I'm not really intending to "defend" linux, I think OSX deserves props for finally bringing unix with a nice front-end on it to the general populace (something all the linux vendors seemed to fail to do for years). But as another pointed out OSX is actually just the latest version of NextStep.
Anyhow, many of your points are due to two things -- 1. OSX like all mac os's only runs on a limited set of hardware, this makes things TONS easier, especially in installation. 2. Linux have NEVER been widely shipped pre-installed. All those things you worry about you laptop not doing and "Top-end hardware support" would pretty much work with a preinstalled, properly system integrated Linux.
I just had an idea. Some linux vendor should start selling boxes with simpler linux distribution pre-installed, with KDE, all the extra crap stripped out with their own picked PC hardware. At least this would give people something to compare fairly against Windows and Mac. Breaking into the PC market by installing on existing machines is just nearly suicide due to the insane variety of hardware.
Of course the strength of the PC is the commodity hardware keeping prices down and avoiding dependency on single vendors.
Actually if you can use nuclear power to produce the 150J but the 100 J of oil can be used in portable devices (gasoline in cars) whereas nuclear cannot it may still be economically viable. (though addmittedly at that point it will be incredibly expensive.)
This is like declaring that batteries are useless because they take more power to charge than they provide when used.
The economic viability argument is more robust, if it costs more money to supply oil than the market demand will bear (due to the availability of other energy sourcess) then oil companies simply won't exist.
1. (as you say) Lindows is required to give the beta testers the code if they ask for it.
2. Also the GPL says that those beta testers have the right to redistribute that code to anyone they please. i.e. Lindows cannot put additional restrictions on it.
That's it. And unless the beta testers are employees, it IS (non-internal) distribution.
He made exactly the point that things are different in the case of patents, they still encourage public disclosure by not granting protection to unpatented devices. That covers pretty much all but the particle accelerators, in which case the community (physicists) is so oriented toward public disclosure that you can always find out how something was done, it's essential in order to judge the validity of the result.
This is like saying we need to "fix" language because it's possible to write a bad essay!
Programming languages would be a hinderance if they *prevented* you from doing it right. They don't, it's just the first try at something is usually not the optimal. Or you were in a hurry, or some people are just sloppy.
oh really? show me where the raw data is. All I've ever found are web apps that think they know how you want to interface with the date (2-3hr blocks and keyword search, title and other info presented in weird places) and are impossible to integrate with anything. The reason the software available is a long way from a Tivo like service is that the data is not generally available in the form such an app would need.
If you know a place the raw data is available please let me know!
It seems obvious to me that there are two reasons that Tivo hasn't been "embraced"
1. It's hard to understand the advantages over a VCR. This doesn't mean there aren't any or that their impossible for normal people to use, it's just a hard sell. Nearly everyone already has multiple VCRs.
2. THE BIG ONE -- The absence of a removable media (like tape on a VCR) is a BIG minus. VCR's are essentially used for 3 things, time-shifting shows, "copying" shows/movies (i.e. recording them to keep for a while or to transport), and for playing rented tapes. Tivo does the first but due to the lack of a removable media it can't do the other two. A Tivo owner can't record something and then take the recording to his friend's house and watch it. It's locked in the Tivo.
If Tivo would simply be brave enough to also include a CDR/W drive that would make this thing a 100% feature-for-feature VCR replacement, wide adoption would be much less painful. A combo Tivo/DVD player is what is needed to actually *replace* a VCR in full functionality, but they don't sell these.
If a company wants to use your XML-library to built a big app (far beyond your code), why should they be disallowed to do so?
Because you wrote it, and you have just as much right to forbid them from selling your product as they do to forbid you from stealing a copy of theirs.
Is it
not the ultimate compliment to you and an excellent way to spread good code/standards?
Ah, but with a free (lowercase) license you have relinquished control of the source to the point where you can't even be sure that they haven't completely screwed it up. You can't see what they've done to your code.
Similarly, why should application developers be barred from creating commercial apps for Linux or KDE?
This is just false.
What's preventing them? They can compile AGAINST the libraries no problem, they just can't distribute modified versions of the code or it's products. But they're writing a new app so this does nothing to them.
The only reason seems a hatred of commercial developers, protecting your code from abuse and hidden extensions can be achieved with LGPL.
No the reason is to prevent a proprietary vendor from taking my code, adding 1% added value and selling it back to me. With GPL I get that 1% in exchange for giving them the 99%.
The GPL simply allows programmers to protect their work in exactly the way the previous poster indicated.
The entire reason to have a strong license like the GPL is for situations like the one the Ximian is in. The are up against a very powerful, well-funded, entrenched proprietary vendor. I predict the downfall will happen like this:
As soon as anyone starts to use Mono instead of MS's stuff, MS will just steal Mono's code, put it in their proprietary stuff and then tell their customers there is no reason to change because they just appropriated the feature that the customer was seeking. This is not so important in itself, but after this happens a few times Mono will likely be orphaned by the developer community. People don't like having their donated code sold for profit.
What is sad is that the "other companies" that want this license change do not understand that it will be the dowfall of the project.
It sounds like Ximian has decided that vendor support is worth sacrificing the viabiblity of a large section of their project. That's their decision, I just have the impression that it might be short-sighted.
Interesting how all the highly moderated comments are anti-GPL. This is slashdot, somethings wrong.
2 things, I think that was his point. If you've watched TV lately you know they're already biased.
Also it's worse than you think, the Audio Home Recording Act explicitly makes it legal to copy CDs for "non-commercial" purposes. that's even broader than just fair use.
I firmly believe that anyone who creates an "ultimate" linux system with only one display must have something wrong with them. What about a great display? Like 3 pci based video cards that play nice together and some whopper LCD panels?
And I should note that there are two more words that they should think about after that: multi PROCESSOR!
And my opinion on this system -- nice decoration, but PC boxes belong in the CLOSET!!
I'd say it must not have been a planned move on the part of MS simply because as of when I post this comment, the pro-MS astroturf is not being posted on slashdot. It must just be some random employee stuffing the box. On most MS stories there are plenty posts on MS's side, here there are only pot-shots, a few notes about how bad online polls are and some genuinely concerned opinions.
The purpose of the AHRA was to make sure that consumers (i.e. "the people") were legally allowed to simply copy music (i.e. non-commercially). In exchange the music companies got the blank media "royalty".
I think that we have all been decieved by the RIAA into believing that copying digital audio is illegal, it's not! Read the AHRA!
No action may be brought under this title (Copyrights) alleging infringement
of copyright based on the manufacture, importation, or distribution
of a digital audio recording device, a digital audio recording
medium, an analog recording device, or an analog recording medium,
or based on the noncommercial use by a consumer of such a device or
medium for making digital musical recordings or analog musical
recordings.
The point of Linux is to be incredibly adaptable and configurable. This means that your ideally designed user interface world is at odds with what Linux is.
That said, you have a valid comment, but in different terms. There probably should be A Linux which has a designed, consistent user interface. One distribution or a set of them should have this goal in mind, and some do. An OS is not just a user interface, and as such one should not design an OS for a single consistent user interface. An OS should efficiently execute system calls and manage computer resources. The user interface is a separate issue.
A LOT of the confusion on this subject is due to lack of the distiction between "a Linux" and "Linux".
On a side note, I think the biggest difference between linux and windows (now that we have KDE and GNOME for UIs) is device support. The device support culture is totally different, in MS drivers are written by the manufacturer, in Linux, they're written by the community. There has been, and will continue to be, a major culture clash here. When I get the newest wiz-bang device from shmo inc., in windows they can give me some driver that will probably hose my system in the end (a lot of Win instability comes from this) but it will work, and I'll blame MS for the instability. In Linux, new wiz-bang devices are generally not supported in the main distributions because it takes time and public review of code to get it there. What is a company to do. This problem has not been solved yet, and that is THE difference between windows and linux for the general user.
(this all comes down to the issue of how do you create a computer system that doesn't need an administrator to maintain it. Win PCs "freed" us from the admins, and MS pushed this hard, but now all those freed people are realising that tha admin might actually be needed (witness current MS security))
To further your point. Heiroglyphics is actually a syllabary system, the pictures actually have relatively little to do with the meaning, they just tell you what syllables to say, which converts into language. (this is why it took us so long to figure out how to translate them, we were trying to figure out what the pictures meant.)
You're thinking more like ideograms, which is how chinese and related languages are written, and for the reasons you say, even those have been simplified and adapted into something complicated and impossible to understand for the uninitiated.
Re:You may want the data more than you think
on
Comparing the DVRs?
·
· Score: 2, Interesting
even $20 a year would be fine with me for this data. I'm thinking some site I can login to with a name/password and they give me a file in a nice database format that I can use. No their box in my house. The point is I want the all the data (prevents profiling) with no strings attached other than standard copyright.
my question is where is this place now? Is anybody doing this?
I know about tvguide.com or tv.yahoo.com, but as you say, and another poster points out, they wouldn't like it (need ads) and the format probably changes constantly. This is probably what I would resort to if I started writing this thing now.
(you know, it's a sad state of affairs when I have essentially no expectation of somebody relying on "standard copyright")
Re:You may want the data more than you think
on
Comparing the DVRs?
·
· Score: 1
You do realize that you have just succesfully argued that the box is irrelevant and all that matters is having a piece of sophisticated software for managing recording schedules.
I've actually been thinking about this (I already record all my TV on my own computer). The big hurdle seems to be getting the listings. Do you know where one could get widespread, accurate, TV listings free for download? It sure seems like there should be a place, but I have a suspicion that the *only* way to get this kind of information is through a pay service, that of course not only wants your money but want *you*. (i.e. wants to make it as hard as possible for you to use a competitor's system in the future).
Imagine a commoditized Tivo where the recorder/storage hardware, the scheduling software and the listing service were all separate and competition could occur on each individually. We already have a Free software solution for the first of these, but as I said above I think the nature of the third makes the second sticky.
from comments mentioned in kernel traffic this week I'd say it's because ALSA is so easily compiled outside the kernel tree. I'm using it with a debian stock kernel now, no kernel compile necessary. Just download the alsa packages, compile and modprobe.
A better question is probably why does redhat muck with their kernels' sound so much that the above isn't true for them? (is it?) They should just use ALSA.
That and the alsa guys seem commited to having a STABLE system before it goes into the kernel.
There are other posts here that probably explain things better but I'll be brief:
1. You don't know what you're doing and neither does your friend.
2. If you took a 2 year old windows install with no updates and tried to install the newest patch, I'm pretty sure you would have similar problems. It's not a very smart thing to do.
I have brought a RH6.2 box up to date. It doesn't require force or nodeps, it just requires installing all the updates. It IS really hard to do because of the RPM version change, but that is redhat's fault.
Sounds like you actually ran into exactly the problem with adults vs. children. This lady assumed that she knew how computers worked, and that nothing had changed in 20 years (a pretty self-centered assumption I think). She was probably never able to convince herself that she should just humbly give up her preconceptions and start from scratch -- which a child does automatically.
The comment about operating systems being wasteful is pretty revealing. She had succeeded in categorizing "new" computers as somehow lesser than the computers she was used to. This kind of attitude makes it almost impossible to learn new things, because now she is resistant to abandoning her old assumptions because she has convinced herself that they are somehow better.
I don't know about your particular case, but in my experience DVD *decoders* really suck. I've never seen a set-top box that I felt was up to snuff. Try xine or oms or anything that uses libmpeg2 for decoding on the computer. I think the hardware manufacurers skimped to make their players cheap and that you can see it in quality. Until I started fiddling with libmpeg2 I didn't know that the accuracy of your iDCT matters, and until recently the one published by intel was not up to spec.
The irony is probably that the better the encoding is (i.e. the better use it makes of the bitrate) probably the more damage a shoddy decoder will do.
The worship Linus faction won't like it, but I think Linus is just out of his league. He is not an academic, but he is trying to comment on an academic topic.
There is a difference between design and implementation. Linus is confusing the two, and thereby simply confusing himself and anyone "arguing" with him. (I think that it's impossible to argue with someone who maintains a self-inconsistent stance.) Let me try to clarify the distinction
Design: The process of determining requirements, classifying them, then planning out a system which satisfies these requirements to a desireable degree.
Implementation: The process of building a system which performs certain actions.
In the case of implementing a design, the actions performed somehow lead to satisfaction of the design requirements.
Now that that is clear, there is one more point of to clarify:
Subsystem: Part of a system, which typically performs a certain action, i.e. partially implements.
Critical point: One can DESIGN a subsystem. This can often lead to confusion on the difference between design and implementation.
One MUST realize that one of the most critical parts of design is the creation of the requirements! Improving implementation NEVER helps if you're solving the wrong problem!!!!
One well-known way of designing is to generate a bunch of ideas which might satisfy the requirements, you then go through, rank the ideas and then eliminate the ones that don't satisfy the requirements as well. This is a "natural selection" approach to design.
With all of that said, Linus' problem is that he is an implementor. The design is DONE! It's unix. All he has to do is implement. Now, in implementation subsystems will be designed and implemented on and on. And better designs for subsystems will be created all the time. The kernel itself is a subsystem, and as such isn't really part of the design, but part of the implementation. He has designed the kernel - Modularity, timeslicing, portability requirements, even making it open-source was a design decision. Oh and designs can evolve, i.e. be amended when more knowledge is gleaned. Example: One of the principle differences between the two VM's is that they set out to accomplish different goals. Many of the requirements are the same (they're both VM's after all) but some are different. By changing to a new VM, one is simply changing the design of a sub-system because it provides a better implementation of its functionality than the previous design. So the design of the kernel hasn't changed, part of its implementation has, it needs a VM subsystem just like before. Changing the kernel design would have been something along the lines of removing the VM, say because memory is so cheap that it is simply unecessary complication.
And for those wondering, his argument that we've never been able to design anything as good as ourselves is stupid. Can you crunch numbers as fast as a computer? Didn't think so, guess what, it was DESIGNED for that purpose. If we knew what conciousness was, I'm sure we could design something that's better at it than us since we just got it by accident.
You missed it too. The reason this was not "just a terrorist act" is because 2 of those planes were targetted at the white house and the capitol building. THAT is an attack on the US state, bordering on an act of war. The trouble is figuring out who is really "responsible" by supporting the action with (or possibly without) knowledge of its intended outcome. This isn't impossible, it's just difficult. If there is a nation which fits the criteria for "resposibility" then it WAS an act of war and you let the military loose to take 'em out. (which we have done, the problem being that like in Vietnam we skipped the whole decaration of war by congress thingy which would have made it clear to everyone EXACTLY why we were undertaking such an arduous and bloody campain and why we picked that particular nation.)
Watch what you say if you want to be taken seriously:
1. "basically indistiguishable" if you computer and your CD player are not absolutely indistinguishable running through the same DAC then either you don't know how to rip cd's, your setup is busted, or you're full of it. A cynical person would just pick the last.
2. As someone else here pointed out, only the latest version of Ogg even supports cross-channel correlations, until then it was two independent streams. Presumably you're hearing phase distortion due to slightly different things being done to each channel, but that's not what you said.
You're on the right track, (listen, listen, listen). You may have good ears and good equipment, but you should be more careful what you say if you want to not be disregarded.
(my current belief is that discussions like this are pointless without references to the quality of the original recording)
Slashdot Mirror
I'm not really intending to "defend" linux, I think OSX deserves props for finally bringing unix with a nice front-end on it to the general populace (something all the linux vendors seemed to fail to do for years). But as another pointed out OSX is actually just the latest version of NextStep.
Anyhow, many of your points are due to two things -- 1. OSX like all mac os's only runs on a limited set of hardware, this makes things TONS easier, especially in installation. 2. Linux have NEVER been widely shipped pre-installed. All those things you worry about you laptop not doing and "Top-end hardware support" would pretty much work with a preinstalled, properly system integrated Linux.
I just had an idea. Some linux vendor should start selling boxes with simpler linux distribution pre-installed, with KDE, all the extra crap stripped out with their own picked PC hardware. At least this would give people something to compare fairly against Windows and Mac. Breaking into the PC market by installing on existing machines is just nearly suicide due to the insane variety of hardware.
Of course the strength of the PC is the commodity hardware keeping prices down and avoiding dependency on single vendors.
Hmmm.
Actually if you can use nuclear power to produce the 150J but the 100 J of oil can be used in portable devices (gasoline in cars) whereas nuclear cannot it may still be economically viable. (though addmittedly at that point it will be incredibly expensive.)
This is like declaring that batteries are useless because they take more power to charge than they provide when used.
The economic viability argument is more robust, if it costs more money to supply oil than the market demand will bear (due to the availability of other energy sourcess) then oil companies simply won't exist.
-
1. (as you say) Lindows is required to give the beta testers the code if they ask for it.
- 2. Also the GPL says that those beta testers have the right to redistribute that code to anyone they please. i.e. Lindows cannot put additional restrictions on it.
That's it. And unless the beta testers are employees, it IS (non-internal) distribution.He made exactly the point that things are different in the case of patents, they still encourage public disclosure by not granting protection to unpatented devices. That covers pretty much all but the particle accelerators, in which case the community (physicists) is so oriented toward public disclosure that you can always find out how something was done, it's essential in order to judge the validity of the result.
This is like saying we need to "fix" language because it's possible to write a bad essay!
Programming languages would be a hinderance if they *prevented* you from doing it right. They don't, it's just the first try at something is usually not the optimal. Or you were in a hurry, or some people are just sloppy.
'available now via the internet'
oh really? show me where the raw data is. All I've ever found are web apps that think they know how you want to interface with the date (2-3hr blocks and keyword search, title and other info presented in weird places) and are impossible to integrate with anything. The reason the software available is a long way from a Tivo like service is that the data is not generally available in the form such an app would need.
If you know a place the raw data is available please let me know!
It seems obvious to me that there are two reasons that Tivo hasn't been "embraced"
1. It's hard to understand the advantages over a VCR. This doesn't mean there aren't any or that their impossible for normal people to use, it's just a hard sell. Nearly everyone already has multiple VCRs.
2. THE BIG ONE -- The absence of a removable media (like tape on a VCR) is a BIG minus. VCR's are essentially used for 3 things, time-shifting shows, "copying" shows/movies (i.e. recording them to keep for a while or to transport), and for playing rented tapes. Tivo does the first but due to the lack of a removable media it can't do the other two. A Tivo owner can't record something and then take the recording to his friend's house and watch it. It's locked in the Tivo.
If Tivo would simply be brave enough to also include a CDR/W drive that would make this thing a 100% feature-for-feature VCR replacement, wide adoption would be much less painful. A combo Tivo/DVD player is what is needed to actually *replace* a VCR in full functionality, but they don't sell these.
Because you wrote it, and you have just as much right to forbid them from selling your product as they do to forbid you from stealing a copy of theirs.
Is it not the ultimate compliment to you and an excellent way to spread good code/standards?
Ah, but with a free (lowercase) license you have relinquished control of the source to the point where you can't even be sure that they haven't completely screwed it up. You can't see what they've done to your code.
Similarly, why should application developers be barred from creating commercial apps for Linux or KDE?
This is just false. What's preventing them? They can compile AGAINST the libraries no problem, they just can't distribute modified versions of the code or it's products. But they're writing a new app so this does nothing to them.
The only reason seems a hatred of commercial developers, protecting your code from abuse and hidden extensions can be achieved with LGPL.
No the reason is to prevent a proprietary vendor from taking my code, adding 1% added value and selling it back to me. With GPL I get that 1% in exchange for giving them the 99%.
The GPL simply allows programmers to protect their work in exactly the way the previous poster indicated.
The entire reason to have a strong license like the GPL is for situations like the one the Ximian is in. The are up against a very powerful, well-funded, entrenched proprietary vendor. I predict the downfall will happen like this:
As soon as anyone starts to use Mono instead of MS's stuff, MS will just steal Mono's code, put it in their proprietary stuff and then tell their customers there is no reason to change because they just appropriated the feature that the customer was seeking. This is not so important in itself, but after this happens a few times Mono will likely be orphaned by the developer community. People don't like having their donated code sold for profit.
What is sad is that the "other companies" that want this license change do not understand that it will be the dowfall of the project.
It sounds like Ximian has decided that vendor support is worth sacrificing the viabiblity of a large section of their project. That's their decision, I just have the impression that it might be short-sighted.
Interesting how all the highly moderated comments are anti-GPL. This is slashdot, somethings wrong.
2 things, I think that was his point. If you've watched TV lately you know they're already biased.
Also it's worse than you think, the Audio Home Recording Act explicitly makes it legal to copy CDs for "non-commercial" purposes. that's even broader than just fair use.
I firmly believe that anyone who creates an "ultimate" linux system with only one display must have something wrong with them. What about a great display? Like 3 pci based video cards that play nice together and some whopper LCD panels?
And I should note that there are two more words that they should think about after that: multi PROCESSOR!
And my opinion on this system -- nice decoration, but PC boxes belong in the CLOSET!!
I'd say it must not have been a planned move on the part of MS simply because as of when I post this comment, the pro-MS astroturf is not being posted on slashdot. It must just be some random employee stuffing the box. On most MS stories there are plenty posts on MS's side, here there are only pot-shots, a few notes about how bad online polls are and some genuinely concerned opinions.
The purpose of the AHRA was to make sure that consumers (i.e. "the people") were legally allowed to simply copy music (i.e. non-commercially). In exchange the music companies got the blank media "royalty".
I think that we have all been decieved by the RIAA into believing that copying digital audio is illegal, it's not! Read the AHRA!
From Title 17 (Copyrights), Sec. 1008 of the U.S. CodeIf you wonder how this could be true it's because the royalty killed DAT but not CDR. doh! blindsided by the computer industry.
The point of Linux is to be incredibly adaptable and configurable. This means that your ideally designed user interface world is at odds with what Linux is.
That said, you have a valid comment, but in different terms. There probably should be A Linux which has a designed, consistent user interface. One distribution or a set of them should have this goal in mind, and some do. An OS is not just a user interface, and as such one should not design an OS for a single consistent user interface. An OS should efficiently execute system calls and manage computer resources. The user interface is a separate issue.
A LOT of the confusion on this subject is due to lack of the distiction between "a Linux" and "Linux".
On a side note, I think the biggest difference between linux and windows (now that we have KDE and GNOME for UIs) is device support. The device support culture is totally different, in MS drivers are written by the manufacturer, in Linux, they're written by the community. There has been, and will continue to be, a major culture clash here. When I get the newest wiz-bang device from shmo inc., in windows they can give me some driver that will probably hose my system in the end (a lot of Win instability comes from this) but it will work, and I'll blame MS for the instability. In Linux, new wiz-bang devices are generally not supported in the main distributions because it takes time and public review of code to get it there. What is a company to do. This problem has not been solved yet, and that is THE difference between windows and linux for the general user.
(this all comes down to the issue of how do you create a computer system that doesn't need an administrator to maintain it. Win PCs "freed" us from the admins, and MS pushed this hard, but now all those freed people are realising that tha admin might actually be needed (witness current MS security))
To further your point. Heiroglyphics is actually a syllabary system, the pictures actually have relatively little to do with the meaning, they just tell you what syllables to say, which converts into language. (this is why it took us so long to figure out how to translate them, we were trying to figure out what the pictures meant.)
You're thinking more like ideograms, which is how chinese and related languages are written, and for the reasons you say, even those have been simplified and adapted into something complicated and impossible to understand for the uninitiated.
even $20 a year would be fine with me for this data. I'm thinking some site I can login to with a name/password and they give me a file in a nice database format that I can use. No their box in my house. The point is I want the all the data (prevents profiling) with no strings attached other than standard copyright.
my question is where is this place now? Is anybody doing this?
I know about tvguide.com or tv.yahoo.com, but as you say, and another poster points out, they wouldn't like it (need ads) and the format probably changes constantly. This is probably what I would resort to if I started writing this thing now.
(you know, it's a sad state of affairs when I have essentially no expectation of somebody relying on "standard copyright")
You do realize that you have just succesfully argued that the box is irrelevant and all that matters is having a piece of sophisticated software for managing recording schedules.
I've actually been thinking about this (I already record all my TV on my own computer). The big hurdle seems to be getting the listings. Do you know where one could get widespread, accurate, TV listings free for download? It sure seems like there should be a place, but I have a suspicion that the *only* way to get this kind of information is through a pay service, that of course not only wants your money but want *you*. (i.e. wants to make it as hard as possible for you to use a competitor's system in the future).
Imagine a commoditized Tivo where the recorder/storage hardware, the scheduling software and the listing service were all separate and competition could occur on each individually. We already have a Free software solution for the first of these, but as I said above I think the nature of the third makes the second sticky.
from comments mentioned in kernel traffic this week I'd say it's because ALSA is so easily compiled outside the kernel tree. I'm using it with a debian stock kernel now, no kernel compile necessary. Just download the alsa packages, compile and modprobe.
A better question is probably why does redhat muck with their kernels' sound so much that the above isn't true for them? (is it?) They should just use ALSA.
That and the alsa guys seem commited to having a STABLE system before it goes into the kernel.
There are other posts here that probably explain things better but I'll be brief:
1. You don't know what you're doing and neither does your friend.
2. If you took a 2 year old windows install with no updates and tried to install the newest patch, I'm pretty sure you would have similar problems. It's not a very smart thing to do.
I have brought a RH6.2 box up to date. It doesn't require force or nodeps, it just requires installing all the updates. It IS really hard to do because of the RPM version change, but that is redhat's fault.
Sounds like you actually ran into exactly the problem with adults vs. children. This lady assumed that she knew how computers worked, and that nothing had changed in 20 years (a pretty self-centered assumption I think). She was probably never able to convince herself that she should just humbly give up her preconceptions and start from scratch -- which a child does automatically.
The comment about operating systems being wasteful is pretty revealing. She had succeeded in categorizing "new" computers as somehow lesser than the computers she was used to. This kind of attitude makes it almost impossible to learn new things, because now she is resistant to abandoning her old assumptions because she has convinced herself that they are somehow better.
Draconian Measures which Compromise Amendment I
I don't know about your particular case, but in my experience DVD *decoders* really suck. I've never seen a set-top box that I felt was up to snuff. Try xine or oms or anything that uses libmpeg2 for decoding on the computer. I think the hardware manufacurers skimped to make their players cheap and that you can see it in quality. Until I started fiddling with libmpeg2 I didn't know that the accuracy of your iDCT matters, and until recently the one published by intel was not up to spec.
The irony is probably that the better the encoding is (i.e. the better use it makes of the bitrate) probably the more damage a shoddy decoder will do.
The worship Linus faction won't like it, but I think Linus is just out of his league. He is not an academic, but he is trying to comment on an academic topic.
There is a difference between design and implementation. Linus is confusing the two, and thereby simply confusing himself and anyone "arguing" with him. (I think that it's impossible to argue with someone who maintains a self-inconsistent stance.) Let me try to clarify the distinction
Design: The process of determining requirements, classifying them, then planning out a system which satisfies these requirements to a desireable degree.
Implementation: The process of building a system which performs certain actions.
In the case of implementing a design, the actions performed somehow lead to satisfaction of the design requirements.
Now that that is clear, there is one more point of to clarify:
Subsystem: Part of a system, which typically performs a certain action, i.e. partially implements.
Critical point: One can DESIGN a subsystem. This can often lead to confusion on the difference between design and implementation.
One MUST realize that one of the most critical parts of design is the creation of the requirements! Improving implementation NEVER helps if you're solving the wrong problem!!!!
One well-known way of designing is to generate a bunch of ideas which might satisfy the requirements, you then go through, rank the ideas and then eliminate the ones that don't satisfy the requirements as well. This is a "natural selection" approach to design.
With all of that said, Linus' problem is that he is an implementor. The design is DONE! It's unix. All he has to do is implement. Now, in implementation subsystems will be designed and implemented on and on. And better designs for subsystems will be created all the time. The kernel itself is a subsystem, and as such isn't really part of the design, but part of the implementation. He has designed the kernel - Modularity, timeslicing, portability requirements, even making it open-source was a design decision. Oh and designs can evolve, i.e. be amended when more knowledge is gleaned. Example: One of the principle differences between the two VM's is that they set out to accomplish different goals. Many of the requirements are the same (they're both VM's after all) but some are different. By changing to a new VM, one is simply changing the design of a sub-system because it provides a better implementation of its functionality than the previous design. So the design of the kernel hasn't changed, part of its implementation has, it needs a VM subsystem just like before. Changing the kernel design would have been something along the lines of removing the VM, say because memory is so cheap that it is simply unecessary complication.
And for those wondering, his argument that we've never been able to design anything as good as ourselves is stupid. Can you crunch numbers as fast as a computer? Didn't think so, guess what, it was DESIGNED for that purpose. If we knew what conciousness was, I'm sure we could design something that's better at it than us since we just got it by accident.
You missed it too. The reason this was not "just a terrorist act" is because 2 of those planes were targetted at the white house and the capitol building. THAT is an attack on the US state, bordering on an act of war. The trouble is figuring out who is really "responsible" by supporting the action with (or possibly without) knowledge of its intended outcome. This isn't impossible, it's just difficult. If there is a nation which fits the criteria for "resposibility" then it WAS an act of war and you let the military loose to take 'em out. (which we have done, the problem being that like in Vietnam we skipped the whole decaration of war by congress thingy which would have made it clear to everyone EXACTLY why we were undertaking such an arduous and bloody campain and why we picked that particular nation.)
Watch what you say if you want to be taken seriously:
1. "basically indistiguishable" if you computer and your CD player are not absolutely indistinguishable running through the same DAC then either you don't know how to rip cd's, your setup is busted, or you're full of it. A cynical person would just pick the last.
2. As someone else here pointed out, only the latest version of Ogg even supports cross-channel correlations, until then it was two independent streams. Presumably you're hearing phase distortion due to slightly different things being done to each channel, but that's not what you said.
You're on the right track, (listen, listen, listen). You may have good ears and good equipment, but you should be more careful what you say if you want to not be disregarded.
(my current belief is that discussions like this are pointless without references to the quality of the original recording)