But a fairly small number of words contain the sequence "rt" or "tr". Absolute frequency is irrelevent - what matters is how ofter you use them in combination.
Note that the author claims that the manufacturers will only allow the release of the code that actually interfaces with the hardware in binary form. It looks like he's happy to provide modules for different architectures so it ought to be possible to run one of these things in an Alpha (for instance), but as-is it's i386 only. This is supposedly due to FCC regulations forbidding radio devices being entirely under user control, so manufacturers outside the US may be a better bet for a driver that can be integrated into the OSs.
Why would desktop/laptop users with full keyboards need to use this? It seems difficult for people with disabilities to use such a system.
You'd be surprised - large numbers of disabled users do use standard hardware with additional input devices. There's other potential advantages, too - we've had RSI sufferers tell us that it's preferable to a keyboard (though this is going to depend on mouse setup, I guess).
One other thing is that kids tend to find it significantly more entertaining than a typing tutor, and it's fairly good at encouraging correct spellings...
but with Dasher, one will always need to concentrate on the input... just by the nature of it
The transparency feature of the OS X version actually helps here. Make it sufficiently see-through and put it on top of the text that you're writing, and you ought to be able to flick your attention between the two without getting lost. It helps if you make the text window font a fairly thick bold one.
Hmm - it's seemed fairly stable here. Could you possibly let me know if you're doing anything specific before it crashes? Are there any graphical glitches?
If Linksys sells me a device that has a firmware image on it, they've distributed the image to *me.*
The firmware is downloadable from Linksys's website. That counts as them distributing it to anyone who downloads it.
However, if someone *using* that same device discovers a filesystem tucked away in there, containing Linux binaries, and makes a copy... well, Linksys hasn't distributed anything to *them.*
True. In that case, the onus is on the person doing the distribution to either provide the source, a written offer for the source or a copy of the offer that they received from Linksys. If they have none of these things, then they're not allowed to distribute it themselves.
Why should it make any difference what form the code is being distributed in? The GPL doesn't permit distribution of executables in any form without accompanying source or an offer of same.
And I could sell that computer to someone, with Linux installed, along with my own software, and still not be obligated to release any source.
No, you wouldn't. The GPL would require you to provide either the source code to GPLed code on your computer, or an offer of the source code. You could quibble over section 3c and whether it's commercial distribution if you're selling the computer rather than the software on it, and in that case you could get away with just passing on the offer that you received with your copy of Linux.
That's fine. If it's impossible for Linksys to release the source, they just have to stop distributing the code (and suffer potential lawsuits about copyright infringement and the like from anyone who holds the copyright on various parts of the kernel). If the GPL doesn't apply, plain copyright law does.
Doesn't matter. The firmware is downloadable regardless of whether you own the product or not - and the firmware includes a CRAMFS filesystem containing the kernel and busybox, so it's binary distribution of GPLed code. The offer of source has to be available to anyone who could obtain it from them. Read section 3 of the GPL.
I should clarify. Games are playable using the open source drivers. My Radeon 9000 (which is a cheap, low end card) gives me far, far better framerates with UT than the Voodoo 3 it replaced did. Do you get as many FPS as you do with the binary drivers? No. Do you get a high enough number of FPS to play 3D games at reasonable detail levels? Yes. Do you get the ability to run 3D applications without crippling the system? Yes. Can you do 3D visualisation on the desktop? Yes.
So, it's entirely acceptable for desktop use even if you're using 3D applications on the desktop. Note that "entirely acceptable" only defines the base level of performance. If your primary goal is gaming, then the binary drivers are probably a better bet - if not, then the open ones are fine.
I wouldn't say pitiful. Less good than the binary drivers, perhaps. You probably lose out somewhat for gaming, but for desktop use it's entirely acceptable.
The open source drivers do work with GLX. I'm using them quite happily now. They don't support the 9500 and up, but that's because the DRI team haven't had time rather than because ATI are witholding documentation. As far as graphics cards go, ATI support the Free Software community far better than Nvidia do.
This depends a lot on how much spam you get. Over the past year I've been averaging over 50 a day (I should really graph this to see which direction it's going in). Spamassassin generally stops somewhere around 95% of this, though it's lower at the moment. That's about 4 times the amount of legitimate mail I get. Without spamassassin, I'd be finding email almost entirely useless, and at that point I'd be happy to bounce (but not drop) some legitimate mail in order to reduce the spam load.
I learned Dvorak a few years ago. At the time I learned it on a physically QWERTY keyboard, which helped enforce proper finger positioning. I ended up being about 15WPM faster in Dvorak (85WPM overall), which certainly wasn't bad - however, I was also typing "correctly", while my QWERTY was an ad-hoc mess that I'd learned as I went along. Spending less time just forcing myself to learn QWERTY properly would probably have resulted in much the same speed increase.
Nowadays my desktop machines have IBM keyboards with removable keycaps, so they're all physically Dvorak - on the other hand, my laptop is both physically and logically QWERTY because other people want to use it occasionally. Switching takes a few seconds, but isn't a major problem.
The complaints are, almost entirely, about libraries. But there's already a robust mechanism for determining that a library dependency is satisfied - the SONAME defines its binary compatibility. So if stuff is breaking, it's because library authors are changing binary compatibility without changing the SONAME. How about we just get library authors to stop breaking stuff?
So run it under qemu. Sure, it's not going to be wonderfully fast, but most people with non-x86 hardware aren't going to be that fussed about gaming (otherwise they'd have an x86 already...)
I don't see why you think that Linux is relegated to the low to medium level servers. I have found that ported apps run 5 to 10 times faster on Linux (with PC hardware) that on Sun's best machines.
You've got the wrong idea about what a high end server is. Solaris scales better to 64 processors and above than Linux does, which is partly why Linux creams it at the lower end.
No, I don't claim that I'll have earned that (I'm still a couple of years away from the MA). It's a ridiculous anachronism. Any suggestion that it's because of the supposed higher quality of the degree is pretty bogus - employers can work out where you were from your CV anyway...
(I'm actually under the impression that the cash goes to the colleges rather than the university, but I wouldn't swear to it)
In various parts of the world, an engineer has a level of professional liability and expected ethics that the software industry refuses to accept. When that's no longer the case, I'll have a good deal more sympathy.
For what it's worth, I have a high level of academic and industry experience. I design and write code for a living, which makes me a professional programmer. I EARNED the right to put "BA" after my name[1], and I EARNED some cash. I did not EARN the right to call myself a Software Engineer, any more than I EARNED the right to call myself an MD, a PhD or any other title that may give the perception of competence.
[1] My university awards BAs for all non-Masters degree courses, even science ones
Slashdot Mirror
But a fairly small number of words contain the sequence "rt" or "tr". Absolute frequency is irrelevent - what matters is how ofter you use them in combination.
Note that the author claims that the manufacturers will only allow the release of the code that actually interfaces with the hardware in binary form. It looks like he's happy to provide modules for different architectures so it ought to be possible to run one of these things in an Alpha (for instance), but as-is it's i386 only. This is supposedly due to FCC regulations forbidding radio devices being entirely under user control, so manufacturers outside the US may be a better bet for a driver that can be integrated into the OSs.
Erk, sorry about that. It seems to work fairly well here, but that's rarely a guarantee in the software world. I'll get that looked into.
:) )
(Of course, now it's been mentioned, I'll probably be able to get it to do so at will
Why would desktop/laptop users with full keyboards need to use this? It seems difficult for people with disabilities to use such a system.
You'd be surprised - large numbers of disabled users do use standard hardware with additional input devices. There's other potential advantages, too - we've had RSI sufferers tell us that it's preferable to a keyboard (though this is going to depend on mouse setup, I guess).
One other thing is that kids tend to find it significantly more entertaining than a typing tutor, and it's fairly good at encouraging correct spellings...
but with Dasher, one will always need to concentrate on the input... just by the nature of it
The transparency feature of the OS X version actually helps here. Make it sufficiently see-through and put it on top of the text that you're writing, and you ought to be able to flick your attention between the two without getting lost. It helps if you make the text window font a fairly thick bold one.
if you are good with the alphabet and get the hang of the interface, you can probably get pretty good speed
The best we've measured is about 40WPM, so it's pretty close to a moderately fast non-touch typist
Hmm - it's seemed fairly stable here. Could you possibly let me know if you're doing anything specific before it crashes? Are there any graphical glitches?
To the same extent that it does on other platforms - that is, no. Only the left mouse button does anything.
If Linksys sells me a device that has a firmware image on it, they've distributed the image to *me.*
The firmware is downloadable from Linksys's website. That counts as them distributing it to anyone who downloads it.
However, if someone *using* that same device discovers a filesystem tucked away in there, containing Linux binaries, and makes a copy... well, Linksys hasn't distributed anything to *them.*
True. In that case, the onus is on the person doing the distribution to either provide the source, a written offer for the source or a copy of the offer that they received from Linksys. If they have none of these things, then they're not allowed to distribute it themselves.
Why should it make any difference what form the code is being distributed in? The GPL doesn't permit distribution of executables in any form without accompanying source or an offer of same.
And I could sell that computer to someone, with Linux installed, along with my own software, and still not be obligated to release any source.
No, you wouldn't. The GPL would require you to provide either the source code to GPLed code on your computer, or an offer of the source code. You could quibble over section 3c and whether it's commercial distribution if you're selling the computer rather than the software on it, and in that case you could get away with just passing on the offer that you received with your copy of Linux.
That's fine. If it's impossible for Linksys to release the source, they just have to stop distributing the code (and suffer potential lawsuits about copyright infringement and the like from anyone who holds the copyright on various parts of the kernel). If the GPL doesn't apply, plain copyright law does.
Doesn't matter. The firmware is downloadable regardless of whether you own the product or not - and the firmware includes a CRAMFS filesystem containing the kernel and busybox, so it's binary distribution of GPLed code. The offer of source has to be available to anyone who could obtain it from them. Read section 3 of the GPL.
I should clarify. Games are playable using the open source drivers. My Radeon 9000 (which is a cheap, low end card) gives me far, far better framerates with UT than the Voodoo 3 it replaced did. Do you get as many FPS as you do with the binary drivers? No. Do you get a high enough number of FPS to play 3D games at reasonable detail levels? Yes. Do you get the ability to run 3D applications without crippling the system? Yes. Can you do 3D visualisation on the desktop? Yes.
So, it's entirely acceptable for desktop use even if you're using 3D applications on the desktop. Note that "entirely acceptable" only defines the base level of performance. If your primary goal is gaming, then the binary drivers are probably a better bet - if not, then the open ones are fine.
I wouldn't say pitiful. Less good than the binary drivers, perhaps. You probably lose out somewhat for gaming, but for desktop use it's entirely acceptable.
The open source drivers do work with GLX. I'm using them quite happily now. They don't support the 9500 and up, but that's because the DRI team haven't had time rather than because ATI are witholding documentation. As far as graphics cards go, ATI support the Free Software community far better than Nvidia do.
There are open source 3D drivers for the Radeon. There are none for the Nvideas. Why do you think this is?
Darwin is not MacOS X. There's a distinct lack of Aqua sourcecode, for instance.
This depends a lot on how much spam you get. Over the past year I've been averaging over 50 a day (I should really graph this to see which direction it's going in). Spamassassin generally stops somewhere around 95% of this, though it's lower at the moment. That's about 4 times the amount of legitimate mail I get. Without spamassassin, I'd be finding email almost entirely useless, and at that point I'd be happy to bounce (but not drop) some legitimate mail in order to reduce the spam load.
I learned Dvorak a few years ago. At the time I learned it on a physically QWERTY keyboard, which helped enforce proper finger positioning. I ended up being about 15WPM faster in Dvorak (85WPM overall), which certainly wasn't bad - however, I was also typing "correctly", while my QWERTY was an ad-hoc mess that I'd learned as I went along. Spending less time just forcing myself to learn QWERTY properly would probably have resulted in much the same speed increase.
Nowadays my desktop machines have IBM keyboards with removable keycaps, so they're all physically Dvorak - on the other hand, my laptop is both physically and logically QWERTY because other people want to use it occasionally. Switching takes a few seconds, but isn't a major problem.
The complaints are, almost entirely, about libraries. But there's already a robust mechanism for determining that a library dependency is satisfied - the SONAME defines its binary compatibility. So if stuff is breaking, it's because library authors are changing binary compatibility without changing the SONAME. How about we just get library authors to stop breaking stuff?
Not to mention that I HAD to do the backup under Windows as OnStream's backup software doesn't support Linux( my main OS ).
1) Get drivers from http://www.linux1onstream.nl/
2) Use existing wide range of Linux backup software
3) Insert lame "profit" gag here
So run it under qemu. Sure, it's not going to be wonderfully fast, but most people with non-x86 hardware aren't going to be that fussed about gaming (otherwise they'd have an x86 already...)
I don't see why you think that Linux is relegated to the low to medium level servers. I have found that ported apps run 5 to 10 times faster on Linux (with PC hardware) that on Sun's best machines.
You've got the wrong idea about what a high end server is. Solaris scales better to 64 processors and above than Linux does, which is partly why Linux creams it at the lower end.
No, I don't claim that I'll have earned that (I'm still a couple of years away from the MA). It's a ridiculous anachronism. Any suggestion that it's because of the supposed higher quality of the degree is pretty bogus - employers can work out where you were from your CV anyway...
(I'm actually under the impression that the cash goes to the colleges rather than the university, but I wouldn't swear to it)
In various parts of the world, an engineer has a level of professional liability and expected ethics that the software industry refuses to accept. When that's no longer the case, I'll have a good deal more sympathy.
For what it's worth, I have a high level of academic and industry experience. I design and write code for a living, which makes me a professional programmer. I EARNED the right to put "BA" after my name[1], and I EARNED some cash. I did not EARN the right to call myself a Software Engineer, any more than I EARNED the right to call myself an MD, a PhD or any other title that may give the perception of competence.
[1] My university awards BAs for all non-Masters degree courses, even science ones