C++ are a red herring, languages have moved on since then. Python is both far faster to write (5-10x) and faster in actual use (since the performance-critical extensions are C/C++, wheras all but the very basics of the Java class library are done in Java.
I use it, and I'm happy using high level languages, I'm a big fan of python for example. Nevertheless, I still say java is slow. Why? Because I'm actually trying to use it for actual programs on sub-ghz machines.
Re:Java is practical for some applications
on
James Gosling on Java
·
· Score: 2, Insightful
Yes, it's practical for some applications. Just like Delphi, or Eiffel, or Haskell. It's a decent, somewhat specialised language that's worth knowing and using where appropriate. However, that's not the public perception of it, that's not the PHB perception of it. Java is an enormous triumph of marketing. People think things are better because they're java. People think arbitrary programs would be better if rewritten in java, where it only makes them worse. I will stop dissing java when people stop thinking it's the best thing since sliced bread, the new C++, the basic language everyone should know.
It's interpreted. It looks like a duck, smells like a duck, quacks like a duck, it's a duck. I don't care about your "jit", it requires an interpreter program, it's not a native binary, it has the slowdown you get from interpreter stuff. (Yes I know the benchmarks say it doesn't. Bite me.) It's interpreted.
There's no "considered to be GPL'd". The libraries are GPLed or not GPLed, and if they are GPLed then you have to use the GPL for any programs that use them, that's the entire point of them being GPL. They're there as a carrot to encourage you to use GPL for your programs. People who use the GPL for their libraries don't want you using them in a propriety work. It's not like your any worse off than if they'd never been written, and without the GPL they may well not have been.
The reason we need it is we want open source in our lifetimes. Like why the soviets supported revolutions in other countries. Ideologically, they knew that everyone would rise up and overthrow their capitalistic masters, but they wanted people to come to marxism now.
If they were trying to be obstructive someone would have sued them for it eventually. They may have opened up CVS to be good guys, but remember they didn't at first. I think part of the reason they did open it was they realised that legally, they didn't have a leg to stand on if they didn't.
And even before they did that, khtml devs were better off with the big patches they sent than without them. Not that much better off, but a bit better off.
And notice how the community got them to stop? Wheras if it was BSD they would have been on the right side of the law and gotten away with it, and other people could have done so too.
What if they don't worry about integrating their changes? If they make the fork once, and then maintain their own divergent version, like Apple is doing with Webcore. We now have what is effectively another open source web renderer - webcore has diverged sufficiently from khtml to count it separately. If khtml had been BSD, that would (probably, since they haven't displayed much concern for keeping the codebases compatible, so they wouldn't have much need to open it up) be one more closed source web renderer and one less open one.
What if the original maintainer has given up, and it needs a lot of porting to work on modern systems? With BSD software, where's the incentive to make the "new" version open source?
Is it me or does their firewall block them? So wouldn't it be simpler, rather than having them listening by default and then a firewall blocking them by default, to just have the services not listening by default? I've heard people say it's so it's easier to re-enable them - but how much harder is it to activate a service than to create a firewall rule permitting it? The only vulnerabilities the firewall will protect you from but turning the services off won't are TCP stack vulnerabilities and you'd have thought MS had got rid of those by now. For the rest, as soon as the firewall permits access to a service it's as vulnerable as if it were running on an unfirewalled machine.
You think it's reasonable for MS to sell a product that requires people to buy a separate router and not tell them? (Plenty of people can be infected in the time it takes to download SP2)
Can't speak for redhat, but I put a fresh slackware 9.1 in the DMZ of my router not so long ago, with no infection. (I think there's some proftpd vulnerability that could have caused a bit of trouble (IIRC readonly filesystem access) but not arbitrary code execution. No one hit me with it (unless they used another vulnerability to get rid of the log entry)in the 2 months before I turned off proftpd)
Isn't there an issue where windows brings up the network near the start of the boot process but doesn't turn on the firewall until near the end, so there's still a window for compromise?
Not to mention that downloading SP2 takes a lot more than 12 minutes for a dialup user.
It's sad how MS has got people to think that needing a $30 router is a normal thing for a PC and OS you bought. I got a router with my ADSL connection, first thing I did was put my linux box in the DMZ (I won't post the IP here, I think that would be a bit too much, but it's there with apache serving webpages and sshd taking remote logins and so on). Haven't had any problems.
You can see these by using your samba logs, if you have it running. I see failed login attempts at a rate of roughly 2 per second (put my PC in the DMZ of my router)
Since they coded it to one browser at a time, I don't think they're following any standards. My browser (konqueror) is pretty standards compliant, almost certainly more so than a few of the supported browsers (Safari and of course IE). I think it's them not bothering to write standards compliant code, not the browsers.
I had the opposite experience when switching to Linux, I really did. My (nvidia riva TNT2) video card gets a slightly higher resolution than in windows (1200x960 rather than 1152x864). I had to download drivers for it, but I have to do that in windows anyway, other than that just X -configure. Sound card (onboard via 82c686b) worked out of the box rather than having to use a borky manufacturer's installer that failed 4 times out of 5 and required 2 reboots between each attempt. Printer (epson stylus C20UX) worked fine (I just ran the kde add printer wizard). Modem (connexant one of some sort) worked although lost its call waiting functionality. I've since upgraded to ADSL using a router, the network card I use worked fine with again no driver download in linux, wheras to use it in windows I would have had to use another pc to download the driver (as it was I just used linux to download the windows driver installer). Linux is far more stable than the windows I switched from (crashes are a quarterly rather than weekly ocurrence). So, as far as I'm concerned, why on earth ANYONE would use windows for a home system is a mystery.
Slashdot Mirror
Yes, it may well be swing (but it's not an old version, I'm using the 1.5.0.04 and it's still slow) but that's the only java gui you can use without losing the whole point of using java. So the effect is that java programs are slow.
Nope. Source under the GPL is defined as the preferred form for making modifications, if a tarball isn't what you work with then you can't just dump that.
I don't have all the answers, but the fact is that if it's a bot google shouldn't be charging them for it, and if the people getting the ads think they can show in a court of law that it was a bot they have every right to sue google.
There are, they're just not popular. And since the community is small, the discussions aren't as good, so less people like them. Slashdot's value isn't in its content but in its communities. You can get better actual stories (better written, and often more interesting) at osnews or ars technica or for the less directly technical ones plastic and K5 any day of the week.
You're an idiot. Try something else for goodness sakes. You won't realise how crappy java is until you've tried using a decent language for things.
My provider for the last year was among the cheapest available at the time and yet included a free static IP. If they can do that, why the hell would I want IPv6?
The fact that MS tried doesn't mean it isn't a problem. It is a problem, the language has no free implementation, I can't recall anyone ever trying to exert as much control over a whole language as Sun does with Java.
Why not? For a start, using a scripting language would have probably resulted in a factor of 10 reduction in the number of lines needed. And perl or python run on as many platforms, possibly more (java didn't work on beos last time I checked). Java may have open specs but it isn't open, wheras almost every other language has a truly free implementation available. And why do you claim linux owes its success to Java? There's no J in LAMP.
Comparisons to C++ are a red herring, languages have moved on since then. Python is both far faster to write (5-10x) and faster in actual use (since the performance-critical extensions are C/C++, wheras all but the very basics of the Java class library are done in Java.
I use it, and I'm happy using high level languages, I'm a big fan of python for example. Nevertheless, I still say java is slow. Why? Because I'm actually trying to use it for actual programs on sub-ghz machines.
Yes, it's practical for some applications. Just like Delphi, or Eiffel, or Haskell. It's a decent, somewhat specialised language that's worth knowing and using where appropriate. However, that's not the public perception of it, that's not the PHB perception of it. Java is an enormous triumph of marketing. People think things are better because they're java. People think arbitrary programs would be better if rewritten in java, where it only makes them worse. I will stop dissing java when people stop thinking it's the best thing since sliced bread, the new C++, the basic language everyone should know.
It's interpreted. It looks like a duck, smells like a duck, quacks like a duck, it's a duck. I don't care about your "jit", it requires an interpreter program, it's not a native binary, it has the slowdown you get from interpreter stuff. (Yes I know the benchmarks say it doesn't. Bite me.) It's interpreted.
There's no "considered to be GPL'd". The libraries are GPLed or not GPLed, and if they are GPLed then you have to use the GPL for any programs that use them, that's the entire point of them being GPL. They're there as a carrot to encourage you to use GPL for your programs. People who use the GPL for their libraries don't want you using them in a propriety work. It's not like your any worse off than if they'd never been written, and without the GPL they may well not have been.
The reason we need it is we want open source in our lifetimes. Like why the soviets supported revolutions in other countries. Ideologically, they knew that everyone would rise up and overthrow their capitalistic masters, but they wanted people to come to marxism now.
And even before they did that, khtml devs were better off with the big patches they sent than without them. Not that much better off, but a bit better off.
And notice how the community got them to stop? Wheras if it was BSD they would have been on the right side of the law and gotten away with it, and other people could have done so too.
What if the original maintainer has given up, and it needs a lot of porting to work on modern systems? With BSD software, where's the incentive to make the "new" version open source?
Is it me or does their firewall block them? So wouldn't it be simpler, rather than having them listening by default and then a firewall blocking them by default, to just have the services not listening by default? I've heard people say it's so it's easier to re-enable them - but how much harder is it to activate a service than to create a firewall rule permitting it? The only vulnerabilities the firewall will protect you from but turning the services off won't are TCP stack vulnerabilities and you'd have thought MS had got rid of those by now. For the rest, as soon as the firewall permits access to a service it's as vulnerable as if it were running on an unfirewalled machine.
You think it's reasonable for MS to sell a product that requires people to buy a separate router and not tell them? (Plenty of people can be infected in the time it takes to download SP2)
Can't speak for redhat, but I put a fresh slackware 9.1 in the DMZ of my router not so long ago, with no infection. (I think there's some proftpd vulnerability that could have caused a bit of trouble (IIRC readonly filesystem access) but not arbitrary code execution. No one hit me with it (unless they used another vulnerability to get rid of the log entry)in the 2 months before I turned off proftpd)
Not to mention that downloading SP2 takes a lot more than 12 minutes for a dialup user.
It's sad how MS has got people to think that needing a $30 router is a normal thing for a PC and OS you bought. I got a router with my ADSL connection, first thing I did was put my linux box in the DMZ (I won't post the IP here, I think that would be a bit too much, but it's there with apache serving webpages and sshd taking remote logins and so on). Haven't had any problems.
You can see these by using your samba logs, if you have it running. I see failed login attempts at a rate of roughly 2 per second (put my PC in the DMZ of my router)
Since they coded it to one browser at a time, I don't think they're following any standards. My browser (konqueror) is pretty standards compliant, almost certainly more so than a few of the supported browsers (Safari and of course IE). I think it's them not bothering to write standards compliant code, not the browsers.
I had the opposite experience when switching to Linux, I really did. My (nvidia riva TNT2) video card gets a slightly higher resolution than in windows (1200x960 rather than 1152x864). I had to download drivers for it, but I have to do that in windows anyway, other than that just X -configure. Sound card (onboard via 82c686b) worked out of the box rather than having to use a borky manufacturer's installer that failed 4 times out of 5 and required 2 reboots between each attempt. Printer (epson stylus C20UX) worked fine (I just ran the kde add printer wizard). Modem (connexant one of some sort) worked although lost its call waiting functionality. I've since upgraded to ADSL using a router, the network card I use worked fine with again no driver download in linux, wheras to use it in windows I would have had to use another pc to download the driver (as it was I just used linux to download the windows driver installer). Linux is far more stable than the windows I switched from (crashes are a quarterly rather than weekly ocurrence). So, as far as I'm concerned, why on earth ANYONE would use windows for a home system is a mystery.
making a viewer that works in all browsers?