Shared source and open source are not mutually exclusive. According to the only document I can find on the subject, the Wikipedia page (IronPython's webpage sucks) the license is the CPL, an IBM-authored license which is incompatible with the GPL but is nonetheless considered a Free Software license by the FSF, and Open Source by the OSI.
Despite it's incompatibility with their own GPL, the FSF sounds like they actually rather like it:
This is a free software license but it is incompatible with the GPL.
The Common Public License is incompatible with the GPL because it has various specific requirements that are not in the GPL.
For example, it requires certain patent licenses be given that the GPL does not require. (We don't think those patent license requirements are inherently a bad idea, but nonetheless they are incompatible with the GNU GPL.)
So I really wouldn't worry about the "shared source" write-up. It's an unusual choice of license, but it is considered Free Software and Open Source, the patent license requirements are actually fairly positive from the point of view of protections from Microsoft itself. Microsoft have chosen the same license in the past when releasing other code they want to be seen as completely open.
No, actually, you've got it exactly backwards -- they haven't segmented tastes by determining how much a customer is willing to spend, they've instead examined customers willing to spend more and have attempted to build products to their tastes.
Nope. Putting the words in a different order doesn't make it backward, 2=1+1 isn't wrong because 1+1=2. I said Apple believes there is a corrolation between "ability to spend" and "needs and tastes of a consumer". You've just said I'm wrong and then repeated that principle, claiming that Apple believes (and apparently you believe) there's a corrolation.
This is not the case. There is no corrolation. There is no legitimate reason for believing that someone with less than $1,000 to spend wants a closed box that needs to connect to an external monitor. There's no reason for believing that someone with over $2,000 wants an expandable box that hooks up to an external monitor. There's no reason for believing that someone who has somewhere between $1,000 and $2,000 wants an unexpandable box with a built-in monitor.
Apple's premise is ridiculous. It works for them at the moment because they control such a small percentage of the market that any product will sell as long as it covers some group of people who exist. Undoubtedly there are some people who want all-in-ones who have $1,500 to spend. Undoubtedly there are some people who want expandable Macs who have $2,500 to spend. Their market isn't saturated. But if they don't produce products that cover other market needs, their market never will be - not by them, anyway.
I don't think a serious argument can be made for Windows 95 being superior to OS/2 3.0, except in the area of available software.
As someone who used both I can honestly say that the serious argument can be made. OS/2 3.0 lacked native networking (it was frequently shipped with a dial-up networking stack, but that's literally what it was, if you wanted a more generic stack that supposed Ethernet and other modern networking systems, you had to buy it as a bolt-on from IBM.)
OS/2's UI was funky too, with a lot of aspects being designed of the "We can do this therefore we should" variety. Generally, for all of its faults, Windows 95 was pretty intuitive and the right set of UI features were provided in a form most users would find easy to find. OS/2, on the other hand... well, if you liked dragging colours and fonts to buttons the look of which you wanted to change, temporarily, to something else, well it kind of worked for you, but who'd want to? On that point, Windows 95 had an extremely clean, AmigaOS 2.0/NEXTSTEP, look, whereas OS/2 looked like a cross between Motif and Windows 2.0. (There's a reason for that.)
OS/2 3.0 was a very good alternative to Windows 3.1, but OS/2 3.x was significantly less usable than Windows 95 (software, remember, wasn't an issue in 1995, as both ran Windows 3.1 software, and there wasn't really any more Win32 than OS/2 software at that point.) You had to wait for OS/2 4.0 before you got a clean UI fronting a feature complete operating system.
That's correct. Normally Microsoft's free new features are available seperately rather than part of service packs, usually for a variety of versions of Windows some of which aren't even sold any more.
You may have to upgrade to the latest pay-for version of Mac OS X to use the latest Safari or Xcode, but Microsoft will make IE,.NET, Visual Studio Express, and other useful tools for Windows for free. And obviously, with things like virtual desktops (Spaces) and automatic back-ups (Time Machine), Apple is, to some extent, still adding features that have been available, from Microsoft, for free, either as part of released operating systems or as a free bolt-on pack, for many years.
Ignoring any issues about offensiveness or whatever, that's not the problem with it. The problem is that it's easily broken.
How do you break it? Easy. Just pick a random number between one and the number of options you have. For a three option CAPTCHA, you have a one-in-three chance of getting through. You're a spammer remember, so these odds do not deter you, all you have to do is run your automated script three times and you'll be close to sending out the same number of spamvertisements as you would have sent without the CAPTCHA.
Realistically no multiple choice system, as advocated by a number of posters here, will succeed unless it has so many choices that it's improbable a real user will be able to use the system without issues.
CAPTCHAs are a bad idea in general. Yet again they're a poor, unwieldy, temporary "solution" to a problem the inventors barely understand that causes more problems than it fixes. Like 99% of anti-spam solutions. The only thing worse than a CAPTCHA is what'll replace them.
The Cube was released at a price close to Apple's PowerMac range. The Cube certainly wasn't a "Headless iMac" in terms of price, it was actually more expensive than most iMacs. At that time, both the PowerMac and the Cube were priced in the mid $1,000s.
The problem with the Cube is that it emphatically wasn't what was being asked for. People wanted a low cost, expandable, headless Mac, not a high cost, stylish headless Mac with poor expandability. Apple currently doesn't cater to that market segment. It's made the deliberate decision not to. The Mac mini is the closest thing they have to such a thing, and it, frankly, is a poor choice, especially right now.
Apple's primary concern is about profits, and with such a small percentage of the market, they can get away with ignoring large consumer segments while appearing to do very well. The fact is though that there's a fundamental flaw in their marketing system. They believe that you can segment a market's tastes and requirements through how much they're prepared to spend. As such if someone can afford an iMac, the Apple belief is that they want an integrated, non-expandable, Mac. If someone can afford a PowerMac/Mac Pro, then they must want an expandable, headless, machine (or a laptop.)
No other company does this. I'm bewildered by the attitude to be honest. This kind of thing would kill them if they had a serious percentage of the market.
Or can you merge GPL code and non-GPL code in a single codebase?
Can you do that if you do not agree to the GPL?
The GPL isn't what's restricting you from doing that. The GPL is making some acts legal that otherwise wouldn't have been. The fact it doesn't allow you, for example, to murder Eric Raymond doesn't mean the GPL is restricting you from doing something.
I wonder how many people realise that the same is true in the US.
Software patents were never explicitly legalized in the US, instead a court decision ruled that a computer running a program was patentable. Algorithms are still, theoretically, unpatentable in US patent law, but the law has been, essentially, "creatively interpretted" by the courts (without being slapped down by Congress) to include pretty much anything.
So if you're relying upon EU rules about algorithms and mathematical truths to hold back the tide, you're playing a strategy that's already lost once.
I know what you're trying to say, but that's kind of wrong because it presumes all cases of "testing" are of the form "The GPL is invalid, therefore I can do what I want." This is not true. Not all cases are of that form.
If someone interprets the GPL as meaning it's ok to release a proprietary extension to a GPL'd program using some type of unusual linkage mechanism, then yes, it may be necessary to test this in court depending on the beliefs of the person who wrote the code they're licensing. The person who's interpretting the GPL that way isn't arguing that the GPL is invalid, and that they don't have a license, they're arguing that they do have a license, and it means something other than that the licensor thinks it means.
Whether that's applicable in this case is for someone who read TFA to answer.
But as someone pointed out elsewhere in this thread, Debian includes other non-GPL compatible licensed software in its distribution like Apache, openssl, PHP for a few examples. Why be so specific about CDDL incompatibilty? Or is this just an issue about a clash of personalities?
Reread the parent. He said that a project that has both code licensed only under the GPL and code only licensed with {a license incompatible with the GPL} cannot be in Debian, because it would be illegal to distribute.
This isn't about putting Apache and GNU C in the same distribution. It's about putting filemanager.c and documentview.c in the same binary when filemanager.c is licensed under the XGL, and documentview.c is licensed under the XGL-incompatible YGL. That's the core of the problem here.
Apple stopped doing that when they went over to 1.44M disk drives, back in the late eighties. The sectors are in the same places as for PC disks, it's just the file system is different.
400k and 800k floppies work the way you describe, but you're unlikely to come across them. (Well, you're unlikely to come across Mac floppies at all, but you're even less likely to come across 400k and 800k disks)
And it wasn't an 'accident', the ray killed him quite deliberately.
Really? Stingrays aren't really aggressive or defensive, they generally swim away if threatened. The sting is controlled by something akin to reflexes rather than deliberate thought.
My Dish Network 625, which supposedly, according to TiVo advocates, is nothing more than a poor rip-off of TiVo's totally original wouldn't-have-been-invented-anyway technology does, in fact, have a commercial skip feature, which doesn't require any registry hacks (or secret key presses) to enable.
I'm kind of baffled by it. Why aren't the networks pissed at Dish Network? How is TiVo way better than Dish as TiVo advocates claim? I wonder if the "secret" commercial skip thing is itself viral marketing (Where's the fun in talking about a normal feature? A feature that's hidden because "The Man" doesn't want you to have it, on the other hand, is worth sharing with your friends...)?
When you say "lots of small processes", do you mean "I'm generally using a few hundred megs, but I'm running a few hundred processes so that's pretty good" (ie you're using a significant percentage of your memory, but nowhere near enough to worry about getting more) or do you mean "I run 30 or 40 processes each of which is maybe a meg at most"? (ie virtually all your memory is free, the amount you added was overkill)
I ask because it's a peculiarity of Linux, and some other VM implementations, that it's frequently better to have swap in the first situation than not have it. Why? Because of disk caching. VM allows the kernel to swap out idle processes (or idle memory from processes anyway) in favour of caching the sectors of recently processed files. This can make for dramatic speed-ups. For a desktop workstation, this can be good and can be bad, a search through a bunch of files that make up a large part of the disk can cause every major application to crawl as they're reloaded into memory.
But in your example, (a) you're not using it as a workstation, you're using it as a server (which tends to imply it'll be much more file intensive) and (b) you're running small processes, not large ones, so the time to recovery for each process will generally be much quicker.
If you're barely using your memory, then ignore this advice. But if you're using a significant percentage of your memory, and you're using your system as a server, the probability is extremely high that you'd benefit from virtual memory.
That doesn't match my experience. As an example, most of the programmers in my department are avid game players. Ages vary from early twenties to mid forties.
And most of them have actually told me that games are their major reason for not switching to GNU/Linux. (Yes, "solutions" like WineX or whatever its called today exist, but who wants to subscribe to a service if they don't need to.)
I agree. And I find it pretty sad I've had to scroll down half way through the first page to find someone who isn't posting "OMG!!!1! Free Music! Finally teh RIAA si giving us what we want!"
I am not a freeloader. I do not use Kazaa and every actual freeloader (as opposed to occasional innocent who's been caught up in the thing) who's paying $3,500 to the RIAA right now to settle lawsuit is getting what they deserve, as far as I'm concerned. I want to donate towards the costs of producing music. I'm happy to be required to buy a CD or MP3 in order to have the right to listen to that piece of music. But I do want certain reasonable rights once I've contributed to the funding, including content shifting and conversion.
This isn't giving me what I want. It's responding to the "Music needs to be free - as in freeloader" mob rather than the "Content needs more freedom" group. It's the logical outcome of the Napster mess, not the idealistic nonsense peddled by Napster and Kazaa's apologists.
Do you actually know what UMA is? I have absolutely no idea why this complete confusion of ideas keeps coming up. I've even read the once excellent Ars Technica claim UMA is something carriers are scared of. Right now, the only major carriers that might be scared of it in the US are Verizon and Sprint. Because they can't use it. It's a GSM technology.
UMA does not cut into an operator's revenue stream. It frees up revenue because the operator is not having to put up towers to get coverage and capacity for every single building in the world. If YOU, the user, save money, it'll only be because the operator is giving you discounts for using UMA, not because you're sticking it to the man by using it, somehow bypassing the carrier. Far from it. You're using the carrier either way.
UMA is not "I can bypass the cellphone company to make free calls", it's "I can route the last mile of my calls through either the radio waves to a tower or via the Internet to a gateway at my carrier, either way getting to my carrier who'll then route the call as necessary." It's a great technology, but what makes it great is that it means that people can make coverage where they currently have blackspots.
What's confused some people is they've read all this crap about Skype phones, and think that UMA is this. It isn't. It's GSM routed over the Internet. Skype phones are something else entirely.
Other people are confused because they've heard it's VoIP. VoIP does not mean "Cheap ways to bypass the phone companies", it's a just a name given to any form of two-way voice traffic routed over IP packets. Just because using Vonage over cable is "sticking it" to AT&T&T doesn't mean that all forms of VoIP are.
This is why T-Mobile and Cingular are members of the UMA consortium and are planning to roll it out here in the US. Yes, they are. Yes, they've made announcements to that effect. It may make calls cheaper. More importantly though it'll make calling more reliable. No more blackspots in the kitchen. Nice.
Good lord. You mean ESR is insulting and alienates the very people he needs to get on his side?
I'm glad he's no longer the official spokesman for the Open Source movement any more (he has some honory position but nothing important.) I think what he had to say was kind of interesting (and said so when this article was published the first time around (albeit with a misleading title) but it needs to be put forward by people a little more knowledgable and with better communication skills than himself.
Even the FSF doesn't entirely disagree with the stance, despite the number of comments here suggesting "compromise" is some great chasm between the FSF and OSI. The FSF created the LGPL and while they officially discourage its use, it's still very much a live license that's seen as appropriate in many cases.
So: GNU/Linux users are self-absorbed geeks who learn Klingon and attend science fiction cons. You can't get iPods to work with GNU/Linux. GNU/Linux needs to compromise and often allow proprietary software. Now is the time to strike, because with the move to 64 bit operating systems, everyone will throw out what they already have and buy entirely incompatible 64 bit systems. Great work there ESR, you are truly are relevent in today's age, hip and with-it. Knowin' the trends.
Here's the funny thing. Everyone's acting like "compromise" is a big thing. And it isn't. We all want to be free, but few people haven't compromised on the Free Software thing at one point or another no matter how strong our beliefs in terms of what we use. The first time this was posted, Slashdot went one step further and claimed ESR was an advocate for proprietary software, which was heavily misleading, if not an outright lie.
What ought to be questioned is what never is. The insults. The snide comments. The pro-factionalization. The implication that he's promoting something new when he's repeating what almost everyone in the community has been doing from day #1.
Oh c'mon! WTF is wrong with you all? Is the world so black and white in Slashdot that if someone protests about being punished too hard, they can only be advocating no punishment at all?
The (ex)Apple employees are protesting that they came clean and yet endured the same punishment they'd have endured if they had not come forward but been caught anyway. The complaint is not that they were punished at all, it's that the punishment was excessive and gives nobody any incentive to be honest.
And they have a point. And this not about murder, where arguably the action is so severe that the appropriate punishment should always be dealt, it's about a case of copyright infringement. Yes, there's room for Apple to take a more lenient line with truth tellers than with those who lie.
Especially when given the case is ultimately about whether an employee can be trusted with the company's proprietary inside information, the issue of whether they lied or not in an investigation is actually relevent.
Apple has arguably over-reacted. And whether it did or not, it has most certainly cut off its own nose to spite the face of others. Firing is an expensive act. Apple can expect to lose the productivity the fired employee would have given to the company during the time it recruits and trains the replacement, and recruiting is hardly cheap either. Further, it has made its own future investigations harder because it will not get the cooperation of employees who see themselves as ultimately loyal.
Apple can hire and fire whoever they want, for whatever (legal) reason. But that doesn't make this anything other than, at face value, assuming there's not more to it than TFA, a dumb decision. And certainly, the logic Slashdotters promote of "IF THEY HATE THERE PUNISHMANT, TEHY MUST FINK BEENG PUNICHED IS RONG!!1" is utterly irrelevent and idiotic.
A lot of the "holes" in recent Windows have to do with design problems. The problem is it's one thing to go around and fix coding bugs, it's another to fix design issues because programs are built around designs.
That said, Vista isn't the rewrite it was originally intended to be.
Each generation of Intel integrated graphics system has gotten steadily better, and supposedly the next generation, present in the 965, uses a design (the X3000) that can hold its own.
My guess is that Intel will probably end up sweeping the market. If this kind of thing is being incorporated into computers for "free", then the number of people who "need" better graphics is going to drop dramatically. Tech people will go for Intel because it's open and "good enough". Games developers will optimize for it because it's open, good enough, and deployed everywhere.
In that kind of market, I suspect AMD will want to use ATI for:
1. Developing niche cards for certain industries.
2. Developing designs for console manufacturers.
3. Developing designs for integration into motherboard chipsets (like the X200) to support AMD CPUs.
In that playing field, AMD isn't really going to care about ATI's behaviour concerning older chipsets. For the niche cards, ATI is probably going to develop one-off specific drivers. For the consoles, the manufacturers will probably develop their own drivers, and for the integrated chipset designs, they'll probably open up a little.
I also have an X1300 laden laptop (a Thinkpad T60) and have tested it extensively under Debian Stable, Ubuntu, and RedHat Core 5.
The drivers are shit. The machine will not even boot into X11 if the power is unplugged. Running Unreal Tournament doesn't work (you get massive lag between keyboard/mouse events and things happening on screen.) Unreal Tournament 2004 "works", but frequently with bizarre artifacts that make the display unwatchable. Full on crashes happen every few days (triggered usually by watching a movie in MPlayer.)
This has nothing to do with "F/OSS zealots", and you blaming them (in response to people saying the nVidia drivers, which are equally closed, are better no less!) shows you have an agenda. The fact is I'll never buy ATI again after this experience, and I'm recommending others avoid them too. I prefer open source, but ultimately I want something that works. If nVidia can provide that, go with them.
I'm still having difficulty with this argument. It looks like the argument is "Does it hurt - that is, cause unhappiness in - the child? If so, it's a form of abuse."
Which makes discipline a little difficult if all options that in any way "hurt" the child are off limits. It also ignores the real world costs of providing many of these items, both monetary and otherwise.
I can handle the argument that needless or extreme punishment is abusive, and that acts of violence, intended to be disciplinary or not, are abusive. But this is not a world where we can legitimately give everything to anyone whenever they want. And sometimes, to do so even when we can would be inappropriate and would hurt the receiver in the long term.
You don't have to file trademarks defensively. Well, let me rephrase that. If you're building a product that has an entirely new name, never before used in relation to your market area, then yeah, you need to file a trademark to ensure nobody else starts making things with the same name. Though it's arguable that's "defensive". You're defending your product against knock-offs, but you're not trademarking it for the same reason as someone might defensively file a patent. That is, you're not doing it so that if someone sues you for trademark infringement, you can sue them right back.
Defensive patents are an entirely different area. People file patents so that if someone sues them, they have some potential ammunition against them. For example, ACME Sandwiches sues you for violating their Cheese Between Two Slices of Bread patent, you in turn can see that they have a wide range of sandwiches, some of which violate your own patents, you can sue them back for violating your "Peanut butter between two slices of Ryvita" patent.
O'Reilly isn't really being "defensive" here. He's taking words that relate to things he's doing conferences on, and trademarking them. The net effect is to make it harder to come up with relevent terms to describe competing conferences. That's really being offensive, not defensive. It's not a good thing.
It does actually appear to have been a common term to describe, erm, websites: Examples. Of course, there weren't many in 1994.
There may not have been a massive number of mentions, the count is in the hundreds not thousands, but that appears to be the time the term started to become popular. So it looks like O'Reilly jumped on a technology that was emerging at that point, and decided to trademark a term already in use by those already using the technology.
Slashdot Mirror
Shared source and open source are not mutually exclusive. According to the only document I can find on the subject, the Wikipedia page (IronPython's webpage sucks) the license is the CPL, an IBM-authored license which is incompatible with the GPL but is nonetheless considered a Free Software license by the FSF, and Open Source by the OSI.
Despite it's incompatibility with their own GPL, the FSF sounds like they actually rather like it:
So I really wouldn't worry about the "shared source" write-up. It's an unusual choice of license, but it is considered Free Software and Open Source, the patent license requirements are actually fairly positive from the point of view of protections from Microsoft itself. Microsoft have chosen the same license in the past when releasing other code they want to be seen as completely open.
Nope. Putting the words in a different order doesn't make it backward, 2=1+1 isn't wrong because 1+1=2. I said Apple believes there is a corrolation between "ability to spend" and "needs and tastes of a consumer". You've just said I'm wrong and then repeated that principle, claiming that Apple believes (and apparently you believe) there's a corrolation.
This is not the case. There is no corrolation. There is no legitimate reason for believing that someone with less than $1,000 to spend wants a closed box that needs to connect to an external monitor. There's no reason for believing that someone with over $2,000 wants an expandable box that hooks up to an external monitor. There's no reason for believing that someone who has somewhere between $1,000 and $2,000 wants an unexpandable box with a built-in monitor.
Apple's premise is ridiculous. It works for them at the moment because they control such a small percentage of the market that any product will sell as long as it covers some group of people who exist. Undoubtedly there are some people who want all-in-ones who have $1,500 to spend. Undoubtedly there are some people who want expandable Macs who have $2,500 to spend. Their market isn't saturated. But if they don't produce products that cover other market needs, their market never will be - not by them, anyway.
As someone who used both I can honestly say that the serious argument can be made. OS/2 3.0 lacked native networking (it was frequently shipped with a dial-up networking stack, but that's literally what it was, if you wanted a more generic stack that supposed Ethernet and other modern networking systems, you had to buy it as a bolt-on from IBM.)
OS/2's UI was funky too, with a lot of aspects being designed of the "We can do this therefore we should" variety. Generally, for all of its faults, Windows 95 was pretty intuitive and the right set of UI features were provided in a form most users would find easy to find. OS/2, on the other hand... well, if you liked dragging colours and fonts to buttons the look of which you wanted to change, temporarily, to something else, well it kind of worked for you, but who'd want to? On that point, Windows 95 had an extremely clean, AmigaOS 2.0/NEXTSTEP, look, whereas OS/2 looked like a cross between Motif and Windows 2.0. (There's a reason for that.)
OS/2 3.0 was a very good alternative to Windows 3.1, but OS/2 3.x was significantly less usable than Windows 95 (software, remember, wasn't an issue in 1995, as both ran Windows 3.1 software, and there wasn't really any more Win32 than OS/2 software at that point.) You had to wait for OS/2 4.0 before you got a clean UI fronting a feature complete operating system.
That's correct. Normally Microsoft's free new features are available seperately rather than part of service packs, usually for a variety of versions of Windows some of which aren't even sold any more.
You may have to upgrade to the latest pay-for version of Mac OS X to use the latest Safari or Xcode, but Microsoft will make IE, .NET, Visual Studio Express, and other useful tools for Windows for free. And obviously, with things like virtual desktops (Spaces) and automatic back-ups (Time Machine), Apple is, to some extent, still adding features that have been available, from Microsoft, for free, either as part of released operating systems or as a free bolt-on pack, for many years.
Ignoring any issues about offensiveness or whatever, that's not the problem with it. The problem is that it's easily broken.
How do you break it? Easy. Just pick a random number between one and the number of options you have. For a three option CAPTCHA, you have a one-in-three chance of getting through. You're a spammer remember, so these odds do not deter you, all you have to do is run your automated script three times and you'll be close to sending out the same number of spamvertisements as you would have sent without the CAPTCHA.
Realistically no multiple choice system, as advocated by a number of posters here, will succeed unless it has so many choices that it's improbable a real user will be able to use the system without issues.
CAPTCHAs are a bad idea in general. Yet again they're a poor, unwieldy, temporary "solution" to a problem the inventors barely understand that causes more problems than it fixes. Like 99% of anti-spam solutions. The only thing worse than a CAPTCHA is what'll replace them.
Absolutely not!
The Cube was released at a price close to Apple's PowerMac range. The Cube certainly wasn't a "Headless iMac" in terms of price, it was actually more expensive than most iMacs. At that time, both the PowerMac and the Cube were priced in the mid $1,000s.
The problem with the Cube is that it emphatically wasn't what was being asked for. People wanted a low cost, expandable, headless Mac, not a high cost, stylish headless Mac with poor expandability. Apple currently doesn't cater to that market segment. It's made the deliberate decision not to. The Mac mini is the closest thing they have to such a thing, and it, frankly, is a poor choice, especially right now.
Apple's primary concern is about profits, and with such a small percentage of the market, they can get away with ignoring large consumer segments while appearing to do very well. The fact is though that there's a fundamental flaw in their marketing system. They believe that you can segment a market's tastes and requirements through how much they're prepared to spend. As such if someone can afford an iMac, the Apple belief is that they want an integrated, non-expandable, Mac. If someone can afford a PowerMac/Mac Pro, then they must want an expandable, headless, machine (or a laptop.)
No other company does this. I'm bewildered by the attitude to be honest. This kind of thing would kill them if they had a serious percentage of the market.
Can you do that if you do not agree to the GPL?
The GPL isn't what's restricting you from doing that. The GPL is making some acts legal that otherwise wouldn't have been. The fact it doesn't allow you, for example, to murder Eric Raymond doesn't mean the GPL is restricting you from doing something.
I wonder how many people realise that the same is true in the US.
Software patents were never explicitly legalized in the US, instead a court decision ruled that a computer running a program was patentable. Algorithms are still, theoretically, unpatentable in US patent law, but the law has been, essentially, "creatively interpretted" by the courts (without being slapped down by Congress) to include pretty much anything.
So if you're relying upon EU rules about algorithms and mathematical truths to hold back the tide, you're playing a strategy that's already lost once.
I know what you're trying to say, but that's kind of wrong because it presumes all cases of "testing" are of the form "The GPL is invalid, therefore I can do what I want." This is not true. Not all cases are of that form.
If someone interprets the GPL as meaning it's ok to release a proprietary extension to a GPL'd program using some type of unusual linkage mechanism, then yes, it may be necessary to test this in court depending on the beliefs of the person who wrote the code they're licensing. The person who's interpretting the GPL that way isn't arguing that the GPL is invalid, and that they don't have a license, they're arguing that they do have a license, and it means something other than that the licensor thinks it means.
Whether that's applicable in this case is for someone who read TFA to answer.
Reread the parent. He said that a project that has both code licensed only under the GPL and code only licensed with {a license incompatible with the GPL} cannot be in Debian, because it would be illegal to distribute.
This isn't about putting Apache and GNU C in the same distribution. It's about putting filemanager.c and documentview.c in the same binary when filemanager.c is licensed under the XGL, and documentview.c is licensed under the XGL-incompatible YGL. That's the core of the problem here.
400k and 800k floppies work the way you describe, but you're unlikely to come across them. (Well, you're unlikely to come across Mac floppies at all, but you're even less likely to come across 400k and 800k disks)
My Dish Network 625, which supposedly, according to TiVo advocates, is nothing more than a poor rip-off of TiVo's totally original wouldn't-have-been-invented-anyway technology does, in fact, have a commercial skip feature, which doesn't require any registry hacks (or secret key presses) to enable.
I'm kind of baffled by it. Why aren't the networks pissed at Dish Network? How is TiVo way better than Dish as TiVo advocates claim? I wonder if the "secret" commercial skip thing is itself viral marketing (Where's the fun in talking about a normal feature? A feature that's hidden because "The Man" doesn't want you to have it, on the other hand, is worth sharing with your friends...)?
When you say "lots of small processes", do you mean "I'm generally using a few hundred megs, but I'm running a few hundred processes so that's pretty good" (ie you're using a significant percentage of your memory, but nowhere near enough to worry about getting more) or do you mean "I run 30 or 40 processes each of which is maybe a meg at most"? (ie virtually all your memory is free, the amount you added was overkill)
I ask because it's a peculiarity of Linux, and some other VM implementations, that it's frequently better to have swap in the first situation than not have it. Why? Because of disk caching. VM allows the kernel to swap out idle processes (or idle memory from processes anyway) in favour of caching the sectors of recently processed files. This can make for dramatic speed-ups. For a desktop workstation, this can be good and can be bad, a search through a bunch of files that make up a large part of the disk can cause every major application to crawl as they're reloaded into memory.
But in your example, (a) you're not using it as a workstation, you're using it as a server (which tends to imply it'll be much more file intensive) and (b) you're running small processes, not large ones, so the time to recovery for each process will generally be much quicker.
If you're barely using your memory, then ignore this advice. But if you're using a significant percentage of your memory, and you're using your system as a server, the probability is extremely high that you'd benefit from virtual memory.
That doesn't match my experience. As an example, most of the programmers in my department are avid game players. Ages vary from early twenties to mid forties.
And most of them have actually told me that games are their major reason for not switching to GNU/Linux. (Yes, "solutions" like WineX or whatever its called today exist, but who wants to subscribe to a service if they don't need to.)
I agree. And I find it pretty sad I've had to scroll down half way through the first page to find someone who isn't posting "OMG!!!1! Free Music! Finally teh RIAA si giving us what we want!"
I am not a freeloader. I do not use Kazaa and every actual freeloader (as opposed to occasional innocent who's been caught up in the thing) who's paying $3,500 to the RIAA right now to settle lawsuit is getting what they deserve, as far as I'm concerned. I want to donate towards the costs of producing music. I'm happy to be required to buy a CD or MP3 in order to have the right to listen to that piece of music. But I do want certain reasonable rights once I've contributed to the funding, including content shifting and conversion.
This isn't giving me what I want. It's responding to the "Music needs to be free - as in freeloader" mob rather than the "Content needs more freedom" group. It's the logical outcome of the Napster mess, not the idealistic nonsense peddled by Napster and Kazaa's apologists.
Do you actually know what UMA is? I have absolutely no idea why this complete confusion of ideas keeps coming up. I've even read the once excellent Ars Technica claim UMA is something carriers are scared of. Right now, the only major carriers that might be scared of it in the US are Verizon and Sprint. Because they can't use it. It's a GSM technology.
UMA does not cut into an operator's revenue stream. It frees up revenue because the operator is not having to put up towers to get coverage and capacity for every single building in the world. If YOU, the user, save money, it'll only be because the operator is giving you discounts for using UMA, not because you're sticking it to the man by using it, somehow bypassing the carrier. Far from it. You're using the carrier either way.
UMA is not "I can bypass the cellphone company to make free calls", it's "I can route the last mile of my calls through either the radio waves to a tower or via the Internet to a gateway at my carrier, either way getting to my carrier who'll then route the call as necessary." It's a great technology, but what makes it great is that it means that people can make coverage where they currently have blackspots.
What's confused some people is they've read all this crap about Skype phones, and think that UMA is this. It isn't. It's GSM routed over the Internet. Skype phones are something else entirely.
Other people are confused because they've heard it's VoIP. VoIP does not mean "Cheap ways to bypass the phone companies", it's a just a name given to any form of two-way voice traffic routed over IP packets. Just because using Vonage over cable is "sticking it" to AT&T&T doesn't mean that all forms of VoIP are.
This is why T-Mobile and Cingular are members of the UMA consortium and are planning to roll it out here in the US. Yes, they are. Yes, they've made announcements to that effect. It may make calls cheaper. More importantly though it'll make calling more reliable. No more blackspots in the kitchen. Nice.
Good lord. You mean ESR is insulting and alienates the very people he needs to get on his side?
I'm glad he's no longer the official spokesman for the Open Source movement any more (he has some honory position but nothing important.) I think what he had to say was kind of interesting (and said so when this article was published the first time around (albeit with a misleading title) but it needs to be put forward by people a little more knowledgable and with better communication skills than himself.
Even the FSF doesn't entirely disagree with the stance, despite the number of comments here suggesting "compromise" is some great chasm between the FSF and OSI. The FSF created the LGPL and while they officially discourage its use, it's still very much a live license that's seen as appropriate in many cases.
So: GNU/Linux users are self-absorbed geeks who learn Klingon and attend science fiction cons. You can't get iPods to work with GNU/Linux. GNU/Linux needs to compromise and often allow proprietary software. Now is the time to strike, because with the move to 64 bit operating systems, everyone will throw out what they already have and buy entirely incompatible 64 bit systems. Great work there ESR, you are truly are relevent in today's age, hip and with-it. Knowin' the trends.
Here's the funny thing. Everyone's acting like "compromise" is a big thing. And it isn't. We all want to be free, but few people haven't compromised on the Free Software thing at one point or another no matter how strong our beliefs in terms of what we use. The first time this was posted, Slashdot went one step further and claimed ESR was an advocate for proprietary software, which was heavily misleading, if not an outright lie.
What ought to be questioned is what never is. The insults. The snide comments. The pro-factionalization. The implication that he's promoting something new when he's repeating what almost everyone in the community has been doing from day #1.
Oh c'mon! WTF is wrong with you all? Is the world so black and white in Slashdot that if someone protests about being punished too hard, they can only be advocating no punishment at all?
The (ex)Apple employees are protesting that they came clean and yet endured the same punishment they'd have endured if they had not come forward but been caught anyway. The complaint is not that they were punished at all, it's that the punishment was excessive and gives nobody any incentive to be honest.
And they have a point. And this not about murder, where arguably the action is so severe that the appropriate punishment should always be dealt, it's about a case of copyright infringement. Yes, there's room for Apple to take a more lenient line with truth tellers than with those who lie. Especially when given the case is ultimately about whether an employee can be trusted with the company's proprietary inside information, the issue of whether they lied or not in an investigation is actually relevent.
Apple has arguably over-reacted. And whether it did or not, it has most certainly cut off its own nose to spite the face of others. Firing is an expensive act. Apple can expect to lose the productivity the fired employee would have given to the company during the time it recruits and trains the replacement, and recruiting is hardly cheap either. Further, it has made its own future investigations harder because it will not get the cooperation of employees who see themselves as ultimately loyal.
Apple can hire and fire whoever they want, for whatever (legal) reason. But that doesn't make this anything other than, at face value, assuming there's not more to it than TFA, a dumb decision. And certainly, the logic Slashdotters promote of "IF THEY HATE THERE PUNISHMANT, TEHY MUST FINK BEENG PUNICHED IS RONG!!1" is utterly irrelevent and idiotic.
The WMF flaw is a design flaw, not a coding flaw.
A lot of the "holes" in recent Windows have to do with design problems. The problem is it's one thing to go around and fix coding bugs, it's another to fix design issues because programs are built around designs.
That said, Vista isn't the rewrite it was originally intended to be.
Each generation of Intel integrated graphics system has gotten steadily better, and supposedly the next generation, present in the 965, uses a design (the X3000) that can hold its own.
My guess is that Intel will probably end up sweeping the market. If this kind of thing is being incorporated into computers for "free", then the number of people who "need" better graphics is going to drop dramatically. Tech people will go for Intel because it's open and "good enough". Games developers will optimize for it because it's open, good enough, and deployed everywhere.
In that kind of market, I suspect AMD will want to use ATI for:
1. Developing niche cards for certain industries.
2. Developing designs for console manufacturers.
3. Developing designs for integration into motherboard chipsets (like the X200) to support AMD CPUs.
In that playing field, AMD isn't really going to care about ATI's behaviour concerning older chipsets. For the niche cards, ATI is probably going to develop one-off specific drivers. For the consoles, the manufacturers will probably develop their own drivers, and for the integrated chipset designs, they'll probably open up a little.
I also have an X1300 laden laptop (a Thinkpad T60) and have tested it extensively under Debian Stable, Ubuntu, and RedHat Core 5.
The drivers are shit. The machine will not even boot into X11 if the power is unplugged. Running Unreal Tournament doesn't work (you get massive lag between keyboard/mouse events and things happening on screen.) Unreal Tournament 2004 "works", but frequently with bizarre artifacts that make the display unwatchable. Full on crashes happen every few days (triggered usually by watching a movie in MPlayer.)
This has nothing to do with "F/OSS zealots", and you blaming them (in response to people saying the nVidia drivers, which are equally closed, are better no less!) shows you have an agenda. The fact is I'll never buy ATI again after this experience, and I'm recommending others avoid them too. I prefer open source, but ultimately I want something that works. If nVidia can provide that, go with them.
I'm still having difficulty with this argument. It looks like the argument is "Does it hurt - that is, cause unhappiness in - the child? If so, it's a form of abuse."
Which makes discipline a little difficult if all options that in any way "hurt" the child are off limits. It also ignores the real world costs of providing many of these items, both monetary and otherwise.
I can handle the argument that needless or extreme punishment is abusive, and that acts of violence, intended to be disciplinary or not, are abusive. But this is not a world where we can legitimately give everything to anyone whenever they want. And sometimes, to do so even when we can would be inappropriate and would hurt the receiver in the long term.
You don't have to file trademarks defensively. Well, let me rephrase that. If you're building a product that has an entirely new name, never before used in relation to your market area, then yeah, you need to file a trademark to ensure nobody else starts making things with the same name. Though it's arguable that's "defensive". You're defending your product against knock-offs, but you're not trademarking it for the same reason as someone might defensively file a patent. That is, you're not doing it so that if someone sues you for trademark infringement, you can sue them right back.
Defensive patents are an entirely different area. People file patents so that if someone sues them, they have some potential ammunition against them. For example, ACME Sandwiches sues you for violating their Cheese Between Two Slices of Bread patent, you in turn can see that they have a wide range of sandwiches, some of which violate your own patents, you can sue them back for violating your "Peanut butter between two slices of Ryvita" patent.
O'Reilly isn't really being "defensive" here. He's taking words that relate to things he's doing conferences on, and trademarking them. The net effect is to make it harder to come up with relevent terms to describe competing conferences. That's really being offensive, not defensive. It's not a good thing.
It does actually appear to have been a common term to describe, erm, websites: Examples. Of course, there weren't many in 1994.
There may not have been a massive number of mentions, the count is in the hundreds not thousands, but that appears to be the time the term started to become popular. So it looks like O'Reilly jumped on a technology that was emerging at that point, and decided to trademark a term already in use by those already using the technology.
That doesn't strike me as acceptable.