I've got one too - and I have to say, I'm mainly very happy with it. I bought mine for the sound quality - it's vastly better than the iPod. (The iPod has "issues" with Classical music with a large dynamic range at high bitrates which make the sound utterly excruciating - yes, this is a bug in the design, and no, Apple don't give a Monkey's).
All I'd like to see added is a way to remote-control it via ethernet (i.e. to actually make it start playing!)
Another nice feature - it's easy to disassemble. I took the back off mine so as to write my name inside it (in case of loss/theft) - and it uses *proper* screws !
This is insane: there are far more of us geeks than there are bad lawyers. What does it take to either:
1)Change the law to ban patents (in their entirety) or 2)Counterattack the patent system in such a way that MAD (mutually assured destruction) is assured for anyone bringing patent litigation.
Thank you for that - that was interesting to read. Of course, for all its faults, X is still pretty good, in an old-fashioned kind of way. I wonder what the XOrg guys have to say about this? It would be nice to have such enhancements without "throwing the baby out with the bathwater". It would be important to be backward compatible to the extent of working on older hardware, and having some sort of compatibility library so that porting X-apps would be easy.
The other thing is, do we really want a 3d interface? I personally quite like having my windows flat!
Richard
Fair points: I admit that I never see these issues because I've never written any X-programs. From a user perspective, I can't complain. As for network performance, have you looked at nomachine? It looks pretty good, and I was impressed by the demo.
http://www.nomachine.com/testdrive.php
Yes, except that a novice should simply do things as per the distro. For example, my Aunt is having no difficulties with KDE/Mandrake. (I did install it for her, but after that, it will basically just work.)
Besides which, X doesn't have that many config files you'd need to look at. I'm a tinkerer, and even I have only ever tweaked XF86Config-4 (now XOrg-config),/etc/X11/fs/config, and/etc/X11/Xresources. Anyway, anything you replaced X with would need some config files!
Lastly, you make a serious logical error: it is precisely by "playing with the rope" that novice users get to grow up. So the "excess rope" is as necessary as it is dangerous.
Seriously, what's so bad about X that you want to see it replaced. OK, it isn't perfect, but it has only, in my view, minor flaws, and these are being worked on (The new XOrg seems to be making rapid progress). Some possible faults:
Drivers: good, but we still need the latest hardware support, and yesterday. nVidia's binary driver is actually very good, but a GPL one would be far nicer. Not really X's fault.
Configuration: XF86Config isn't especially pleasant. That said, most distros set this up anyway, so not really a problem.
Resources: X is "supposedly" a resource hog. Well, GPE (using X) runs fine on my little Zaurus. All the bloat is in modular extensions.
Copy/Paste: Some people find it confusing. It's not, when you remember that X has *2* buffers, one for select->middle click; the other for Ctrl-C->Ctrl-V.
Ugliness: Yes, if you are using some of the older widgets, I'd agree. But you have a choice! QT/GTK are both good looking.
On the good side, X is compatible with lots of things, it works well over a fast network, it's stable, and we have it here and now! XOrg are developing new features. (See here: http://www.freedesktop.org/XOrg/X11R68ScreenShots )
So if you really don't like 'x' about 'X', it's better to change it than to throw it out altogether!
Actually, Linux has *better* support for digital cameras than MS-Windows (you just plug in the camera, and it works, never mind the manufacturer's driver). You can use the fonts with Linux too (KDE has a nice graphical font installer). For newsletters, there is scribus, which is already pretty decent. As for home movies, it is possible to edit and burn the result to DVD (I just did it), although I will concede that cinelerra isn't very user friendly.
[6]: Backlight *LEDs* ?
Surely cold-cathode fluorescent lamp?
(Yes, the CCFL tubes do eventually die; worse, they tend to yellow with age. However, they are usually replaceable.)
Thanks - really useful!
I have 2 identical LCD monitors, one from DVI and the other VGA (due to Nvidia weirdness) - thanks to this trick, the analog one no longer looks radically worse!
BTW, you can do exactly the same with a png file, and reduce your bandwidth from
163902 bytes to 863 bytes (yes, really it's that good!)
Richard
You mean, they didn't notice he went mad, and he didn't start trying to kill them until the spoke about disconnecting him. Also, the entire point is that he was designed to seek truth, and be unable to lie, yet he was made to conceal the purpose of the mission. That was the cause of his "cognitive dissonance".
HAL didn't make an erroneous diagnostic, he lied about it. The reason was to break the line of communication with earth, which was the cause of his schizophrenia.
There is actually a "standard" rest frame for the Universe: it's the frame at rest with respect to the cosmic microwave background (i.e. a frame in which the CMB would have the same temperature in each direction). However, this doesn't contradict S.R.
The problem with publication (although it is a very good idea), is that it doesn't go far enough. A published idea is sufficient prior art to invalidate some patents, if we are lucky. But my idea of a "free patent foundation" would have to counter attack. For example,
1) The "FPF" obtains a patent on widgets of type "X". 2) The bad guys, (lets call them "M") invent a genuinely new widget, type "Y", for which no prior art may be found. 3) This new widget is also independently invented by a free software developer, "D". 4) M sues D over widget Y. 5) The FPF then threatens to sue M over their use of widget X, unless M promise never to sue over Y.
The advantage of this is that it would, in time, mean that nobody needs to take up defensive patents, and that nobody need fear being sued for patent infringement. Everyone(*) wins.
While my primary point is, as you said, to make software patents untenable in practice, I'd really like to see it apply to everything, whether software or hardware, and whether or not the invention is genuinely new. As a scientist, and an inventor, I think that all patents are harmful, both to society, and to the inventor concerned.
* There is an oft-quoted, hypothetical case of the lone genius inventor whose invention is stolen by a big company. He would hypothetically have been protected, had he patented it. Of course this is not true. The big company can almost always force him to cross-license it with them for free, by finding a patent of their own which he infringes. And he hasn't got the resources for a big legal fight. Thus, patents do not help the small inventor.
Remember that, to have a patent, the idea must be secret until the patent is applied for; only then is it publishable. So you can't put it on the web right away.
Also, the reason I *think* you'd need GPL is that, at least in software, if you offer your patent license free to BSD-licensed software, it can then be taken proprietary (even incorporated into MS software)
It sounds to me as though there is a missing link here. We have the programmers with ideas, and we have plenty of people who'd donate money. (I would for a start!). What is needed is a foundation to handle the legal part. (I'm slightly surprised that the FSF isn't doing it already - except that this would then "validate" the idea of software patents?).
The aim here would ideally be to totally undo the patent system: if we cannot have a wold without patents, we should extend the "MAD" ("mutually assured destruction") principle
so that no company ever dare to use a patent offensively. The organisation (let's call it the Free Patent Foundation for now) would be able to use its large resources to defend, initially just Free Software, and as it manages to gain a portfolio, could expand into other fields. I know plenty of academics who would be delighted to add their ideas, if the process could be simplified to under an hour's work per inventor.
Is it possible to use the patent system against itself, in the same way that the GPL counteracts the principle of copyright using copyright law itself?
For example, a public foundation dedicated to holding patents in the public interest. Anyone with an idea could submit it to them; they would then obtain a patent on it, and license it freely to the public, with the exception of companies who use their patent portfolios offensively.
Also, as an attack against software patents, would it be possible for a free developer to patent their own algorithms/widgets etc, and license it *only* for use in GPL'd software?
Is the above legally possible? There's an obvious problem in that most FOSS developers have neither the time nor the expertise, nor the money to apply for patents, but if an organisation could be formed to do that part, I have no doubt that the intellectual citizens of the world can out-invent the corporations obtaining patents.
[I am personally convinced that patents per se are a very bad thing, in any field of endeavour. They run contrary to the spirit of scientific endeavour, and they create a "tragedy of the commons" on a global scale. They also lead to monopolies. Pragamatically, one might make a special exemption for pharmaceutical patents, on the grounds of the huge investment required up front, but even then, such patents should be unenforceable in the 3rd world on grounds of humanitarian necessity.]
Fair point, but it's *really* easy to change to something you prefer. I think you've missed something: By default, each "Theme" such as Keramik comes as a package. But there are actually several parts, which you can mix and match.
For example, I love the Keramik widgets, but I hate the default Keramik colours. I personally like the following:
But you can make the entire thing look like MacOSX if you prefer. If you want "simple", then put everything back to KDE Classic.
Seriously, KDE is extremely configurable. I think the defaults are good, but if you don't like them, the LookNFeel control panel is there for your pleasure!
Good point, but I think this was a design choice. I doubt they'll want to re-do the entire design of the control centre, but your point about the "poorly marked" button is a good one. Have you filed a bug on it?
Actually, you *can* suck water higher than 10 metres. That's how it goes up tall trees - by a mixture of osmotic pressure from the roots, and transpirational pull from the leaves. The surface tension of the water is principally responsible: it prevents the "column" of water from tearing.
If you can get hold of one, you could use an IBM microdrive with a CF adapter. This behaves exactly like an IDE disk. The advantage compared to flash is that you won't wear it out with multiple writes eg swap.
I did that, just by installing Mozilla. In my newbie days, I wanted to install a more recent version, and told it to install as/usr/bin/mozilla. When it said that the directory wasn't empty, and "could it please delete the other files", I said yes. Oops! I ended up needing a reinstall.
I did actually replace it with an Antec as you suggest.
But what I really want to be able to buy is an expensive HDD. (How about an IDE/SATA disk with a failure rate of less than 0.5 % in 5 years?)
My guess is that the current failure rate is about 10% in 2 years.
Anyone would think that when a PSU dies, it should die silently. Not this one. 240V on all the auxilliary connectors - bang go 3 HDDs and a CDROM.
The motherboard is apparently OK (which is a relief, since it has 2GB RAM and 2 x 1GHz P3 on it, and you can't buy motherboards to replace it anymore).
It turns out that the MB has its PCI system damaged so that the PCI network card won't. (The card is detected, and configured, but sending packets - no way!). Fortunately, I obtained a USB-ethernet adaptor (sitetech - really cute), and it works.
This happened to me 2 weeks ago. I'm *really* glad that I had a full backup every night.
Moral: buy expensive components.
Corollary: where can one buy expensive components? I'd really like to pay a 50% premium for reliability, but none of the manufacturers seem to offer it!
I recently inherited a chat system (java clients, java server) where the server needed to run on Linux, (had been Windows). I was delighted that the server ran, without modifications, but was appalled to discover that under Linux, it required 6MB RAM on the server *per connected client* !! Apparently, this is a limitation of Java Multithreading under Linux. In the end, I re-wrote the server in PHP (php-cli, using sockets), and the result is vastly better!
Some other observations: Java isn't just "object-oriented" - it's "object-obsessive". Why can't we have, for example, a string primitive.
Also, in Swing, why is there no way to make the scrollbar in an applet's JScrollPane just stay scrolled down! It won't automatically scroll, when text is appended to the textarea, and there is no way to make it do so. There's an ugly hack, but even this isn't reliable: mainPane.getVerticalScrollBar().setValu e(mainPane.getVerticalScrollBar().getMaximum());
I should perhaps add that I'm not a Java expert, (I learned Java specifically for that task) and I may have missed an obvious trick, despite google.
Lastly, I don't care whether Java is GPL, but please, Sun, can you fix the licensing so that it will at least play nice with the Linux distros (eg so that Mandrake can release RPMS of it).
Slashdot Mirror
I've got one too - and I have to say, I'm mainly very happy with it. I bought mine for the sound quality - it's vastly better than the iPod.
(The iPod has "issues" with Classical music with a large dynamic range at high bitrates which make the sound utterly excruciating - yes, this is a bug in the design, and no, Apple don't give a Monkey's).
All I'd like to see added is a way to remote-control it via ethernet (i.e. to actually make it start playing!)
Another nice feature - it's easy to disassemble. I took the back off mine so as to write my name inside it (in case of loss/theft) - and it uses *proper* screws !
This is insane: there are far more of us geeks than there are bad lawyers. What does it take to either:
1)Change the law to ban patents (in their entirety)
or
2)Counterattack the patent system in such a way that MAD (mutually assured destruction) is assured for anyone bringing patent litigation.
Thank you for that - that was interesting to read. Of course, for all its faults, X is still pretty good, in an old-fashioned kind of way. I wonder what the XOrg guys have to say about this? It would be nice to have such enhancements without "throwing the baby out with the bathwater". It would be important to be backward compatible to the extent of working on older hardware, and having some sort of compatibility library so that porting X-apps would be easy. The other thing is, do we really want a 3d interface? I personally quite like having my windows flat! Richard
Fair points: I admit that I never see these issues because I've never written any X-programs. From a user perspective, I can't complain. As for network performance, have you looked at nomachine? It looks pretty good, and I was impressed by the demo. http://www.nomachine.com/testdrive.php
Yes, except that a novice should simply do things as per the distro. For example, my Aunt is having no difficulties with KDE/Mandrake. (I did install it for her, but after that, it will basically just work.) Besides which, X doesn't have that many config files you'd need to look at. I'm a tinkerer, and even I have only ever tweaked XF86Config-4 (now XOrg-config), /etc/X11/fs/config, and /etc/X11/Xresources. Anyway, anything you replaced X with would need some config files!
Lastly, you make a serious logical error: it is precisely by "playing with the rope" that novice users get to grow up. So the "excess rope" is as necessary as it is dangerous.
Seriously, what's so bad about X that you want to see it replaced. OK, it isn't perfect, but it has only, in my view, minor flaws, and these are being worked on (The new XOrg seems to be making rapid progress). Some possible faults:
Drivers: good, but we still need the latest hardware support, and yesterday. nVidia's binary driver is actually very good, but a GPL one would be far nicer. Not really X's fault.
Configuration: XF86Config isn't especially pleasant. That said, most distros set this up anyway, so not really a problem.
Resources: X is "supposedly" a resource hog. Well, GPE (using X) runs fine on my little Zaurus. All the bloat is in modular extensions.
Copy/Paste: Some people find it confusing. It's not, when you remember that X has *2* buffers, one for select->middle click; the other for Ctrl-C->Ctrl-V.
Ugliness: Yes, if you are using some of the older widgets, I'd agree. But you have a choice! QT/GTK are both good looking.
On the good side, X is compatible with lots of things, it works well over a fast network, it's stable, and we have it here and now! XOrg are developing new features.
(See here: http://www.freedesktop.org/XOrg/X11R68ScreenShots )
So if you really don't like 'x' about 'X', it's better to change it than to throw it out altogether!
Actually, Linux has *better* support for digital cameras than MS-Windows (you just plug in the camera, and it works, never mind the manufacturer's driver). You can use the fonts with Linux too (KDE has a nice graphical font installer). For newsletters, there is scribus, which is already pretty decent. As for home movies, it is possible to edit and burn the result to DVD (I just did it), although I will concede that cinelerra isn't very user friendly.
[6]: Backlight *LEDs* ? Surely cold-cathode fluorescent lamp? (Yes, the CCFL tubes do eventually die; worse, they tend to yellow with age. However, they are usually replaceable.)
Thanks - really useful! I have 2 identical LCD monitors, one from DVI and the other VGA (due to Nvidia weirdness) - thanks to this trick, the analog one no longer looks radically worse! BTW, you can do exactly the same with a png file, and reduce your bandwidth from 163902 bytes to 863 bytes (yes, really it's that good!) Richard
You mean, they didn't notice he went mad, and he didn't start trying to kill them until the spoke about disconnecting him. Also, the entire point is that he was designed to seek truth, and be unable to lie, yet he was made to conceal the purpose of the mission. That was the cause of his "cognitive dissonance".
HAL didn't make an erroneous diagnostic, he lied about it. The reason was to break the line of communication with earth, which was the cause of his schizophrenia.
There is actually a "standard" rest frame for the Universe: it's the frame at rest with respect to the cosmic microwave background (i.e. a frame in which the CMB would have the same temperature in each direction). However, this doesn't contradict S.R.
The problem with publication (although it is a very good idea), is that it doesn't go far enough. A published idea is sufficient prior art to invalidate some patents, if we are lucky. But my idea of a "free patent foundation" would have to counter attack. For example,
1) The "FPF" obtains a patent on widgets of type "X".
2) The bad guys, (lets call them "M") invent a genuinely new widget, type "Y", for which no prior art may be found.
3) This new widget is also independently invented by a free software developer, "D".
4) M sues D over widget Y.
5) The FPF then threatens to sue M over their use of widget X, unless M promise never to sue over Y.
The advantage of this is that it would, in time, mean that nobody needs to take up defensive patents, and that nobody need fear being sued for patent infringement. Everyone(*) wins.
While my primary point is, as you said, to make software patents untenable in practice, I'd really like to see it apply to everything, whether software or hardware, and whether or not the invention is genuinely new. As a scientist, and an inventor, I think that all patents are harmful, both to society, and to the inventor concerned.
* There is an oft-quoted, hypothetical case of the lone genius inventor whose invention is stolen by a big company. He would hypothetically have been protected, had he patented it. Of course this is not true. The big company can almost always force him to cross-license it with them for free, by finding a patent of their own which he infringes. And he hasn't got the resources for a big legal fight. Thus, patents do not help the small inventor.
Remember that, to have a patent, the idea must be secret until the patent is applied for; only then is it publishable. So you can't put it on the web right away. Also, the reason I *think* you'd need GPL is that, at least in software, if you offer your patent license free to BSD-licensed software, it can then be taken proprietary (even incorporated into MS software) It sounds to me as though there is a missing link here. We have the programmers with ideas, and we have plenty of people who'd donate money. (I would for a start!). What is needed is a foundation to handle the legal part. (I'm slightly surprised that the FSF isn't doing it already - except that this would then "validate" the idea of software patents?). The aim here would ideally be to totally undo the patent system: if we cannot have a wold without patents, we should extend the "MAD" ("mutually assured destruction") principle so that no company ever dare to use a patent offensively. The organisation (let's call it the Free Patent Foundation for now) would be able to use its large resources to defend, initially just Free Software, and as it manages to gain a portfolio, could expand into other fields. I know plenty of academics who would be delighted to add their ideas, if the process could be simplified to under an hour's work per inventor.
Is it possible to use the patent system against itself, in the same way that the GPL counteracts the principle of copyright using copyright law itself?
For example, a public foundation dedicated to holding patents in the public interest. Anyone with an idea could submit it to them; they would then obtain a patent on it, and license it freely to the public, with the exception of companies who use their patent portfolios offensively.
Also, as an attack against software patents, would it be possible for a free developer to patent their own algorithms/widgets etc, and license it *only* for use in GPL'd software?
Is the above legally possible? There's an obvious problem in that most FOSS developers have neither the time nor the expertise, nor the money to apply for patents, but if an organisation could be formed to do that part, I have no doubt that the intellectual citizens of the world can out-invent the corporations obtaining patents.
[I am personally convinced that patents per se are a very bad thing, in any field of endeavour. They run contrary to the spirit of scientific endeavour, and they create a "tragedy of the commons" on a global scale. They also lead to monopolies. Pragamatically, one might make a special exemption for pharmaceutical patents, on the grounds of the huge investment required up front, but even then, such patents should be unenforceable in the 3rd world on grounds of humanitarian necessity.]
Can you explain to me how this works? My bash scripting is fairly good, but I can't figure this out, and it's too dangerous to experiment! Thanks
Fair point, but it's *really* easy to change to something you prefer. I think you've missed something: By default, each "Theme" such as Keramik comes as a package. But there are actually several parts, which you can mix and match.
For example, I love the Keramik widgets, but I hate the default Keramik colours. I personally like the following:
Colours: Mandrake Galaxy (mainly blue)
Icons: KDE Classic
Style: Keramic
Window decorations: Keramik
Mouse: CrystalCursors
But you can make the entire thing look like MacOSX if you prefer. If you want "simple", then put everything back to KDE Classic.
Seriously, KDE is extremely configurable. I think the defaults are good, but if you don't like them, the LookNFeel control panel is there for your pleasure!
Good point, but I think this was a design choice. I doubt they'll want to re-do the entire design of the control centre, but your point about the "poorly marked" button is a good one. Have you filed a bug on it?
Actually, you *can* suck water higher than 10 metres. That's how it goes up tall trees - by a mixture of osmotic pressure from the roots, and transpirational pull from the leaves. The surface tension of the water is principally responsible: it prevents the "column" of water from tearing.
If you can get hold of one, you could use an IBM microdrive with a CF adapter. This behaves exactly like an IDE disk. The advantage compared to flash is that you won't wear it out with multiple writes eg swap.
I did that, just by installing Mozilla. In my newbie days, I wanted to install a more recent version, and told it to install as /usr/bin/mozilla. When it said that the directory wasn't empty, and "could it please delete the other files", I said yes. Oops! I ended up needing a reinstall.
I did actually replace it with an Antec as you suggest. But what I really want to be able to buy is an expensive HDD. (How about an IDE/SATA disk with a failure rate of less than 0.5 % in 5 years?) My guess is that the current failure rate is about 10% in 2 years.
Anyone would think that when a PSU dies, it should die silently. Not this one. 240V on all the auxilliary connectors - bang go 3 HDDs and a CDROM. The motherboard is apparently OK (which is a relief, since it has 2GB RAM and 2 x 1GHz P3 on it, and you can't buy motherboards to replace it anymore). It turns out that the MB has its PCI system damaged so that the PCI network card won't. (The card is detected, and configured, but sending packets - no way!). Fortunately, I obtained a USB-ethernet adaptor (sitetech - really cute), and it works. This happened to me 2 weeks ago. I'm *really* glad that I had a full backup every night. Moral: buy expensive components. Corollary: where can one buy expensive components? I'd really like to pay a 50% premium for reliability, but none of the manufacturers seem to offer it!
I recently inherited a chat system (java clients, java server) where the server needed to run on Linux, (had been Windows). I was delighted that the server ran, without modifications, but was appalled to discover that under Linux, it required 6MB RAM on the server *per connected client* !! Apparently, this is a limitation of Java Multithreading under Linux. In the end, I re-wrote the server in PHP (php-cli, using sockets), and the result is vastly better!
u e(mainPane.getVerticalScrollBar().getMaximum());
Some other observations: Java isn't just "object-oriented" - it's "object-obsessive". Why can't we have, for example, a string primitive.
Also, in Swing, why is there no way to make the scrollbar in an applet's JScrollPane just stay scrolled down! It won't automatically scroll, when text is appended to the textarea, and there is no way to make it do so. There's an ugly hack, but even this isn't reliable:
mainPane.getVerticalScrollBar().setVal
I should perhaps add that I'm not a Java expert, (I learned Java specifically for that task) and I may have missed an obvious trick, despite google.
Lastly, I don't care whether Java is GPL, but please, Sun, can you fix the licensing so that it will at least play nice with the Linux distros (eg so that Mandrake can release RPMS of it).