Those features aren't missing? Hmm, ok, mea culpa. I might give it a crack then.
As for binary modules, I believe RMS alluded to some proprietary firmware inside some linux module source code (eg stuff of the form char *firmware[] = "\x17\xa4\x29\x77... etc). And as for BK, there's already one example where the owner of BitKeeper has forced one kernel developer to choose between bitkeeper access to the Linux repository, and developing a challenger to BitKeeper. Yes he is being reasonable in the circumstances but that still does not make it the sort of choice that ANY open source developer should have to make. I for one am cheering for Subversion.
I agree with the developers in the choice. I also wish nVidia, along with the other companies, GPLed their drivers. After all, they are _drivers_ for the hardware I bought. And if they don't want to, they could at least release the complete specifications for the hardware so we can build our own 3D accelerated drivers.
Why don't you ask them to release their hardware designs too while you're at it? The driver is as integral a part of the nVidia architecture as the hardware is -- this driver is nV's competitive edge. You can't honestly expect them to release that. I wholly agree about point two though.
The rest of these developments are... well, sad really. Why this GPL mania all of a sudden? As far as I'm concerned, Linus has absolutely no right to claim the moral highground here -- he uses BitKeeper to manage the kernel; this is a piece of software written by a talented, though bigoted jerk who has made it quite clear that he intends to stick around and is quite intent on beating out any competition (see the Subversion fiasco). Then he constantly moans about how everyone hates him and his license but he cannot operate a business under any other one. My advice to him is not to impose a total anathema on this community: pull out of the Linux development process or Arch, Subversion, or something else entirely will grow into a monster and destroy your SCM product, and any market opportunity it might have, inside Linux or out.
So on the one hand it's perfectly fine for one kind of encroaching commercial interest in Linux, but not another. Sorry, but this is pinning my bullshit-o-meter. I use Linux on my machine because it's the best UNIX clone for the desktop, or rather my desktop (show me a FreeBSD kernel with ACPI, Direct Rendering, FireWire and AGP and then we'll talk). But it's so incredibly sad to see such a great project turn into one great big egofest (because in light of the BitKeeper situation, that's all it is. An otherwise brilliant bunch of developers out for some sad little power trip. Why don't we get rid of that proprietary, binary firmware in the kernel source and then you can tell me all about your lofty goals of defending the GPL)
There have been great strides in artificial eye technology: I think Slashdot even posted a link to a rudimentary artificial eye prototype already: the person with these artificial eyes percieves sight as flashes around the outlines of objects. It's amazing stuff and it's certainly an area of active research.
The main problem though is interfacing with the brain, and that's not an easy nut to crack. In particular, there is pretty much no hope at all for someone who is blind from birth; such people have a stunted visual cortex so even if they were given eagle eye artificial optics, it would be useless as their brains would be incapable of processing the information, so they wouldn't even understand the _concept_ of vision. I'd link to the story if I could find it, but suffice it to say it'll be quite some time before all forms of blindness can be eradicated.
And on the topic of vision concept, here's an idle thought: imagine if everything red looked green to you, everything green looked blue and everything blue looked red. How would you know that your perception of a given colour wasn't the same as someone else's? (answer: you can't. Think about it. Maybe this explains why some people have totally screwed up senses of aesthetics. Or in fact you could extend it further; how do you know that other people even 'see' the same way as you do? A few very rare people can see four primary colours. Imagine what THAT would feel like. But I've gone even further off topic now than I was originally...)
Yes we know. Yes we've heard it all before. Yes it's nice that some people are realising this. But we also know that no matter how many humble recommendations the actually knowledgeable people submit, Microsoft is still going to bribe and strong-arm their way into this market as well.
Show me a nationwide IT deployment that doesn't run Microsoft and then I'll REALLY be impressed. As it stands though... yeah it's nice but it ultimately doesn't mean much. I haven't heard much about that Peruvian Free Software bill making much progress either. (then again, Peru hates American influence just about enough for that to have a hope of happening, I think)
Buzzwords mainly, but basically some bloke picked over the specs for TCP/IP, put together some tools that do really pathological things with packets and take advantage of what various TCP/IP implementations expect and use that to agressively map networks.
Uh... in other words, nothing new whatsoever. NMAP's been doing this for ages, this is just more of the same. At least that's what it looks like, the submitter did an absolutely lousy job of actually getting to the point (what the fuck does "Paketto Keiretsu" actually DO!?)
Kids these days. Until quite recently I was hosting several PHP based websites off a P75 with 24MB RAM and it was actually working remarkably well. Ok then some tit went and installed vBulletin so I had to upgrade but I personally blame the fact that vBulletin is written like crap rather than the server itself - 111 days uptime it got: went up when I built it, the next time it went down is when I had to decommission it.
Had a look at the API, it's... messy to say the least. There's a function that lets you touch off a list of async io operations (ie your "read on every single connection") with one system call, however there's no corresponding function that says "Here's a list of pending AIO ops, give me, now, a list of all the completed or errored-out ones". You can opt to have a realtime signal delivered with the corresponding AIO struct (or fd or something...) but from what I've seen of signals they're pretty evil.
Dunno, I can see how this library would be useful but it stinks a bit too much of "design by committee" for my liking, and it isn't terribly well supported anyway (Linux, for instance, only has an AIO that is emulated by creating a thread for each AIO op. Which is precisely what we're trying to avoid in the first place)
AIO is a way of beginning a large series of IO operations and leaving the kernel to complete them while you get on with something else (or that's the best definition I can find so far). That still doesn't solve the problem of how to efficiently serve a small number of active connections without ignoring the inactive connections for any extended period of time.
From what I understand NGPT is mainly a user space thing. Why not go with the 1:1 one in the kernel (NPTL or whatever), just have a libpthread.so (NPTL runtime) and libpthread-mn.so (NGPT). From a programmer's standpoint, when I say pthread_create() I want to know exactly what that does: with NPTL I know what happens. With NGPT I don't. Also, the old rule of "Don't pay for what you don't use" applies. If I'm going to have just, say, four threads, those four threads are going to run better as four kernel threads as opposed to 2 LWP's dynamically mapped between 4 CPU contexts.
But, again, I might want to write a server of some sort which handles hundreds of thousands of connections at once, but 99% are idle at any given time and the other 1% require some nontrivial processing sometimes and/or a long stream of data to be sent without prejudicing the other 99%. Now, for ANY 1:1 threading system, I can't just create x * 10^5 threads because the overhead would be colossal. But equally so, implementing this with poll() is going to be horrid, and if the amount of processing done on a connection is nontrivial and/or DoS'able, there's going to be tons of hairy context management code in there, until lo and behold you end up with a 1:N or M:N scheduling implementation yourself. NGPT could be very useful as a portable userspace library here, as these people have implemented an efficient M:N scheduler under GPL, something that hasn't existed before and could be very useful. I think these libraries might be much more complimentary than the article makes out.
Of course not. But in a perverse sort of way, I actually respect the writer of this letter for not heaping on a ton of condescending euphemisms and doubletalk. Why the hell do companies even bother with it anyway? because the consumer will get "offended"? Fair enough up to a point, but the only thing I get offended by is when somebody talks down to me as if I live in some sort of idealistic fantasy land.
As for the music industry's attitudes... well, deal with it. Cut away everything you can and you'll still have the simple fact that huge amounts of music are indeed traded online. Much of it is not a lost sale because the downloading parties are typically teenagers (who can't afford the current extortionate prices) but much of it is. You can argue about try before you buy or deflect the issue by going into cartel discussions all you want but it's still not legal. If you don't like that, lobby to change the law. Not enough people give a crap to change the law? Then that's unfortunate but then, that's democracy for you, and the current state of voter apathy is... beyond the scope of this discussion anyway.
Or you could make your own 'label' of sorts, get some talent together and prove that a community-oriented, P2P friendly, grass roots music distribution system really can work, and will yield much greater profits to the artists. I I'm not sitting around whinging about the music industry, I'm putting my money where my mouth is and actually doing just that. f enough people got together and made such a thing a reality, it'll send the traditional RIAA companies flying like so many dead leaves; isn't that what we all want? Then again if it's just me and a few others it's quite probably a foregone conclusion that it will fail considering the immense problems inherent in getting some listeners, but it can't hurt to try can it.
This isn't just a problem in IRC channels, on messageboards you'll often get a few trolls hell bent on crap flooding the forums (wait... this is Slashdot why am I saying something that's gotta be freaking obvious. Achem)
Anyways, what I certainly think might be nice is to have an RBL-like system somewhere that scans for open proxies and automatically blacklists them. When your server recieves a connection, it just sticks.rbl.openproxy-rbl.org or whatever on the end of the IP and sees if there's a response. If there is it drops the connection like it's carrying the plague (or Code Red as the case might be). Simple, and easy to cache seeing as you can just have a local BIND running to cache results for hosts who commonly connect.
Yes but you're not copying anything copyrighted are you. The data that goes back and forth between the server and yourself isn't copyrighted. How on earth the DMCA applies here I don't know, and why anyone would want to prosecute you under it is also beyond me entirely.
The presence of "Open" in something's name seems almost as big a guarantee that it's proprietary as seeing "Microsoft" in its name. Just about the only exception I can think of is OpenOffice.
Btw: the last actual war the U.S. had was WWII. Everything since then has been a "conflict" or "police action" or some other term.
Yeah, they've moved on from attacking just sovereign states to attacking entire abstract concepts (any nation states that might happen to involve are just a side issue)
It's a simple fact that 95% of "attacks" are quite harmless game-playing by "script kiddies", against which there's no need to defend.
Last I checked having some HTML file written in FrontPage saying "j00 h4v3 b33n 0wnz0r3d" in red on black where your index page is supposed to be doesn't do wonders for your company's reputation.
Why the hell do mobile phone companies keep harping on about their integrated cameras. Come on, think about it can you honestly imagine even a contrived situation when a mobile phone camera can do something that a disposable or digital camera can't do just as well if not better. I get tired of it. My current mobile phone is a Nokia 7110. I plan to continue using this thing until it falls apart: predictive text messaging, vibration function and even WAP is in there, and I got it dirt cheap too because it was already discontinued when I got mine. And I don't even use the WAP.
This isn't really ideal I suppose, though the manufacturers really need to just focus on more useful things. Broadband-speed, permanent near flat rate wireless access would rock: some other poster in a different story mentioned that his student brother/friend/acquaintance was working in Japan and could stream MP3's off his home machine and off his mobile phone on the way to work (and this guy was a student so he's hardly loaded). I want to be able to do that. Then there's all this fuss about personal area networks: you've got a mobile phone in your shirt pocket, connected to the internet and you can check your email and the like on your PDA. Or if you're in the car you can have a headset which uses a wireless bluetooth link to let you talk to people behind the wheel or initiate calls by voice: I want that too. All of these things would seriously rock.
But for crying out loud, as for these worthless gimmicks take your damn FM radios and digital cameras and integrated mobile phone/PDA jobs and shove them.
Debian might be slow with releases but I hope they're not going to go backwards. I'm typing this under KDE 2.2.2 and Linux 2.4.18, both came from binary packages straight from Woody.
Just today I was talking about how Linux really needs this sort of thing (well, that and a decent network filesystem; NFS is vile and AFS is... idiosyncratic. Coda is apparently not particularly suited to real world needs and Intermezzo is something completely different)
The LDAP integration is actually something they could really lever here, KDE could seriously do with a graphical LDAP admin system (ie one specifically designed for managing users in particular). That and they would do well to stick Kerberos up every concieveable orifice too - Single Sign On is a good thing, and I dont mean one controlled by a company with the letters M and S in its name.
Slashdot Mirror
...were these students once aspiring to be Enron employees or something? (Seriously though, A: who cares and B: what's the point)
Those features aren't missing? Hmm, ok, mea culpa. I might give it a crack then.
As for binary modules, I believe RMS alluded to some proprietary firmware inside some linux module source code (eg stuff of the form char *firmware[] = "\x17\xa4\x29\x77... etc). And as for BK, there's already one example where the owner of BitKeeper has forced one kernel developer to choose between bitkeeper access to the Linux repository, and developing a challenger to BitKeeper. Yes he is being reasonable in the circumstances but that still does not make it the sort of choice that ANY open source developer should have to make. I for one am cheering for Subversion.
Why don't you ask them to release their hardware designs too while you're at it? The driver is as integral a part of the nVidia architecture as the hardware is -- this driver is nV's competitive edge. You can't honestly expect them to release that. I wholly agree about point two though.
The rest of these developments are... well, sad really. Why this GPL mania all of a sudden? As far as I'm concerned, Linus has absolutely no right to claim the moral highground here -- he uses BitKeeper to manage the kernel; this is a piece of software written by a talented, though bigoted jerk who has made it quite clear that he intends to stick around and is quite intent on beating out any competition (see the Subversion fiasco). Then he constantly moans about how everyone hates him and his license but he cannot operate a business under any other one. My advice to him is not to impose a total anathema on this community: pull out of the Linux development process or Arch, Subversion, or something else entirely will grow into a monster and destroy your SCM product, and any market opportunity it might have, inside Linux or out.
So on the one hand it's perfectly fine for one kind of encroaching commercial interest in Linux, but not another. Sorry, but this is pinning my bullshit-o-meter. I use Linux on my machine because it's the best UNIX clone for the desktop, or rather my desktop (show me a FreeBSD kernel with ACPI, Direct Rendering, FireWire and AGP and then we'll talk). But it's so incredibly sad to see such a great project turn into one great big egofest (because in light of the BitKeeper situation, that's all it is. An otherwise brilliant bunch of developers out for some sad little power trip. Why don't we get rid of that proprietary, binary firmware in the kernel source and then you can tell me all about your lofty goals of defending the GPL)
First funny Soviet Russia post I've seen. Mod this up (not posting anonymously to lend it a bit of credibility)
There have been great strides in artificial eye technology: I think Slashdot even posted a link to a rudimentary artificial eye prototype already: the person with these artificial eyes percieves sight as flashes around the outlines of objects. It's amazing stuff and it's certainly an area of active research.
The main problem though is interfacing with the brain, and that's not an easy nut to crack. In particular, there is pretty much no hope at all for someone who is blind from birth; such people have a stunted visual cortex so even if they were given eagle eye artificial optics, it would be useless as their brains would be incapable of processing the information, so they wouldn't even understand the _concept_ of vision. I'd link to the story if I could find it, but suffice it to say it'll be quite some time before all forms of blindness can be eradicated.
And on the topic of vision concept, here's an idle thought: imagine if everything red looked green to you, everything green looked blue and everything blue looked red. How would you know that your perception of a given colour wasn't the same as someone else's? (answer: you can't. Think about it. Maybe this explains why some people have totally screwed up senses of aesthetics. Or in fact you could extend it further; how do you know that other people even 'see' the same way as you do? A few very rare people can see four primary colours. Imagine what THAT would feel like. But I've gone even further off topic now than I was originally...)
Yes we know. Yes we've heard it all before. Yes it's nice that some people are realising this. But we also know that no matter how many humble recommendations the actually knowledgeable people submit, Microsoft is still going to bribe and strong-arm their way into this market as well.
... yeah it's nice but it ultimately doesn't mean much. I haven't heard much about that Peruvian Free Software bill making much progress either. (then again, Peru hates American influence just about enough for that to have a hope of happening, I think)
Show me a nationwide IT deployment that doesn't run Microsoft and then I'll REALLY be impressed. As it stands though
Maybe, if some shambling pile of flesh with tubes and pins running in and out of it is your idea of a 'hottie'.
Then again, seven of nine didn't look terribly great until they took all those nasty implants off...
Buzzwords mainly, but basically some bloke picked over the specs for TCP/IP, put together some tools that do really pathological things with packets and take advantage of what various TCP/IP implementations expect and use that to agressively map networks.
Uh... in other words, nothing new whatsoever. NMAP's been doing this for ages, this is just more of the same. At least that's what it looks like, the submitter did an absolutely lousy job of actually getting to the point (what the fuck does "Paketto Keiretsu" actually DO!?)
Clearly not a codec implementation then, or not a full one. Besides, Sorenson will sue them into oblivion when/if they do get it working.
Kids these days. Until quite recently I was hosting several PHP based websites off a P75 with 24MB RAM and it was actually working remarkably well. Ok then some tit went and installed vBulletin so I had to upgrade but I personally blame the fact that vBulletin is written like crap rather than the server itself - 111 days uptime it got: went up when I built it, the next time it went down is when I had to decommission it.
Had a look at the API, it's ... messy to say the least. There's a function that lets you touch off a list of async io operations (ie your "read on every single connection") with one system call, however there's no corresponding function that says "Here's a list of pending AIO ops, give me, now, a list of all the completed or errored-out ones". You can opt to have a realtime signal delivered with the corresponding AIO struct (or fd or something...) but from what I've seen of signals they're pretty evil.
Dunno, I can see how this library would be useful but it stinks a bit too much of "design by committee" for my liking, and it isn't terribly well supported anyway (Linux, for instance, only has an AIO that is emulated by creating a thread for each AIO op. Which is precisely what we're trying to avoid in the first place)
AIO is a way of beginning a large series of IO operations and leaving the kernel to complete them while you get on with something else (or that's the best definition I can find so far). That still doesn't solve the problem of how to efficiently serve a small number of active connections without ignoring the inactive connections for any extended period of time.
From what I understand NGPT is mainly a user space thing. Why not go with the 1:1 one in the kernel (NPTL or whatever), just have a libpthread.so (NPTL runtime) and libpthread-mn.so (NGPT). From a programmer's standpoint, when I say pthread_create() I want to know exactly what that does: with NPTL I know what happens. With NGPT I don't. Also, the old rule of "Don't pay for what you don't use" applies. If I'm going to have just, say, four threads, those four threads are going to run better as four kernel threads as opposed to 2 LWP's dynamically mapped between 4 CPU contexts.
But, again, I might want to write a server of some sort which handles hundreds of thousands of connections at once, but 99% are idle at any given time and the other 1% require some nontrivial processing sometimes and/or a long stream of data to be sent without prejudicing the other 99%. Now, for ANY 1:1 threading system, I can't just create x * 10^5 threads because the overhead would be colossal. But equally so, implementing this with poll() is going to be horrid, and if the amount of processing done on a connection is nontrivial and/or DoS'able, there's going to be tons of hairy context management code in there, until lo and behold you end up with a 1:N or M:N scheduling implementation yourself. NGPT could be very useful as a portable userspace library here, as these people have implemented an efficient M:N scheduler under GPL, something that hasn't existed before and could be very useful. I think these libraries might be much more complimentary than the article makes out.
Of course not. But in a perverse sort of way, I actually respect the writer of this letter for not heaping on a ton of condescending euphemisms and doubletalk. Why the hell do companies even bother with it anyway? because the consumer will get "offended"? Fair enough up to a point, but the only thing I get offended by is when somebody talks down to me as if I live in some sort of idealistic fantasy land.
... beyond the scope of this discussion anyway.
As for the music industry's attitudes... well, deal with it. Cut away everything you can and you'll still have the simple fact that huge amounts of music are indeed traded online. Much of it is not a lost sale because the downloading parties are typically teenagers (who can't afford the current extortionate prices) but much of it is. You can argue about try before you buy or deflect the issue by going into cartel discussions all you want but it's still not legal. If you don't like that, lobby to change the law. Not enough people give a crap to change the law? Then that's unfortunate but then, that's democracy for you, and the current state of voter apathy is
Or you could make your own 'label' of sorts, get some talent together and prove that a community-oriented, P2P friendly, grass roots music distribution system really can work, and will yield much greater profits to the artists. I I'm not sitting around whinging about the music industry, I'm putting my money where my mouth is and actually doing just that. f enough people got together and made such a thing a reality, it'll send the traditional RIAA companies flying like so many dead leaves; isn't that what we all want? Then again if it's just me and a few others it's quite probably a foregone conclusion that it will fail considering the immense problems inherent in getting some listeners, but it can't hurt to try can it.
This isn't just a problem in IRC channels, on messageboards you'll often get a few trolls hell bent on crap flooding the forums (wait... this is Slashdot why am I saying something that's gotta be freaking obvious. Achem)
.rbl.openproxy-rbl.org or whatever on the end of the IP and sees if there's a response. If there is it drops the connection like it's carrying the plague (or Code Red as the case might be). Simple, and easy to cache seeing as you can just have a local BIND running to cache results for hosts who commonly connect.
Anyways, what I certainly think might be nice is to have an RBL-like system somewhere that scans for open proxies and automatically blacklists them. When your server recieves a connection, it just sticks
Yes but you're not copying anything copyrighted are you. The data that goes back and forth between the server and yourself isn't copyrighted. How on earth the DMCA applies here I don't know, and why anyone would want to prosecute you under it is also beyond me entirely.
The presence of "Open" in something's name seems almost as big a guarantee that it's proprietary as seeing "Microsoft" in its name. Just about the only exception I can think of is OpenOffice.
I'd hardly consider Deus Ex to be shit. Neither would most reviewers as a matter of fact.
Yeah, they've moved on from attacking just sovereign states to attacking entire abstract concepts (any nation states that might happen to involve are just a side issue)
It's a simple fact that 95% of "attacks" are quite harmless game-playing by "script kiddies", against which there's no need to defend.
Last I checked having some HTML file written in FrontPage saying "j00 h4v3 b33n 0wnz0r3d" in red on black where your index page is supposed to be doesn't do wonders for your company's reputation.
Why the hell do mobile phone companies keep harping on about their integrated cameras. Come on, think about it can you honestly imagine even a contrived situation when a mobile phone camera can do something that a disposable or digital camera can't do just as well if not better. I get tired of it. My current mobile phone is a Nokia 7110. I plan to continue using this thing until it falls apart: predictive text messaging, vibration function and even WAP is in there, and I got it dirt cheap too because it was already discontinued when I got mine. And I don't even use the WAP.
This isn't really ideal I suppose, though the manufacturers really need to just focus on more useful things. Broadband-speed, permanent near flat rate wireless access would rock: some other poster in a different story mentioned that his student brother/friend/acquaintance was working in Japan and could stream MP3's off his home machine and off his mobile phone on the way to work (and this guy was a student so he's hardly loaded). I want to be able to do that. Then there's all this fuss about personal area networks: you've got a mobile phone in your shirt pocket, connected to the internet and you can check your email and the like on your PDA. Or if you're in the car you can have a headset which uses a wireless bluetooth link to let you talk to people behind the wheel or initiate calls by voice: I want that too. All of these things would seriously rock.
But for crying out loud, as for these worthless gimmicks take your damn FM radios and digital cameras and integrated mobile phone/PDA jobs and shove them.
Debian might be slow with releases but I hope they're not going to go backwards. I'm typing this under KDE 2.2.2 and Linux 2.4.18, both came from binary packages straight from Woody.
I'm still getting nailed by Code Red. Weird how something can survive for two years without touching a single permanent storage device.
Just today I was talking about how Linux really needs this sort of thing (well, that and a decent network filesystem; NFS is vile and AFS is ... idiosyncratic. Coda is apparently not particularly suited to real world needs and Intermezzo is something completely different)
The LDAP integration is actually something they could really lever here, KDE could seriously do with a graphical LDAP admin system (ie one specifically designed for managing users in particular). That and they would do well to stick Kerberos up every concieveable orifice too - Single Sign On is a good thing, and I dont mean one controlled by a company with the letters M and S in its name.
American democracy is great. Every dollar is represented equally.
I've said it before yes but it seems particularly apt now.