Typically they do 3- to 5-album contracts initially, which is more than most bands put out in their lifetime. When you get to be huge like say Metallica then you *can* do things however you want, but at that point what you want is to stick with the record companies, because their deals are going to be a LOT better at that point. (Simple supply and demand here. When you're nobodies, you have to compete with other bands in order to get a contract, so contracts will be the worst acceptable deal for the artists. When you're Metallica, *record companies* compete over *you*, meaning you will be able to demand the best contract still profitable to them. Those profits are obviously taken out of your share, but the record companies offer advantages that make up for those profits.)
Another driving force is this: if you have an initial shitty deal with your record company for a few albums, and your first album is a hit, you may be able to re-negotiate for a better deal, in exchange for a comittment of more albums. On the surface that doesn't seem wise on the part of bands -- but given that most bands don't stay successful all that long, it may be a good idea for them; you wouldn't want the record company to eat all the income from your 3 hit albums and leave you with just the income from the next two that tank. Or maybe you just need money to eat while you wait!
Of course, many bands, after becoming successful, *do* leave the commercial labels in spite of their economic interests. Nine Inch Nails and Courtney Love come to mind (incidentally, both have chimed in re the Napster issue, on opposite sides).
This is a cop-out, IMHO because RMS saw that if he pushed the issue, more people would realize that X11 is more free than the GPL.
Push the issue!? The RMS has posted an essay specifically on that issue[link] (GPL vs other free software licenses) and he addresses it all the time -- so much that he is criticized for doing so. He even touches on it in the GPL's preamble. How much more could he push it?
And, while it is true that the X11 license is "more free" than the GPL in that it has fewer restrictions, the issue is whether those freedoms are good. There are, after all, bad freedoms (such as the freedom to steal). Some might say that the freedom to steal is bad because it is incompatible with the freedom to own. RMS says the freedom to create proprietary software is bad because it is incompatible with the freedom (of the user) to control his software. Whether or not this is the case I would rather not address here -- but certainly the GPL cannot be dismissed through such an overly-simplistic application of the concept of freedom as "less restrictions == better". It is not incompatible with freedom to restrict a person's ability to infringe upon others' freedom (whether or not it is, in the case of the GPL, a good thing).
Maybe not straight out, but he sure seems to imply it often enough.
Such as when? Why don't you back your accusations with evidence? I've never seen any implication of the sort, and while I have not read everything he has said, I have read several essays by and interviews of RMS.
Stunts like demoting the LGPL to "lesser" status certainly don't argue against this perception.
DUH! "Lesser" does not mean less free. You are, of course, correct if you assert that RMS believes that the GPL is preferable to other free software licenses most of the time (however, you can see from his essay on the LGPL[link] that there are times when he finds the LGPL preferable) -- but that is not what you previously asserted. What you asserted was that RMS believes that non-GPL free software licenses are not free software licenses. This is patently false and given your access to the facts, libelous.
There are only 3 categories GPL-compatible, but non-GPL software: public domain, strict GPL subsets, and software with a specific GPL-surrender (a clause that says "you can also just forget any other terms and distribute it under the GPL"). Anything else contains terms which are incompatible with the GPL.
If I didn't see you post so much I'd conclude from this that you are a troll. Any license that allows relicensing without new restrictions -- e.g. the X11 license -- allows relicensing under the GPL. There doesn't need to be a specific GPL surrender, though obviously a surrender to relicensing must exist, as well as the lack of extra-GPL restrictions. However, I challenge you to find imposed by the X11 license "any further restrictions on the recipients' exercise of the rights granted herein" (to quote the GPL).
Oh right. I forgot about what we were originally discussing -- whether RMS considered Xfree to be non-free. What were you saying about that?
I don't think there are many people using nothing but GPL software. I also don't remember a top-priority effort to replace something evil and unfree like XFree86 (gasp! people are allowed to reuse the code without being required to distribute the source, it's not free!).
I don't know who started this all-too-popular misrepresentation of RMS's views, but it's most absurd and needs to stop. If you will look on the FSF web site on the matter, which you obviously have never done, you will see XFree86 specifically categorized as "GPL-compatible free software"[link]. I quote:
"
The X11 License This is a simple, permissive non-copyleft free software license, compatible with the GNU GPL. XFree86 uses the same license."
Further, you will see the FSF recommend the XFree86 license as an alternative to the GPL[link]. I quote again:
"if you want to release a program as non-copylefted free software, [...] please copy the license from XFree86."
RMS has never said anything along the lines of "non-GPL is not free", or even "non-copyleft is not free". See the FSF pages on different kinds of software licenses[link], see the FSF's page on specific licenses[link] and then stop propagating the straw man.
Ha! Intel may still be king, but it's taken a MAJOR hit in the past 2 years from AMD. AMD used to be insignificant in the PC CPU market -- now they are making the fastest PC CPU and are selling CPUs to several previously all-Intel *major* OEMs. Whether or not Intel recovers, they are *NOT* doing well right now -- at least compared to how they were doing 5, 10, or 15 years ago.
Of course, the number of people buying computers is going up and up, so Intel's revenue is still going to go up -- but they're no longer getting the *percentage* of the consumer CPU revenue that they once did.
The standards clearly allow for one to dynamically generate text on the
clientside using DHTML or XML/XSLT.
Sorry, Lynx ain't up to the standards snuff.
The standards do not REQUIRE that the browser be capable of interpreting DHTML. Standards organizations make very deliberate use of the words "MAY" vs "MUST" -- I think you need to pay a little more attention to that distinction.
Re:If you can clone an extinct animal...
on
TigerCloning
·
· Score: 1
You can make a bunch and release them in the wild, but their genetic diversity will be destroyed,
No, the genetic diversity is the only thing you are preserving. Whatever their behavior, the clones would be geneticly the same as the creatures they are replicated from.
"Genetic diversity" refers to the diversity of the gene pool, i.e. lots of different genes floating around, so that not everyone has genes that make them susceptible to the same disease (for instance).
You know something? I can't find any music on 8 track, and it's amost impossible to find a convenience store that sells leaded gas.
Oh please. The problem is not that lynx is obsolete. Lynx can display standards-compliant sites fine (on the rare occaisons that it can't, I agree that lynx is at fault -- but this is usually not the case). The problem is people violating the standards. And the standards will always require support for all-text interfaces, because the w3c has taken a firm stance on the support of blind users (who can't just upgrade from their text-to-speach browser to mozilla). w3.org/WAI.
(KJV, Jesus speaking) And other sheep I have, which are not of this fold: them also I must bring, and they shall hear my voice; and there shall be one fold, and one shepherd.
In context there's no indication whatsoever that it meant the other sheep were aliens. The other sheep were just people that weren't present at the time.
But then, maybe he's thinking of a different passage.
What to do? Just read
a few phrases back and distribute the source code for Qt! And reading over section 2, I find that works
distinct from the Program do not have to fall under the GPL.
You're right, they don't have to fall under the GPL, *BUT* they can't have any restrictions on redistribution, use, modification, or redistribution of modifications. QT's license restricts these (for commercial developers). Thus, it is not GPL compatible.
Re:use is not proof of necessity
on
KDE Strikes Back
·
· Score: 1
RMS didn't invent free software, he distorted the name Free Software for his own purposes ("it's not really free unless you place restrictions on how you can use it" - good logic, Mr. Stallman).
Uhh, first of all freedom when referred to positively does not generally mean the freedom to take other people's freedom away. Freedom does not connote the freedom to steal, etc, even though from a strict definition of ``freedom'' such a freedom to steal may be sensical (with the side-effect of removing the word's positive connotations). For you to suggest that restrictions and freedom are mutually exclusive is an equivocation of the term ``freedom''.
Second, your implication that RMS does not consider BSD- or Xfree-style licensed software (or public domain unlicensed software) to be free software is simply false. If you see the FSF's definition of ``free software'' on their web site (www.fsf.org), you will see that non-copylefted free software is still called free software (it is merely called non-copylefted free software). Remember that Debian, when it was the official distribution of the FSF, contained Xfree (and many other BSD-licensed packages). RMS (and the FSF) do not consider only GPL'd software to be free software, they merely endorse the GPL as the best way to create free software (and of course, to further their larger goal, the elimination of proprietary software).
Note: no opinions have been expressed anywhere in this post. Only (possibly false) statements of factual nature have been made. If you think your opinion differs, you are reading this post incorrectly, because it espouses no opinion.
Re:Unix _isn't_ an operating system . . .
on
Is UNIX An OS?
·
· Score: 1
It's quite telling that every major OS except Windoze will have some connection to Unix after Mac OS X comes out, be it a kernel, an API, or the standard Unix tools.
Actually, even Windows NT implements the POSIX API... but then nobody uses it, so I guess it doesn't really count.
Exactly. And I'm not underinformed, which is why I put "standards" in quotes. Users are more concerned with functionality than arbitrary
HTML "standards", which is why IE 5.x has been so wildly successful despite the bitchings of HTML standards purists.
"HTML standards purists" should not be "bitching" about IE 5. It is the most standards-compliant that can stay open for more than 15 minutes without crashing. But as far as functionality, I don't think IE is anything special. Speed is one of its major advantages, but other than that (and the above-mentioned standards compliance) it doesn't have much that NS doesn't have. The ability to block javascript/java from a blocklist (using internet zones) is nice, as well as some of the features that can be added through other Microsoft utilities, but these are things most users don't know how and wouldn't care to spend effort enabling anyway. So other than the above-mentioned, what does IE have feature-wize that's so special?
We expect them to want to know
this, so start with an NDA and then take them through the whole horse and pony show. I haven't found anyone that thought it was strange, and
only one person actually read it.
If I was running a business, I wouldn't want to hire someone who would sign a contract without reading it.
As a preempt to any OSS comments, remember, this isn't just a driver, it's an OpenGL implementation. It's also the highest quality GL implemenation on a consumer level card. It is simply too much to ask them to OSS it. They'd be giving away all the tweeks to the GL pipeline, not just register level info. Matrox is really hurting for a good OpenGL ICD. Is it really fair to ask NVIDIA to give them one?
Then why don't they just release hardware specs? Should we really support any hardware company who doesn't release specs? Maybe you don't mind hardware companies preventing free software drivers from being implemented but through reverse engineering, but you must understand that people who care about or use free software will not find this acceptable. I only use free software OSes, and nVidia has made it so that even if I have the time, talent, and will to help my OS of choice compete with proprietary OSes by supporting nVidia cards, I cannot do this without the severe disadvantage of having to reverse engineer. Of course, that's nVidia's prerogative, but it's not deserving of praise -- it's actually quite rude.
Maybe you will say that they have to do it in order to survive; that their competition will use hardware specs to reverse engineer and copy their hardware, and that they won't be able to do the same, making the situation unfair. Well, I want to make it so that companies have to release hardware specs in order to survive. That's why I won't buy closed hardware unless I absolutely must, and will urge others to do the same. If everyone does that, then all the hardware manufacturers will release specs, and it will hurt them all equally and thus not at all.
Now that's bad enough, but on top of that, the Debian system makes hell on users who try to install other software. If you happened to download a.tar.gz file and do the make / install yourself, the Debian installer had no way to let you update the dependencies. So next time you went and grabbed a.deb release, you'd end up overwriting the newer libs from your.tar.gz file. Simply brain dead. The Debian team needs to realize that people MIGHT just want to install software OTHER than what they provide. What a concept.
If you don't want dpkg to manage a package, you put the package on "hold". If you want to ignore dependencies, you use dpkg --ignore-depends=package
So like I said, read the docs before you criticize. Just because you don't know how to do something, does not mean Debian does not let you do it.
Of course, there is a lot more you can do with ipchains than that. I recommend you block all ports below 1024, except for the ones you need, block 6000-6010, and go ahead and block any GNOME ports if you don't know what they're for.
A more radical policy which many people use, is to block *all* incoming TCP connections, and UDP packets, *except* for ones explicitly allowed. You can do that too, but it may cause some problems (it won't cause any problems that wouldn't also be caused by using IP MASQ. In fact, this would be pretty much the functional equivalent of IP MASQ, but with only one computer.)
Kernel 2.4 will change the entire way networking is adminstered, btw, so if you're using 2.4 those docs will be worthless. But everything you can do in 2.2 you can do in 2.4, so the same basic strategy applies.
The real flaw of Freenet, IMHO, isn't the potential for revisionism, it's the idea that only
popular information is valuable. That might make sense in a market context, but it doesn't
really have any place in an intellectual context if by "intellectual" you mean to imply a
search for truth instead just popularity. Moreover, it is often the most revolutionary,
cutting-edge, ahead-of-their-time ideas that are the most unpopular.
It's not the idea that has to be popular, it's the actual document. By `popular' all is meant is in-demand. If people want to read something, then it will stay on the FREENET -- whether or not its ideas are popular or not. Take, for instance, that fellow who wrote the article falsely claiming BUGTRAQ listed more bugs for GNU/Linux than Windows, and fallaciously using this mis-fact to conclude that GNU/Linux is less secure. Among the slashdot crowd, that fellow was not popular in the sense you intend, nor were his ideas. But his article was popular -- that is, it was in demand.
At one time, the ideas of democracy and freedom of speech were extremely unpopular ideas. In some places, they still are. Freenet-like systems would not have helped the rise of democracy very much.
That's not true at all. While the ideas may not have been popular, obviously the documents extolling those ideas must have been in demand enough to get read.
Mind you, it's great to see that popular ideas will be more resistant to government/corporate suppression, but they already were. It is ideas held by small minorities that are the most vulnerable.
Minority opinions will have just as much exposure, as long as people are willing to read them. A solution to what I agree is the larger problem -- not the ability to speak freely, but the ability to be heard -- is not provided by FREENET. But it's not at all prevented, either.
That said, I think the primary effect of FREENET, if it is successful, will not be the ability of minority ideas to evade censorship, but the de facto eradication of copyrights on digital media. (And that fact makes its success all that less probable, of course.)
Approval voting is flawed. Rather than merely a pass-fail, what is necessary is a ranking of the voters. Approval voting still allows for tactical voting. Let's say you want Gore, but you want Nader more. How do you express this desire? You don't want Bush to win, but should the election come down to Gore vs. Nader, you want to be able to choose Nader. But the only way you can do this is by approving Nader, but not Gore. This increases the chances that Bush will win. Your only way to move your vote towards the left is to risk a win by the right! Solving this is only possible in a system not based on pass/fail approval, but rather based on ranking -- which is the minimum amount of data required of voters for a fair vote. Condorcet is one method of analysing that data, but that's another issue. The point is, unless you can explicitly vote e.g. Nader #1, MacReynolds #2, Gore #3, then the government may not have enough information to know what candidate is most preferred by the public. It will not, for instance, know whether you prefer Gore over than Nader. If Gore and Nader are the top two candidates, not knowing which you prefer means trouble!
For more info on Condorcet, which solves the problems in approval voting, see the Voting Systems FAQ. [LINK]
In Condorcet, you never incur a greater risk by voting your true preference. With FPTP (the current system) you do (by "throwing away" your vote on a third party, increasing the chances of victory for the greater of two republicrat evils). With approval voting you do, too (by specifying Nader over Gore and risking Bush victory, or specifying both over Bush and risking Gore victory when you prefer Nader).
Unfortunately, Condorcet often fails to produce a winner. There are tie-breakers for such a situation, though. One of them is an alternative to Condorcet that almost always produces the Condorcet winner if there is one -- IRV (instant run-off voting -- aka alternative vote or AV). Interestingly enough, it is the voting system used by the American Political Science Association -- an organization of political science professors -- to select the president of the association. However, the best system imaginable, from a game theoretic point of view, is Condorcet, with IRV only as a tie-breaker. In those situations where IRV fails to match Condorcet, Condorcet is clearly the preferable choice. In those situations where Condorcet ends in tie, the IRV result always appears equitable. See the Voting Systems FAQ for more info (they refer to IRV as "AV", which I didn't use here to avoid confusion with "approval voting").
There are flaws necessarily inherent in all voting systems (inherent in the very concept of voting), but those appear to be the only flaws from which Condorcet suffers.
Harry Browne is the Libertarian Party candidate. Among other things, if elected he promises to abolish the federal income tax and auction off federal property to pay the national debt. He qualified for matching campaign funds this election, but turned them down. He'll be on the ballot in all 50 states; no other third party candidate will. http://www.harrybrowne2000.org/
(I don't endorse Browne, btw, but I'd infinitely prefer him over a republicrat.)
Surely there must be/. regulars who could work out how to build a very minimalist system for ourselves. What i'd like is something like this, just with the 16meg of flash with ethernet & serial
LART has 32MB DRAM, 4MB Flash ROM, serial, 10BaseT Ethernet (on separate board), PS/2 mouse & keyboard, IDE (44pin laptop IDE) (all on a single separate board, the "Kitchen Sink Board"), and more. Check out this picture: http://www.lart.tudelft.nl/gallery/vt2 20.jpg.
Well obviously... But that's impossible to defend against. You need to find the source of the DoS, and then contact them, and get them to find out how the attacker is using their box.
Slashdot Mirror
Another driving force is this: if you have an initial shitty deal with your record company for a few albums, and your first album is a hit, you may be able to re-negotiate for a better deal, in exchange for a comittment of more albums. On the surface that doesn't seem wise on the part of bands -- but given that most bands don't stay successful all that long, it may be a good idea for them; you wouldn't want the record company to eat all the income from your 3 hit albums and leave you with just the income from the next two that tank. Or maybe you just need money to eat while you wait!
Of course, many bands, after becoming successful, *do* leave the commercial labels in spite of their economic interests. Nine Inch Nails and Courtney Love come to mind (incidentally, both have chimed in re the Napster issue, on opposite sides).
And, while it is true that the X11 license is "more free" than the GPL in that it has fewer restrictions, the issue is whether those freedoms are good. There are, after all, bad freedoms (such as the freedom to steal). Some might say that the freedom to steal is bad because it is incompatible with the freedom to own. RMS says the freedom to create proprietary software is bad because it is incompatible with the freedom (of the user) to control his software. Whether or not this is the case I would rather not address here -- but certainly the GPL cannot be dismissed through such an overly-simplistic application of the concept of freedom as "less restrictions == better". It is not incompatible with freedom to restrict a person's ability to infringe upon others' freedom (whether or not it is, in the case of the GPL, a good thing).
Such as when? Why don't you back your accusations with evidence? I've never seen any implication of the sort, and while I have not read everything he has said, I have read several essays by and interviews of RMS. DUH! "Lesser" does not mean less free. You are, of course, correct if you assert that RMS believes that the GPL is preferable to other free software licenses most of the time (however, you can see from his essay on the LGPL[link] that there are times when he finds the LGPL preferable) -- but that is not what you previously asserted. What you asserted was that RMS believes that non-GPL free software licenses are not free software licenses. This is patently false and given your access to the facts, libelous. If I didn't see you post so much I'd conclude from this that you are a troll. Any license that allows relicensing without new restrictions -- e.g. the X11 license -- allows relicensing under the GPL. There doesn't need to be a specific GPL surrender, though obviously a surrender to relicensing must exist, as well as the lack of extra-GPL restrictions. However, I challenge you to find imposed by the X11 license "any further restrictions on the recipients' exercise of the rights granted herein" (to quote the GPL).Oh right. I forgot about what we were originally discussing -- whether RMS considered Xfree to be non-free. What were you saying about that?
Of course, the number of people buying computers is going up and up, so Intel's revenue is still going to go up -- but they're no longer getting the *percentage* of the consumer CPU revenue that they once did.
But then, maybe he's thinking of a different passage.
Second, your implication that RMS does not consider BSD- or Xfree-style licensed software (or public domain unlicensed software) to be free software is simply false. If you see the FSF's definition of ``free software'' on their web site (www.fsf.org), you will see that non-copylefted free software is still called free software (it is merely called non-copylefted free software). Remember that Debian, when it was the official distribution of the FSF, contained Xfree (and many other BSD-licensed packages). RMS (and the FSF) do not consider only GPL'd software to be free software, they merely endorse the GPL as the best way to create free software (and of course, to further their larger goal, the elimination of proprietary software).
Note: no opinions have been expressed anywhere in this post. Only (possibly false) statements of factual nature have been made. If you think your opinion differs, you are reading this post incorrectly, because it espouses no opinion.
Stability is not a "feature", at least in the usual sense of the word.
X11 listens on 6000 for display :0, 6001 for display :1, etc.
Maybe you will say that they have to do it in order to survive; that their competition will use hardware specs to reverse engineer and copy their hardware, and that they won't be able to do the same, making the situation unfair. Well, I want to make it so that companies have to release hardware specs in order to survive. That's why I won't buy closed hardware unless I absolutely must, and will urge others to do the same. If everyone does that, then all the hardware manufacturers will release specs, and it will hurt them all equally and thus not at all.
So like I said, read the docs before you criticize. Just because you don't know how to do something, does not mean Debian does not let you do it.
ipchains --insert input --destination-port 1030:1040 --jump DENY
Of course, there is a lot more you can do with ipchains than that. I recommend you block all ports below 1024, except for the ones you need, block 6000-6010, and go ahead and block any GNOME ports if you don't know what they're for.
A more radical policy which many people use, is to block *all* incoming TCP connections, and UDP packets, *except* for ones explicitly allowed. You can do that too, but it may cause some problems (it won't cause any problems that wouldn't also be caused by using IP MASQ. In fact, this would be pretty much the functional equivalent of IP MASQ, but with only one computer.)
More info: ipchains(8), IPCHAINS-HOWTO.
Kernel 2.4 will change the entire way networking is adminstered, btw, so if you're using 2.4 those docs will be worthless. But everything you can do in 2.2 you can do in 2.4, so the same basic strategy applies.
That said, I think the primary effect of FREENET, if it is successful, will not be the ability of minority ideas to evade censorship, but the de facto eradication of copyrights on digital media. (And that fact makes its success all that less probable, of course.)
For more info on Condorcet, which solves the problems in approval voting, see the Voting Systems FAQ. [LINK]
In Condorcet, you never incur a greater risk by voting your true preference. With FPTP (the current system) you do (by "throwing away" your vote on a third party, increasing the chances of victory for the greater of two republicrat evils). With approval voting you do, too (by specifying Nader over Gore and risking Bush victory, or specifying both over Bush and risking Gore victory when you prefer Nader).
Unfortunately, Condorcet often fails to produce a winner. There are tie-breakers for such a situation, though. One of them is an alternative to Condorcet that almost always produces the Condorcet winner if there is one -- IRV (instant run-off voting -- aka alternative vote or AV). Interestingly enough, it is the voting system used by the American Political Science Association -- an organization of political science professors -- to select the president of the association. However, the best system imaginable, from a game theoretic point of view, is Condorcet, with IRV only as a tie-breaker. In those situations where IRV fails to match Condorcet, Condorcet is clearly the preferable choice. In those situations where Condorcet ends in tie, the IRV result always appears equitable. See the Voting Systems FAQ for more info (they refer to IRV as "AV", which I didn't use here to avoid confusion with "approval voting").
There are flaws necessarily inherent in all voting systems (inherent in the very concept of voting), but those appear to be the only flaws from which Condorcet suffers.
(I don't endorse Browne, btw, but I'd infinitely prefer him over a republicrat.)
I think you forgot your link. But here's one: fairvote.org. Also see Open Directory Project's listing on voting systems and the Voting Systems FAQ.
LART has 32MB DRAM, 4MB Flash ROM, serial, 10BaseT Ethernet (on separate board), PS/2 mouse & keyboard, IDE (44pin laptop IDE) (all on a single separate board, the "Kitchen Sink Board"), and more. Check out this picture: http://www.lart.tudelft.nl/gallery/vt2 20.jpg.
Well obviously... But that's impossible to defend against. You need to find the source of the DoS, and then contact them, and get them to find out how the attacker is using their box.