I don't want to support the big labels either, and since I listen to pretty non-mainstream music anyway, the solution I've found is to buy almost exclusively from small labels (like Cuneiform, not to name them) where you know that they are "good guys" too (working their asses off to make it possible for non-mainstream music to be heard). I like CDs a lot; I have about 400 of them, and I keep buying them, because I *like* things (it's not a moral choice or anything like that). As much as I like to check things out via mp3s (and download rarities or off-album tracks, like Adrian Belew's new song), for me at least, an mp3 played on a computer can't even being to replace playing a CD in a proper stereo, and having the package, the liner notes, etc.
simple: repeatability. you can hear things for free on the radio, but you can't choose when, and the quality sucks, so you still buy the CD. with napster and mp3's, you can listen to what you want, whenever you want, and then you get to keep it, even burn it on CD-R. and the quality is likely to be good, too (disclaimer: I don't actually use napster so i don't know what the average quality there is), so there's not much of an incentive to still buy the CD. radio works (with their ASCAP/BMI/SESAC artist rates), because it's not good enough to discourage people from buying the CD.
now, what to do with this? well, my take is that the possibility to freely exchange data on the internet is real and is not going to go away, so anything that depends on that not being possible is going to have to change radically to survive. I'm sorry for any artists who get the bad end of the stick, but I don't think there's ultimately any point in trying to work a compromise. there simply is no compromise between the practical ability to share information and get away with it, and a system that depends on that being impossible. no law can stop millions of mostly anonymous file sharing and downloading monkeys.
as for Napster itself, who knows. maybe the current Napster company will get sued to oblivion, maybe not. ultimately, it's not very important either; the possibility is there, and Napster, being a company (with the need for "stragetic partnerships" and all that crap), is vulnerable. but it only takes ONE free, open-source, multi-OS (windows, mac, linux), non-centralized Gnutella-workalike for this to be 100% unstoppable. Gnutella itself almost made it, but got pulled before they had a chance to release source (that's what happens when you ignore "release early, release often"). Something else will replace them.
the whole thing is sad in a way, but refreshing in another. the record labels will not "go the way of the dinosaurs" like some people have been saying, but the whole landscape of popular music and its economics *is* going to be redrawn. it doesn't make sense to be for or against; it's like being for or against the tide. it's happening already.
well, Linux will still have plenty of competition in the Unix world... the 3 BSD's and the really high-end (64+ processors) OSs like Solaris. as for CDE and Motif, I don't hate them, but since their development appears to be as good as dead, I'm glad they're being replaced by open source alternatives. if they weren't dead already, the open source world would probably put more effort in cloning them than in replcing them.
I've found it works better when I *don't* have a local webserver getting the requests. So the thing to do if you're using a local Apache, is to tell your apache explicitly to bind to the 127.0.0.1 address, and use your/etc/hosts to redirect doubleclick and others to 127.0.0.2 (note the 2). it's still on the 'lo' interface, so it doesn't get out to the network, but apache won't get the hit.
That doesn't work for the long term; companies will just learn to make an DNS alias like myads.mysite.com CNAME ads.doubleclick.net. What we need is selective cookie settings, as in "these domains get to set cookies, any others don't, or the other way round), and for embedded content (not only images, but java, html in ilayers, and anything else that a browser will pull automatically when loading a page). Mozilla has something like this for cookies and images, but it doesn't seem to be working yet; at least I couldn't get the user interface for it to work on a daily snapshot a few days ago.
... and that's exactly why Gnutella would be a good idea to replace napster. open source, open protocol, open extensions to all kinds of filetypes, not just mp3. if you're going to condemn this because it makes copyright violations easier, you might as well shutdown the whole web for the same reason.
this article was full of exaggerations. does anyone really belive that
The deal has dealt the much-heralded geek community and its open-source development model a terrible blow, one from which it may never recover.
, really??
I, for one, don't. The worst that can happen is that/. could go down the drain in credibility. If it does, well, *shrug*, something else will come to replace it. It's not liek the free software / open source community is short of people wanting to run news sites. And I don't think VA will be stupid enough to let/. lose its credibility anyway.
actually, I don't buy any of them. but I see your point, and I see how advertising is (relatively) important. it just doesn't strike me as the first thing to worry about, in the list of potential dangers of the formation of a single large linux conglomerate.
your major concerns are about the *banner ads* of all things? well, I read slashdot for the stories, not for the ads.... what I sure hope is that we'll keep seeing stories about SGI's linux boxes and everything else that is interesting but competes with VA-Andover.
I remember reading somewhere that the question with A.E. Van Vogt's novels was not how good they are, but how they can be so fascinating, intense, and ultimately, influential, despite all their shortcomings. I think that sums the situation up. AEVV was mentally more daring and ready to try with his mind than anyone else around, and that's including the Beat Generation people. That put him in contact with ideas and things that have turned out to be less than hoped (e.g General Semantics), or even downright nefarious (e.g Dianetics). But that's the price to pay for being open-minded and ready to experiment. We all owe Van Vogt a collective thanks for the mass of intense ideas that he acquainted us with. And, what the hell, Korzybski still makes an interesting read.
no, you aren't purchasing a LICENSE to anything. you're purchasing recorded medium with a chunk of data on it, which you are allowed (by law, not by any license) to play, listen to, enjoy, make backup copies of, etc. you are also not allowed (by law, not by any license) to redistribute copies.
at least that's the case with CDs. I don't see any reason why DVDs would be any different.
huh? how do *privacy* policies have anything to do with *proprietary info* laws? what the DVD people did is figure something out by reverse engineering, and publish it. what the DoubleClick people are doing is figuring personal things about individuals and using them. there are arguments for the two kinds of protection: protection of privacy, and protection of trade secrets. but the two are not the same, and there's no reason why the laws about each should be exactly the same.
personally, I support something like this:
for trade secrets: employees and NDA-signers are bound by contract not to reveal them. anyone who reverse-engineers a product without being bound in a specific way is free to publish information about it. in other words: I don't want to give to companies the possibility to hide a secret somewhere in a user product, and then make it illegal for the user to figure out the secret. you can make it illegal to actually *use* the secret value in certain ways, but finding it and publishing it is OK. there's one way for companies to protect their research, which is patents; patent something, which publishes itit, and then no-one can use it without paying you. if you don't patent something, then it's up for grabs, via re-discovery, or reverse engineering. of course, this assumes a working, non-broken patent system, which rejects insufficiently specific patents, as well as obvious ones. (I'm not taking a stance as to whether algorithms shoudl be patentable here, that's another piece of debate).
for personal information about individuals: I believe in protection, here. no company, entity or organization should be able to keep information about you without you having strong rights on that information (review, modify and delete rights). opt-out is not enough; for things liek what DoubleClick is doing, they should be legally required to do it in an opt-in way. exception: if you have a commercial tie with a company (you're a customer, etc...), then they can keep info about you. (i.e companies are allowed, and should be, to have a db of client interaction past histories. but not to sell it to another company).
bugtraq sure makes me wish one thing these days: that all MS related bugs were moved to their own mailing list! I mean, after all, there already is an NTBugtraq list, so why not rename it MSBugtraq and keep MS stuff out of Bugtraq itself? right now the situation is annoying: NT people need to read two lists, and Unix people read one but skip more than 50% of it.
Does this mean that we have artificial life, or merely a perfect simulation? The program will only manipulate register contents, which are not connected to actual physical realities.
well, your brain only manipulates chemicals and electric currents, which are no more connected to physical realities than CPU registers. we're already a "simulation" if you will; it's just running on meat-hardware. it also happens to have (presumably) evolved there.
We will not be able to find out unless we (personally) undergo such a transfer...
no, by undergoing such a transfer we will not find out anything deep; we'll just see whether the technology works, and have two entities convinced that they are the real one, with one of the two possibly having some technical difficulties (interface imperfections) with severe psychological consequences. perfect the technical aspect enough, and the simulated one doesn't have a way to know that it is simulated just by introspecting and observing the universe. and then your fundamental problem of "what is an consciousness" remains.
the core problem is that consciousness is strongly tied to short-term memory, which works only one way (you remember the past, not the future -- except for Patrick Moraz's Future Memories). yet we experience time going forwards. the end result is that we have no clear intuitive picture of what it really means to duplicate a consciousness. assuming that the technology does make it possible unobtrusively, no-one doubts that the scanned guy won't feel his consciousness duplicating, or anything like that. he'll just continue to be himself, and if the body dies later, that consciousness will go with it.
if we take a purely external, descriptive stance, there is really no problem at all: the subjective feeling of "me" doesn't count (I think the big word for that is "epiphenomenon"), and you have two individual intelligences living in different universes, that happen to share a past up to a point. no problem. except that that doesn't make any distinction between the two, yet, if they scan me, I will still be mortal, while the scanned copy might well run forever, which is a mighty big difference.
I don't know of any theoretical framework out there that can make sense of this mess. religions that believe in some sort of soul don't solve the problem either, they just change the terms: now the difficulty is understanding if and how it can be duplicated, and if not, if and how there can be consciousness without it.
well, getting their banners and tracking cookies, but removing the cookie every day or more often, is a way of poisoning the db with lots of useless entries; however, these entries will eventually expire (no more hits in a long time => cookie must be lost; not associated to a real name profile => useless, expire it). a stronger way to poison the db would be to have a proxy that randomizes the content of the doubleclick cookie, within its usual syntax. depending on how their system is setup, you could either get ignored in most cases, or manage to assign your hits to other random people's profiles. but you'd need a lot of people doing that to have a significant impact, and most people just don't care enough. hell, *I* don't care enough either; I'm just happy to block them at/etc/hosts.
who needs a proxy! just 1) delete your entire cookie file once a day or so, and 2) stick the 50 or so biggest ad servers in your/etc/hosts (that's \windows\hosts for you windows users), assigned to a bogus IP like 127.0.0.2. start with ad.doubleclick.net, and add hosts as you see ads (under netscape, right-click to "copy image location")
in general, cookies are OK, and quite useful, for short-lived browser/server interaction state keeping. There is no real need for long-term cookies; at worst you'll have to enter a password a few times more. And clearing your cookie file very effectively dissociates any further browsing from any profile doubleclick may have of you.
huh??? have you even looked at Chomsky's minimalist program? it's very much in the line of his earlier efforts; the model has changed, and been simplified considerably, but it doesn't depart one bit from the idea that the basic structure of language is innate.
and no, I don't have a personal opinion on it, eiher way. Goverment&Binding (aka Principles&Params in its later versions) came forward enough to predict non-trivial language features that could be checked; I don't think minimalism has gone this far yet.
Slashdot Mirror
I don't want to support the big labels either, and since I listen to pretty non-mainstream music anyway, the solution I've found is to buy almost exclusively from small labels (like Cuneiform, not to name them) where you know that they are "good guys" too (working their asses off to make it possible for non-mainstream music to be heard). I like CDs a lot; I have about 400 of them, and I keep buying them, because I *like* things (it's not a moral choice or anything like that). As much as I like to check things out via mp3s (and download rarities or off-album tracks, like Adrian Belew's new song), for me at least, an mp3 played on a computer can't even being to replace playing a CD in a proper stereo, and having the package, the liner notes, etc.
now, what to do with this? well, my take is that the possibility to freely exchange data on the internet is real and is not going to go away, so anything that depends on that not being possible is going to have to change radically to survive. I'm sorry for any artists who get the bad end of the stick, but I don't think there's ultimately any point in trying to work a compromise. there simply is no compromise between the practical ability to share information and get away with it, and a system that depends on that being impossible. no law can stop millions of mostly anonymous file sharing and downloading monkeys.
as for Napster itself, who knows. maybe the current Napster company will get sued to oblivion, maybe not. ultimately, it's not very important either; the possibility is there, and Napster, being a company (with the need for "stragetic partnerships" and all that crap), is vulnerable. but it only takes ONE free, open-source, multi-OS (windows, mac, linux), non-centralized Gnutella-workalike for this to be 100% unstoppable. Gnutella itself almost made it, but got pulled before they had a chance to release source (that's what happens when you ignore "release early, release often"). Something else will replace them.
the whole thing is sad in a way, but refreshing in another. the record labels will not "go the way of the dinosaurs" like some people have been saying, but the whole landscape of popular music and its economics *is* going to be redrawn. it doesn't make sense to be for or against; it's like being for or against the tide. it's happening already.
well, Linux will still have plenty of competition in the Unix world... the 3 BSD's and the really high-end (64+ processors) OSs like Solaris. as for CDE and Motif, I don't hate them, but since their development appears to be as good as dead, I'm glad they're being replaced by open source alternatives. if they weren't dead already, the open source world would probably put more effort in cloning them than in replcing them.
I've found it works better when I *don't* have a local webserver getting the requests. So the thing to do if you're using a local Apache, is to tell your apache explicitly to bind to the 127.0.0.1 address, and use your /etc/hosts to redirect doubleclick and others to 127.0.0.2 (note the 2). it's still on the 'lo' interface, so it doesn't get out to the network, but apache won't get the hit.
That doesn't work for the long term; companies will just learn to make an DNS alias like myads.mysite.com CNAME ads.doubleclick.net. What we need is selective cookie settings, as in "these domains get to set cookies, any others don't, or the other way round), and for embedded content (not only images, but java, html in ilayers, and anything else that a browser will pull automatically when loading a page). Mozilla has something like this for cookies and images, but it doesn't seem to be working yet; at least I couldn't get the user interface for it to work on a daily snapshot a few days ago.
... and that's exactly why Gnutella would be a good idea to replace napster. open source, open protocol, open extensions to all kinds of filetypes, not just mp3. if you're going to condemn this because it makes copyright violations easier, you might as well shutdown the whole web for the same reason.
same thing in Swedish. let's block *.se!
now you know what shape to give to your OK buttons.
cat >> $HOME/.netscape/preferences.js
user_pref("browser.chrome.disableMyShopping", true);
^D
I, for one, don't. The worst that can happen is that /. could go down the drain in credibility. If it does, well, *shrug*, something else will come to replace it. It's not liek the free software / open source community is short of people wanting to run news sites. And I don't think VA will be stupid enough to let /. lose its credibility anyway.
actually, I don't buy any of them. but I see your point, and I see how advertising is (relatively) important. it just doesn't strike me as the first thing to worry about, in the list of potential dangers of the formation of a single large linux conglomerate.
there's half of one, using perl instead of C, at my hacks page. if anyone wants to complete it...
your major concerns are about the *banner ads* of all things? well, I read slashdot for the stories, not for the ads.... what I sure hope is that we'll keep seeing stories about SGI's linux boxes and everything else that is interesting but competes with VA-Andover.
good ones. add Greg Egan too; /. readers should find him especially interesting.
I remember reading somewhere that the question with A.E. Van Vogt's novels was not how good they are, but how they can be so fascinating, intense, and ultimately, influential, despite all their shortcomings. I think that sums the situation up. AEVV was mentally more daring and ready to try with his mind than anyone else around, and that's including the Beat Generation people. That put him in contact with ideas and things that have turned out to be less than hoped (e.g General Semantics), or even downright nefarious (e.g Dianetics). But that's the price to pay for being open-minded and ready to experiment. We all owe Van Vogt a collective thanks for the mass of intense ideas that he acquainted us with. And, what the hell, Korzybski still makes an interesting read.
at least that's the case with CDs. I don't see any reason why DVDs would be any different.
command.com was a complete bore to disassemble... the int21 handler was way more interesting :)
personally, I support something like this:
bugtraq sure makes me wish one thing these days: that all MS related bugs were moved to their own mailing list! I mean, after all, there already is an NTBugtraq list, so why not rename it MSBugtraq and keep MS stuff out of Bugtraq itself? right now the situation is annoying: NT people need to read two lists, and Unix people read one but skip more than 50% of it.
the core problem is that consciousness is strongly tied to short-term memory, which works only one way (you remember the past, not the future -- except for Patrick Moraz's Future Memories). yet we experience time going forwards. the end result is that we have no clear intuitive picture of what it really means to duplicate a consciousness. assuming that the technology does make it possible unobtrusively, no-one doubts that the scanned guy won't feel his consciousness duplicating, or anything like that. he'll just continue to be himself, and if the body dies later, that consciousness will go with it.
if we take a purely external, descriptive stance, there is really no problem at all: the subjective feeling of "me" doesn't count (I think the big word for that is "epiphenomenon"), and you have two individual intelligences living in different universes, that happen to share a past up to a point. no problem. except that that doesn't make any distinction between the two, yet, if they scan me, I will still be mortal, while the scanned copy might well run forever, which is a mighty big difference.
I don't know of any theoretical framework out there that can make sense of this mess. religions that believe in some sort of soul don't solve the problem either, they just change the terms: now the difficulty is understanding if and how it can be duplicated, and if not, if and how there can be consciousness without it.
well, getting their banners and tracking cookies, but removing the cookie every day or more often, is a way of poisoning the db with lots of useless entries; however, these entries will eventually expire (no more hits in a long time => cookie must be lost; not associated to a real name profile => useless, expire it). a stronger way to poison the db would be to have a proxy that randomizes the content of the doubleclick cookie, within its usual syntax. depending on how their system is setup, you could either get ignored in most cases, or manage to assign your hits to other random people's profiles. but you'd need a lot of people doing that to have a significant impact, and most people just don't care enough. hell, *I* don't care enough either; I'm just happy to block them at /etc/hosts.
in general, cookies are OK, and quite useful, for short-lived browser/server interaction state keeping. There is no real need for long-term cookies; at worst you'll have to enter a password a few times more. And clearing your cookie file very effectively dissociates any further browsing from any profile doubleclick may have of you.
hey, Unlambda is Turing-complete... how about we rewrite the Linux kernel in that? :)
oh yeah, programming languages is a S-W field if there ever was one. hey, to me, everything's starting to look like a hashref :)
and no, I don't have a personal opinion on it, eiher way. Goverment&Binding (aka Principles&Params in its later versions) came forward enough to predict non-trivial language features that could be checked; I don't think minimalism has gone this far yet.