I didn't claim otherwise. Not all drives are capable of being modified to be regionless, so having the DVD encoded as region 0 would prevent this from being a problem. CSS is hardly a problem nowadays.
I don't see why anyone here is complaining about being forced to actually obey copyright laws
This isn't about copyright law. Playing a region 1 DVD outside region 1 is in no way a breach of the traditional concept of copyright - I'm paying the copyright holder money in order to gain the right to play the film, but the copyright holder is then refusing to allow me to view the film if my DVD hardware is from somewhere other than the same region. Region encoding is not a mechanism for protecting copyright, it's a mechanism for allowing different markets to be sold different products at different times and preventing things moving between these markets in a way that the companies think may harm their profits.
The companies making these [regionless DVD players] are basically stealing from the movie companies
Again, no. Region encoding does not protect copyright. You can produce pirated DVDs with region encoding without any difficulty.
There are legitimate regionless DVD players. The first generation of DVD drives for computers were sold for their ability to allow you to watch DVDs, but are all region 0. Restricting this after the fact means that there are some people (including me) with hardware that was not bought for the purpose of avoiding region encoding who are now being discriminated against. This doesn't make me terribly happy.
Firstly, as other people have pointed out, a lot of "region free" DVD players are actually switchable between regions an infinite number of times. Secondly, Linux DVD players (and I'd assume most Windows ones - the ATI one I played with did) allow you to play the video files directly, without going through the menu system. As a result, there's no opportunity for the script to run. Ironically, if the DVDs really are encoded in region 0 this could make playing them easier than it would otherwise be.
Right. There's reluctance. It's unproven. It's new. But there's awareness. Without the GPL, we wouldn't have had that to the same degree. People would happily take free code and make propriatory apps out of it. The idea is not to get free code used everywhere, the idea is to get free code used everywhere while keeping it free. If that's not what you want, then fine - don't use the GPL. It's not being forced upon you.
But the whole idea of FSF and GPL is to do away with copywrited software and any limitations that come with it. Yet you are stubbornly defending GPL using laws this license was designed to destroy.
It's called pragmatism. Idealistically releasing code into the public domain would just result in corporations folding it into propriatory software. Releasing it under the GPL forces them to redistribute the source to the software and encourages them to view this as a viable practice. Over time, people's attitudes are changed.
It might seem tired and overused argument but GPL really does force people into GPL version of what software freedom is about.
No. Yes, the GPL embodies the concept of "Free" as used by GNU. Why is this a problem? Providing you agree with this definition and see it as a good thing, the GPL is a good choice of license. If you disagree with it, then you're not obliged to use the GPL. The authors of the GPL decided that preventing people from being able to release code released under the GPL in propriatory products was a lesser evil than allowing them to do the same thing, and as a result you can't. I agree with them, and by implication so does everyone that releases software under the GPL. I'm not being forced to do so, and you're not being forced to use my code - nobody is being forced to accept the GNU version of freedom.
Above.net has never made any secret of their usage of the RBL to blackhole traffic. If you don't like it, don't peer with them. Above.net is under no obligation to carry any traffic other than that stipulated in their contracts with other ISPs. If this affected you, somebody somewhere signed a contract on your behalf that lets this happen.
This will encourage people to use proprietary browser plugins for windows, rather than developing native ones for Linux.
Isn't this already the case? Macromedia's Flash plugin (there's a free alternative, but last time I played with it it didn't seem to support anywhere near as much) and Real's Realmedia plugin are both widely used and closed. The moment Netscape supported closed-source plugins, we'd pretty much lost in that respect already.
What's more of a problem is that if companies can get away with just recommending their Windows plugin for Linux use, non-x86 users are left out in the cold.
cantab.net is a great example of a hideously implemented system. Every student at Cambridge has a unique identifier used for their email address, and all hell would break loose if any of these were ever reused. However, there is no official commitment to not reusing these identifiers and as a result a completely different local part is used for the address. This wouldn't be so bad, except that they've chosen
inital.surname.yearofentry
Namespace collisions? What namespace collisions?
Thank God the university development office don't run the university mail servers.
Natural selection says we're breeding to friggin much
Natural selection says nothing of the sort. Natural selection works to create organisms that will produce children that will produce more children than any of their competitors. It's pretty trivial to demonstrate this.
We have two genes, a and b. We have two organisms, X and Y. (I'm going to assume assexual reproduction here. The principle is much the same in sexually reproducing organisms, it would just take a bit longer). a is a gene that results in the parent having 4 children. b is a gene that results in the parent having 8 children. X carries a. Y carries b.
At generation 0, we have 1 a and 1 b.
At generation 1, we have 4 a and 8 b.
At generation 2, we have 16 a and 64 b.
At generation 3, we have 64 a and 256 b.
As can clearly be seen, b will spread throughout the population much faster than a. Now assume that starvation hits the population. No more than 512 organisms can live in the population at once. Assuming organisms carrying a and b are otherwise of similar fitness and hence the same proportion of each type dies per generation, we get:
At generation 4, we have 102 a and 410 b
At generation 5, we have 57 a and 455 b
At generation 6, we have 30 a and 482 b
Even in starvation, the gene giving fewer children is selected against. A mutation that controls the size of the population of a species is of absolutely no use whatsoever because of a single fundamental flaw - if you don't have children, you won't pass on that mutation. If you're going to try to use natural selection as an argument for anything, at least try to learn something about it first.
This is an excellent and important point, but it is based on the premise that Mother Nature knows better than we do what's right and wrong.
Which is demonstrably false. Mother Nature knows nothing. Selection has no higher plan other than the favouring of genes that make you more likely to live long enough to breed and to be able to breed successfully enough that your genes are spread to future generations, based on whatever the environmental conditions are at the time. Which is fine in a situation where the conditions in your ecosystem stay reasonably static over time, but works rather less well in a situation where the favoured conditions vary between pretty much every generation. Like the situation that we've been in for the past few hundred years, for example. We're able to predict the future better than nature is (we stand some chance of being right. Nature "assumes" that conditions are going to stay the same as they currently are, which is pretty much the least likely outcome), so imposing our own selection pressure on the genes at an earlier stage is more likely to lead to "well suited" individuals than just letting nature get on with it.
This doesn't mean that I'm in favour of genetically modified humans. The social considerations are pretty fascinating and damned difficult, and when we screw up it's likely to be pretty graphic.
On the other hand, there's evidence that mice degrade any paternal mitochondria that enter the egg. I don't think that this has been studied to any great extent in humans yet (it's quite a bit harder to get hold of large numbers of naturally fertalised human eggs in order to perform experiments on them, oddly enough), but there is reason to suspect that inheriting mitochondria from sperm is less likely than you might expect.
Most hardware with Macrovision support defaults to it being switched off. That way it has to be deliberately switched on in software, proving that you're using it and letting Macrovision claim their license fee. Oddly enough, most Linux drivers don't bother poking the register that enables Macrovision. The Linux drivers for em8300 based cards (Hollywood Plus, Creative dxr3, a few others) certainly don't enable Macrovision.
Re:I remember something like that...
on
Linux Anecdotes
·
· Score: 2
I've done something similar. While installing an old version of Debian, I needed to edit my fstab. The boot floppies came with a tiny editor called ae. Momentary brain fade led me to do
ae/dev/hda
Ok, my screen's full of binary garbage. Never mind. A brief bit of experimentation got me a file open prompt, and I had my fstab. I made a few alterations, and then saved. And, having just hit enter, saw where it was saving it. I spent the rest of the night rebuilding my partition table by hand.
If they want to set an example they should be
given equal weight.
Thereby increasing development costs significantly as they have to rewrite all their Gnome-based config utilities to use QT as well? Progeny made a decision to support one desktop environment over another in order to make life easier for themselves.
Along with X 4.x which is incompatable with all but the most expensive newest cards.
Pretty much the only vaguely common cards that XFree 4 doesn't support are the old S3 Trios (not the Trio3D, which is supported). Still, if you do own some ancient crufty thing, the Debian packagers have thoughtfully included modified XFree 3 packages that only support the cards that XFree 4 doesn't.
Note that pretty much every cheap graphics card on the market for the past 5 years is supported by XFree 4. Note also that the new Vesa driver in XFree 4 means that pretty much every graphics card on the planet is supported to some degree, something that wasn't true under XFree 3.
Re:Oracle submits a laundry list of changes?
on
Preview Of Linux 2.5
·
· Score: 2
If Sun or Microsoft pointed out that the current kernel code compromised the performance of some of their applications and that it could be fixed without compromising the performance of anything else, I really can't imagine anyone refusing to do anything about it.
Re:Oracle submits a laundry list of changes?
on
Preview Of Linux 2.5
·
· Score: 2
I don't see why. It's to Linux's benefit if it performs well with Oracle, and if this can be done without compromising any other components of the kernel then why not? Oracle aren't demanding anything, they're merely pointing out that if these features were implemented Oracle (and, presumably, similar large-scale apps) preformance will be better. Most of the items listed are flaws in the current kernel, so you could always argue that fixing them is just a matter of neatness...
Apoptosis is, as the article hints, a vital part of development - at various stages of embryogenesis, tissue that has developed needs to be removed for the next step. Apoptosis is also involved in the regulation of cell growth. If a cell detects that it's becoming non-functional, apoptosis is triggered and the cell dies. This is absolutely vital in cancer prevention. Every day cells in your body pick up chance mutations that could potentially lead to cancer, but in the vast majority of cases they kill themselves before this can happen. However, if a mutation occurs in the pathway that triggers apoptosis then this can't happen, and the cell may go on to become cancerous.
AIF is important because it's likely that many cancers will have a defective copy of it. With advances in gene therapy, it should soon be possible to insert new copies of the gene into the cancerous cells thereby triggering apoptosis in them and destroyin the tumour. However, it's not much use for making you live forever. You don't die because cells decide that you've lived for too long - you die due to disease or failure of organs induced by wear and tear and picking up of mutations. It's possible that inhibiting AIF might lead to some individuals living slightly longer, but only because cells that are defective would hang around until they broke down completely rather than killing themselves cleanly. It wouldn't be terribly pleasant.
What could make it easier? I've also never had apt, dpkg or dselect crash on me or behave in strange fashion. The same is true of RPM. What sort of instability are we talking about here?
As for Debian become a standard, I hope the auther is kidding! Debian may not be the worst distrobution, but RedHat trumps it in every catagory...featurs, stability and price.
Features are somewhat in the eye of the beholder, but I'll happily admit that Red Hat's hardware detection and install are much better than anything in Debian. Progeny have a Debian-based dist that solves that one nicely. Other than that, what features do you feel are present in Red Hat that aren't adequately available in Debian? In terms of stability, both are based on the same kernel and use the same XFree. I've had pure Debian systems running for over a year without crashing. Again, what stability issues do you feel exist with Debian? Stability is certainly not something that I'd bae a choice of Linux distribution on, mainly because they're all pretty much identical in this regard. As for price - Debian is free. Completely and utterly. You can find people who will sell you official CDs for the price of duplication, and you'll get the complete distribution. How much cheaper do you want it?
apt-get can also build source - apt-get -b source will build it - however this isn't done like the ports tree - it won't pick up all the dependancies in the same way that the binary packages will, or at least not that I've seen while playing with building apache.
The latest version of apt supports the build-depends field in packages, which fixes this problem.
Downsides: no CVS - if a package gets a 5 line patch you need a whole new binary or source package for it.
That depends. If it's an upstream change, yes. If the package maintainer makes an alteration, you keep the same upstream tarball and just download the new version of the Debian alterations.
Why would we need YAPM (Yet Annother Package Manager)??
At the moment there are RPM's, DEB's (apt) and they are cool enough.. then there's the good old tarballs, and BeOS PKGs', then there's QNX' package manager, and loads of time..
It's not YAPM in that sense. It interfaces with your distribution's existing package management system, presenting a uniform interface to it regardless of whether you're using Debian, Red Hat or SuSe while at the same time giving you various extra features that you may (or may not) need. It's not a replacement for RPMs or Debs, it's a replacement for things like dselect.
My problem is with the issue of bolt-on updaters. I personally have used these things before, and I have found that they are more trouble than they are worth.
Define bolt-on. When it comes to Debian, for instance, you're merely replacing the package management utilities that come with the system (dpkg and apt) with another one that uses the same data files. There's no difference in the amount of underlying control over the operating system.
I would not trust my operating system to a third-party updater; after all, given that Unix is a server OS, you should compile from scratch anyway (to ensure maximum speed and so on) - the value of this is limited.
Ah, right - this would explain your confusion. You're missing the point entirely. Many people don't want to have to download and compile every piece of software they want to use. They want to be able to say "I want that" and for it to appear on their machine. This isn't limited to users. I admin over 20 machines consisting of 4 different architectures. If I had to compile everything by hand it would take me forever. However, the wonders of the Debian package management system mean that I can keep them running, up to date and secure while at the same time doing a full-time degree. Package management is a wonderful example of a labour-saving device. Don't knock it just because you're a die-hard "compile everything from source" freak.
If you want this, you should get an OS where this facility is integrated; for example, Windows 2000 now has built-in management and update facilities, and this might be a more appropriate route to take.
Uhm. Linux (in the shape of Debian, Red Hat, Mandrake, Caldera, SuSe, the vast majority of other distributions) is an OS that has had a package management system integrated for several years. The copy of AIX I have here from 1992 had a package management facility. Doing everything the long, hard, painful way for minimal gain is not the UNIX way.
Slashdot Mirror
I didn't claim otherwise. Not all drives are capable of being modified to be regionless, so having the DVD encoded as region 0 would prevent this from being a problem. CSS is hardly a problem nowadays.
I'm assuming this is a troll, but still...
I don't see why anyone here is complaining about being forced to actually obey copyright laws
This isn't about copyright law. Playing a region 1 DVD outside region 1 is in no way a breach of the traditional concept of copyright - I'm paying the copyright holder money in order to gain the right to play the film, but the copyright holder is then refusing to allow me to view the film if my DVD hardware is from somewhere other than the same region. Region encoding is not a mechanism for protecting copyright, it's a mechanism for allowing different markets to be sold different products at different times and preventing things moving between these markets in a way that the companies think may harm their profits.
The companies making these [regionless DVD players] are basically stealing from the movie companies
Again, no. Region encoding does not protect copyright. You can produce pirated DVDs with region encoding without any difficulty.
There are legitimate regionless DVD players. The first generation of DVD drives for computers were sold for their ability to allow you to watch DVDs, but are all region 0. Restricting this after the fact means that there are some people (including me) with hardware that was not bought for the purpose of avoiding region encoding who are now being discriminated against. This doesn't make me terribly happy.
Firstly, as other people have pointed out, a lot of "region free" DVD players are actually switchable between regions an infinite number of times. Secondly, Linux DVD players (and I'd assume most Windows ones - the ATI one I played with did) allow you to play the video files directly, without going through the menu system. As a result, there's no opportunity for the script to run. Ironically, if the DVDs really are encoded in region 0 this could make playing them easier than it would otherwise be.
I can do this in Windows from the taskbar, with no reboots or anything. Why can't I do it in X?
Because up until recently, there was no way for the X server to notify the clients (which include the window manager) that the resolution had changed.
Right. There's reluctance. It's unproven. It's new. But there's awareness. Without the GPL, we wouldn't have had that to the same degree. People would happily take free code and make propriatory apps out of it. The idea is not to get free code used everywhere, the idea is to get free code used everywhere while keeping it free. If that's not what you want, then fine - don't use the GPL. It's not being forced upon you.
But the whole idea of FSF and GPL is to do away with copywrited software and any limitations that come with it. Yet you are stubbornly defending GPL using laws this license was designed to destroy.
It's called pragmatism. Idealistically releasing code into the public domain would just result in corporations folding it into propriatory software. Releasing it under the GPL forces them to redistribute the source to the software and encourages them to view this as a viable practice. Over time, people's attitudes are changed.
It might seem tired and overused argument but GPL really does force people into GPL version of what software freedom is about.
No. Yes, the GPL embodies the concept of "Free" as used by GNU. Why is this a problem? Providing you agree with this definition and see it as a good thing, the GPL is a good choice of license. If you disagree with it, then you're not obliged to use the GPL. The authors of the GPL decided that preventing people from being able to release code released under the GPL in propriatory products was a lesser evil than allowing them to do the same thing, and as a result you can't. I agree with them, and by implication so does everyone that releases software under the GPL. I'm not being forced to do so, and you're not being forced to use my code - nobody is being forced to accept the GNU version of freedom.
Above.net has never made any secret of their usage of the RBL to blackhole traffic. If you don't like it, don't peer with them. Above.net is under no obligation to carry any traffic other than that stipulated in their contracts with other ISPs. If this affected you, somebody somewhere signed a contract on your behalf that lets this happen.
This will encourage people to use proprietary browser plugins for windows, rather than developing native ones for Linux.
Isn't this already the case? Macromedia's Flash plugin (there's a free alternative, but last time I played with it it didn't seem to support anywhere near as much) and Real's Realmedia plugin are both widely used and closed. The moment Netscape supported closed-source plugins, we'd pretty much lost in that respect already.
What's more of a problem is that if companies can get away with just recommending their Windows plugin for Linux use, non-x86 users are left out in the cold.
I can run Debian on my Macintosh SE/30 machines?
Yup. Debian is the only dist which still keeps M68k up to date.
cantab.net is a great example of a hideously implemented system. Every student at Cambridge has a unique identifier used for their email address, and all hell would break loose if any of these were ever reused. However, there is no official commitment to not reusing these identifiers and as a result a completely different local part is used for the address. This wouldn't be so bad, except that they've chosen
inital.surname.yearofentry
Namespace collisions? What namespace collisions?
Thank God the university development office don't run the university mail servers.
Natural selection says we're breeding to friggin much
Natural selection says nothing of the sort. Natural selection works to create organisms that will produce children that will produce more children than any of their competitors. It's pretty trivial to demonstrate this.
We have two genes, a and b. We have two organisms, X and Y. (I'm going to assume assexual reproduction here. The principle is much the same in sexually reproducing organisms, it would just take a bit longer). a is a gene that results in the parent having 4 children. b is a gene that results in the parent having 8 children. X carries a. Y carries b.
At generation 0, we have 1 a and 1 b.
At generation 1, we have 4 a and 8 b.
At generation 2, we have 16 a and 64 b.
At generation 3, we have 64 a and 256 b.
As can clearly be seen, b will spread throughout the population much faster than a. Now assume that starvation hits the population. No more than 512 organisms can live in the population at once. Assuming organisms carrying a and b are otherwise of similar fitness and hence the same proportion of each type dies per generation, we get:
At generation 4, we have 102 a and 410 b
At generation 5, we have 57 a and 455 b
At generation 6, we have 30 a and 482 b
Even in starvation, the gene giving fewer children is selected against. A mutation that controls the size of the population of a species is of absolutely no use whatsoever because of a single fundamental flaw - if you don't have children, you won't pass on that mutation. If you're going to try to use natural selection as an argument for anything, at least try to learn something about it first.
This is an excellent and important point, but it is based on the premise that Mother Nature knows better than we do what's right and wrong.
Which is demonstrably false. Mother Nature knows nothing. Selection has no higher plan other than the favouring of genes that make you more likely to live long enough to breed and to be able to breed successfully enough that your genes are spread to future generations, based on whatever the environmental conditions are at the time. Which is fine in a situation where the conditions in your ecosystem stay reasonably static over time, but works rather less well in a situation where the favoured conditions vary between pretty much every generation. Like the situation that we've been in for the past few hundred years, for example. We're able to predict the future better than nature is (we stand some chance of being right. Nature "assumes" that conditions are going to stay the same as they currently are, which is pretty much the least likely outcome), so imposing our own selection pressure on the genes at an earlier stage is more likely to lead to "well suited" individuals than just letting nature get on with it.
This doesn't mean that I'm in favour of genetically modified humans. The social considerations are pretty fascinating and damned difficult, and when we screw up it's likely to be pretty graphic.
On the other hand, there's evidence that mice degrade any paternal mitochondria that enter the egg. I don't think that this has been studied to any great extent in humans yet (it's quite a bit harder to get hold of large numbers of naturally fertalised human eggs in order to perform experiments on them, oddly enough), but there is reason to suspect that inheriting mitochondria from sperm is less likely than you might expect.
Most hardware with Macrovision support defaults to it being switched off. That way it has to be deliberately switched on in software, proving that you're using it and letting Macrovision claim their license fee. Oddly enough, most Linux drivers don't bother poking the register that enables Macrovision. The Linux drivers for em8300 based cards (Hollywood Plus, Creative dxr3, a few others) certainly don't enable Macrovision.
I've done something similar. While installing an old version of Debian, I needed to edit my fstab. The boot floppies came with a tiny editor called ae. Momentary brain fade led me to do
/dev/hda
ae
Ok, my screen's full of binary garbage. Never mind. A brief bit of experimentation got me a file open prompt, and I had my fstab. I made a few alterations, and then saved. And, having just hit enter, saw where it was saving it. I spent the rest of the night rebuilding my partition table by hand.
If they want to set an example they should be
given equal weight.
Thereby increasing development costs significantly as they have to rewrite all their Gnome-based config utilities to use QT as well? Progeny made a decision to support one desktop environment over another in order to make life easier for themselves.
Along with X 4.x which is incompatable with all but the most expensive newest cards.
Pretty much the only vaguely common cards that XFree 4 doesn't support are the old S3 Trios (not the Trio3D, which is supported). Still, if you do own some ancient crufty thing, the Debian packagers have thoughtfully included modified XFree 3 packages that only support the cards that XFree 4 doesn't.
Note that pretty much every cheap graphics card on the market for the past 5 years is supported by XFree 4. Note also that the new Vesa driver in XFree 4 means that pretty much every graphics card on the planet is supported to some degree, something that wasn't true under XFree 3.
If Sun or Microsoft pointed out that the current kernel code compromised the performance of some of their applications and that it could be fixed without compromising the performance of anything else, I really can't imagine anyone refusing to do anything about it.
I don't see why. It's to Linux's benefit if it performs well with Oracle, and if this can be done without compromising any other components of the kernel then why not? Oracle aren't demanding anything, they're merely pointing out that if these features were implemented Oracle (and, presumably, similar large-scale apps) preformance will be better. Most of the items listed are flaws in the current kernel, so you could always argue that fixing them is just a matter of neatness...
Apoptosis is, as the article hints, a vital part of development - at various stages of embryogenesis, tissue that has developed needs to be removed for the next step. Apoptosis is also involved in the regulation of cell growth. If a cell detects that it's becoming non-functional, apoptosis is triggered and the cell dies. This is absolutely vital in cancer prevention. Every day cells in your body pick up chance mutations that could potentially lead to cancer, but in the vast majority of cases they kill themselves before this can happen. However, if a mutation occurs in the pathway that triggers apoptosis then this can't happen, and the cell may go on to become cancerous.
AIF is important because it's likely that many cancers will have a defective copy of it. With advances in gene therapy, it should soon be possible to insert new copies of the gene into the cancerous cells thereby triggering apoptosis in them and destroyin the tumour. However, it's not much use for making you live forever. You don't die because cells decide that you've lived for too long - you die due to disease or failure of organs induced by wear and tear and picking up of mutations. It's possible that inhibiting AIF might lead to some individuals living slightly longer, but only because cells that are defective would hang around until they broke down completely rather than killing themselves cleanly. It wouldn't be terribly pleasant.
(Note: IAAB)
Also, RPM is clearly much easier to use and seems more stable, at least to me.
apt-get install foobar3
apt-get remove frotz12-dev
What could make it easier? I've also never had apt, dpkg or dselect crash on me or behave in strange fashion. The same is true of RPM. What sort of instability are we talking about here?
As for Debian become a standard, I hope the auther is kidding! Debian may not be the worst distrobution, but RedHat trumps it in every catagory...featurs, stability and price.
Features are somewhat in the eye of the beholder, but I'll happily admit that Red Hat's hardware detection and install are much better than anything in Debian. Progeny have a Debian-based dist that solves that one nicely. Other than that, what features do you feel are present in Red Hat that aren't adequately available in Debian? In terms of stability, both are based on the same kernel and use the same XFree. I've had pure Debian systems running for over a year without crashing. Again, what stability issues do you feel exist with Debian? Stability is certainly not something that I'd bae a choice of Linux distribution on, mainly because they're all pretty much identical in this regard. As for price - Debian is free. Completely and utterly. You can find people who will sell you official CDs for the price of duplication, and you'll get the complete distribution. How much cheaper do you want it?
apt-get can also build source - apt-get -b source will build it - however this isn't done like the ports tree - it won't pick up all the dependancies in the same way that the binary packages will, or at least not that I've seen while playing with building apache.
The latest version of apt supports the build-depends field in packages, which fixes this problem.
Downsides: no CVS - if a package gets a 5 line patch you need a whole new binary or source package for it.
That depends. If it's an upstream change, yes. If the package maintainer makes an alteration, you keep the same upstream tarball and just download the new version of the Debian alterations.
Why would we need YAPM (Yet Annother Package Manager)??
At the moment there are RPM's, DEB's (apt) and they are cool enough.. then there's the good old tarballs, and BeOS PKGs', then there's QNX' package manager, and loads of time..
It's not YAPM in that sense. It interfaces with your distribution's existing package management system, presenting a uniform interface to it regardless of whether you're using Debian, Red Hat or SuSe while at the same time giving you various extra features that you may (or may not) need. It's not a replacement for RPMs or Debs, it's a replacement for things like dselect.
My problem is with the issue of bolt-on updaters. I personally have used these things before, and I have found that they are more trouble than they are worth.
Define bolt-on. When it comes to Debian, for instance, you're merely replacing the package management utilities that come with the system (dpkg and apt) with another one that uses the same data files. There's no difference in the amount of underlying control over the operating system.
I would not trust my operating system to a third-party updater; after all, given that Unix is a server OS, you should compile from scratch anyway (to ensure maximum speed and so on) - the value of this is limited.
Ah, right - this would explain your confusion. You're missing the point entirely. Many people don't want to have to download and compile every piece of software they want to use. They want to be able to say "I want that" and for it to appear on their machine. This isn't limited to users. I admin over 20 machines consisting of 4 different architectures. If I had to compile everything by hand it would take me forever. However, the wonders of the Debian package management system mean that I can keep them running, up to date and secure while at the same time doing a full-time degree. Package management is a wonderful example of a labour-saving device. Don't knock it just because you're a die-hard "compile everything from source" freak.
If you want this, you should get an OS where this facility is integrated; for example, Windows 2000 now has built-in management and update facilities, and this might be a more appropriate route to take.
Uhm. Linux (in the shape of Debian, Red Hat, Mandrake, Caldera, SuSe, the vast majority of other distributions) is an OS that has had a package management system integrated for several years. The copy of AIX I have here from 1992 had a package management facility. Doing everything the long, hard, painful way for minimal gain is not the UNIX way.