That sounds like a known problem with xfs: when a client requests a ridiculously large font (and mozilla doesn't filter out such requests), xfs can go insane trying to service it.
It's a bug, but it's also a bug in the client - it/shouldn't/ be sending ridiculous requests to the server. After all, the server really needs to try and handle/any/ requests, whereas a client can decide that a font size it's been given is stupid, and drop it . ..
There's a fix for it out, I think, but I don't know if it's been distributed.
Copyright does not restrict access to information in any way. Copyright is designed to protect authorship, not to restrict the flow of information. In fact, you will find that pure information cannot be copyrighted. Classic examples include contest rules, recipes and phone books.
How can these things be anything/but/ information, if digitising them results in a copy that could be considered equivalent to the original? That's the whole point of digital data storage: you take some representation of something, and convert it into a big number. The encoding scheme and that number/are/ that piece of data, and if that's equivalent to the original, then the original was obviously a piece of information that/could/ be encoded that way.
It's ok to get philosophical, but you should do it honestly. This whole "information" charade denigrates artists and authors, pretends that they make no unique contribution. If that were the case, however, that they truly add no value, then there should be no market in pirating -- oops, liberating their product.
Would you like to explain to me how calling something information denigrates it? This seems to be beyond my feeble mind, which sees the word 'information' as the appropriate term for describing meaningful data: that's how the word is used in computing, at least. Meaningful data is hardly an insult, unless you find the idea of calling art 'data' insulting.
I also can't help wondering where you got the idea that I was pretending that people creating new things made no unique contribution. In fact, I explicitly stated that they add their own creativity to the mix, and that this added value. I also stated that people using other's creations added their own creativity, and that this was an endless cycle of building on other's works. Yes, each person is adding value, but each person is also deeply indebted to those they're building on - should we force anyone who wants to create something new to seek out all their possible sources of inspiration and pay them money? Only the first people in the chain? Or the first two, if they happen to be alive? Where does that end? And why should one person get something for free from the public domain, when another has to pay? Was one worth more than the other, because one author died 150 years ago and one died 50 years ago?
I'm not saying there's no value in these things, I'm arguing with the idea that the person who created them deserves some unique right to profit from them. After all, that person was building on other's work - why should he profit when/they/ won't?
In fact, it does more. It turns the moral and just (that the spoils of creation should rightly flow to the creator) on its head (any old bozo should be able to appropriate your work without consideration).
Okay, there you just sank into the realms of bullshit. "moral and just"? What's moral and just about that line of reasoning? It's probably a good idea, if you want those creators to keep on doing whatever it is they're doing (though even that is debatable where 'art' is concerned, since most art is done for personal reasons, not profit), but moral and just? Moral, that one person should profit from something when millions might profit otherwise? Just, that one person should profit when those this person used don't? I'm not entirely sure I like your morality, or your idea of justice.
Bah. I'm disappointed - I was looking for a reasonable discussion, not a collection of straw man arguments and ad-hominem attacks. Thankyou for ignoring my argument and ruining my day.
It might be useful to turn that one around, actually . ..
By what right does one person restrict another's ability to use and share information that they have acquired from somewhere?
That's the core of the problem with the current clash between the IP supporters and the sharers (though they may not really know it): this is pure information that people are trying to control, and once that's been released there are no natural barriers to its dissemination.
Legal barriers to copying were created specifically to make information something that could be owned, in order to encourage people to make it available (which would, in the natural order of things, make that information universally available, and hence impossible to profit from in the normal way). They don't change the fact that information has no natural barriers to dissemination, they merely create legal barriers to that end.
Personally, I'd argue that I have a right to any information I can get my hands on, and once I have that information I have a right to do whatever I want with it. Releasing information to the public is exactly the same as giving it away freely - there are no natural barriers to stop it's spreading, or to stop it's use. That's the way things work when there are no artificial barriers put in place. It's also a very nice way for things to work - it means I can take all the information I have and use it to create more information, without worrying about ownership or compensating anyone else, or anything like that, things that would normally stymie my creativity greatly. It also means that/I/ can't get any compensation for my creativity. But then, why should I get compensation for building on other people's works? Why should the people I built on get compensation for building on/their/ foundations? All everyone ever does is take an idea and add a bit of creativity to it - this is a valuable exercise, but it's entirely useless if you/can't/ take that idea in the first place.
So, what does anyone gain by making information something that can't be shared universally? Well, whoever 'owns' that information presumably gains some profit from selling it, and everyone else gains in some limited way from having the information available, even if they're not allowed to build on it without jumping through hoops.
How does this compare with the case where the information is made freely available to be built on? In this case, the creator of the information gains less, possibly no profit from it (there are definitely models where profit can be gained, and the creator can generally provide some unique qualities related to this information that may allow profits in related areas). At the same time,/everyone/ else gains vastly more than they would have had: they can now take this information and build on it, without any limitations other than their own creativity. Disney can make animated features out of it, playwrites can create salacious parodies laced with all sorts of crap that the originator would never have thought of, writers can create books based on it . . . Anyone can do anything they want. Shakespeare's works are perfect examples of this: think how much has been built on/their/ foundations, that would not have been built if they hadn't been in the public domain.
Copyright laws supposedly strike a balance between these two sides: enough ownership to allow creators to profit, but not so much that other's can't build. And that's a wonderful idea (I won't go on about how current variations on copyright laws destroy the balance completely - this is more a tract on the underlying philosophy and how realities have changed).
The problem with copyright laws in the current technological climate is that the actual barriers to dissemination of information have suddenly dropped to almost nothing. Before the digitisation of everything, there were real physical and economic barriers that meant that the originator of a piece of information had a seriously big head start when it came to profiting from that information, regardless of copyright laws or anything like that. Now, though, that has disappeared. Digitised information comes damned close to matching the idealised notion of information that can be universally shared with no natural barriers. It's now just as easy to send someone a copy of a book as it is to tell them about it - two ways to disseminate information, one of which is perfectly legal, the other not, and both conceptually almost identical.
This doesn't change the fact that the concepts behind copyright are nice. It/is/ generally a good idea to compensate people for their creative work, and no one argues otherwise. What/has/ changed is the workability of current copyright laws: there are no actual barriers remaining, which means that the originator of any piece of information is in almost the same position as everyone else when it comes to profiting from it. No amount of law-making or discussion of the rights of creators will change that - it's just something people have to get used to.
It seems to me that creators can take one of two roads now: they can try to create an artificial scarcity of the information they create, so that the old model will still work, or they can look for some other way to profit from their creativity. That's where things are now, and as I said, no amount of laws or assertions of people's rights will change the reality.
So, to summarise the philosophical side, I don't think the question is what right do you have to someone else's work; rather, the question is what right does anyone have to limit the uses one can make of information one posesses, and what right does anyone have to demand compensation for adding their creative bit to the sources they build on.
The pragmatic side of this is simple: the compromise of copyright is a law which will only work when there are real barriers to the sharing of information, outside the legal controls. Without those barriers, some other compromise needs to be found.
Or, we can lock information up so tightly that it becomes useless to us - this seems to be the favoured approach of the commercial creators . . . I'd rather not see that happen, personally.
That's about right, yes. Though I believe the currently popular terminology is more along the lines of "bush" than tree or ladder - when you look at the details of the fossil record, there tends to be a lot of experimentation which then settles down over time, so you don't get nice clean branches, you get a big messy bush. Which may have changed since I last did and anthropology unit, but hey . ..
As for the other crap the parent poster was throwing out, it's tragic - he (she? it?) could at least have done some research before creating a terribly broken argument like that. And unfortunately I'm too tired to pull it apart more . . .
Most of the core ideas in Unix were developed in the 60's, actually.
Computing in the 50's was a very different thing, so limited that the idea of wasting cycles on things like memory management or protected memory would have been considered insane. It wasn't until hardware developed to the point where there were cycles and memory to spare that anything like Unix (or MULTICS, which is where most of Unix's ideas were developed) became possible.
No one has ever said that the author/should/ be forced to put any time or money or anything into reviving something that falls into the public domain - the/only/ requirement is that any copies that are available become public domain, and hence usable and modifiable by anyone.
I have no idea where you got the idea that there was any kind of requirement on the author - it's/never/ been claimed by anyone, to my knowledge, and the idea really is laughable. I mean, most of the time the author is/dead/ . . .
The projected clock speed at the initial release is 1.6GHz, and it's expected to scale way beyond this. The 800MHz clock is for stepping 0 silicon - ie, this is an example of the first working silicon they've produce from this line, without any of the cleanups and optimisations that will go into it by the time it reaches release status.
These numbers are really extremely impressive - if the perfomance scales with the clock speed, an Opteron will outperform a P4 running at twice the clock speed. Given that it's also been designed to scale, it's going to be hard for Intel to keep up . . .
CDs have a frequency response that goes up to ~22KHz - that (strangely enough) is about where humans stop being able to hear stuff. CD audio/cannot/ encode anything higher than that. So, you can't hide a watermark in a range above what humans can hear on a CD.
In order to implement that kind of thing generally they'd have to get/everyone/ to replace/all/ their audio equipment - they'd need to move everyone onto DVD-audio, which handles up to 96KHz. Now, I'm sure they'd/love/ to be able to do that, but I seriously doubt that consumers would be happy with that idea. CDs are so popular because they're right about good enough for everything people want from them - they're even good enough for high quality music, though that'd probably be helped by going to DVD-audio. For the vast majority of the music sold, anything beyond CD audio is wasted - witness the prevalence of mp3s at 128kb/s, which sound shitty on any reasonable speakers.
With visual watermarking, again, the technology is carefully designed to match up with the limitation of human senses - you could probably encode something in the ranges outside what we can see, but the response of the ADCs is/also/ tuned to what's useful for humans, so it'd likely go missing there, too. Are the movie studios going to mandate that all video cameras be built with detectors that can pick up some particular near-infrared frequency, so that they can see the watermark? Are they going to pay the manufacturers the difference in the price due to using that more capable detector?
This is the thing that has me completely buggered here - how do these people think they can force/everyone/ else to jump through hoops just so that/they/ can make more money? It's ridiculous. It'd be laughable, if they didn't seem to own enough politicians to make it happen.
WindowMaker is a bloated Windows wannabe??? Have you ever/used/ it? It's a NeXT clone, which is/vastly/ different to Windows. It's also not particularly bloated - it weighs in at 1824kB on my box at the moment.
Afterstep is, I believe, another NeXT clone (note the 'step' in the name).
Gnome/KDE I won't comment on - I don't use them, since I like wmaker.
Please at least try and know what you're talking about when you flame things.
Ummmm . . . No, the type of *foo in char *foo; is a char, and foo is a pointer to a char. There's no notion of forcing anything here - you're/declaring/ the variable, not dereferencing it.
In any case, even when you/are/ actually dereferencing, unless you explicitly cast the variable to a different type somewhere along the line the compiler will pick up the mismatched type error. Since C isn't strongly typed, this often gets you a warning and nothing else, but that reflects the fact that C lets you treat a chunk of memory however you want, because it was designed as an OS implementation language where you need direct manipulation of memory. But the language itself/does/ have a strong notion of the type of variables - strong enough that you can do a lot of static analysis. It breaks down around pointers and typecasts because you can throw away information if you choose, but again, the language's notion of types is well defined, with loopholes that make playing games with memory easier.
I could compile C to Caml (a strongly typed language) if I wanted to, and the only difficult bit would be providing generalised typecasts (Caml doesn't allow casting in the general case, only for things like ints to floats or chars to ints, where there are clear, well defined conversions). Most of the type handling would pass straight through untouched.
Don't make the mistake of looking at the runtime behaviour of a language and thinking that defines it's semantics. In this case, a construct that's entirely meaningful will cause a runtime error if some other stuff isn't done, but that doesn't change the fact that it/is/ meaningful, and that the language properly defines that meaning.
And use 8 space tabs - that makes code blocks more obvious than anything else, and it even warns you when you're getting ridiculously deeply nested control structures.
Another good thing to read is the Linux kernel coding style documentation (Documentation/CodingStyle). It's a good discussion of a lot of this stuff.
See, now you're rewriting what you said. you didn't say "patents relating to CMYK prevent its implementaion", you said "CMYK is patented". You're inventing reality as we speak. How is anyone supposed to take you seriously?
And now/you're/ being a stupid prick. I/said/ in my second reply that I'd gotten the details of my claim wrong, and I clarified them, with references. Yes, I changed my claim - that's the point of being able to have a continuing dialogue. I cleared up my misremembering of the situation, and I did so in a reasonable manner, with references.
Furthermore, if the GIMP is not a poorly written hack, then why are they planning a massive rewrite for version 2.0? If it's so good, why rewrite it?
This demonstrates a severe lack of understanding of how real world software is developed. They did an initial implementation of the core, and it worked for what they were interested in at the time. Later experience showed that it/didn't/ work for areas the software grew into, and as a result of the changed demands they decided to reimplement things. That seems quite sane to me, and to a lot of other people who develop software.
Anyway, if this is so terribly horriffic, maybe you should villify Microsoft - they did exactly the same thing when they started developing NT.
Ok, how can you make an informed decision about which is the best product to use if you don't "know anything much about any of them". How can you be sure you've chosen the best product if you have no knowledge of the competition? Once again, you're proving your blind allegiance to the inferior.
Well, given I get paid to develop Linux software, I really don't know how something like VC++ could ever be the best choice . . .
I/do/, however, use BitKeeper - commercial software, which just happens to be vastly superior to any of the free alternatives. At the same time, I use GNU emacs, because I prefer its interface to Xemacs, and I like the way emacs in general works. I use gcc because it's a good compiler, freely available, and because it's necessary for compiling kernels. I use mutt as my mua, because when I started reading lkml it was the best choice - now, I use it because I just happen to like it.
All of these are my personal preferences. I'm not standing up and shouting to the world "You must all use GNU emacs, because Xemacs is evil!" I'm simply saying/I/ prefer it. That's definitely not zealotry.
Your delusion is self-defeating twofold: 1) If you're not willing to admit that OSS has a ways to go before it's "as good as" commercial software, it's never going to progress. 2) Other people will take OSS less seriously, because information about it from proponents like you is not 100% factual, further hindering its adoption.
I'm not only willing to admit that open source software has a ways to go in many cases before it's as good as commercial software, I'm willing to put time and effort into helping get it there. Or, as is the case with my using BitKeeper, I simply use the superior commercial software.
As for my information not being 100% factual, you're right, I/did/ screw up my original post - I should have done that half hour of research on google/before/ posting. Thankyou for calling me on that.
Not that it would have made any difference - you would most likely still have been trolling, and I'd probably have ended up just ignoring you completely.
You're trolling quite well, for what it's worth. You're just a little too vehement in ignoring what I've actually said, and that's making you look like a fuckwit.
Blindly defending something with a completely uninformed argument? Was I doing that?
I was off a little in the details, but half an hour's research on google came up with more information, which backed up my original claim: the gimp didn't implement CMYK because of patent problems related to it.
Would I have defended a Microsoft product this way? Perhaps, if I'd seen someone trashing it the way you were trashing the gimp. Most likely not, though - I don't use any MS products, nor do I know anything much about any of them, so I'm not likely to step in with an explanation of why something is the way it is.
I have problems with MS's approach to business, based on what I've read about them (including reading through the findings of fact from their current court case). But then, I have problems with many businesses based on my experiences with them. That doesn't make me a zealot, by any stretch of the imagination. If anything, it makes me an informed consumer - something to be proud of, I think.
I realise you're trolling, by the way, so I won't respond to any more comments you make unless there's some real substance to them.
Hokay . . . Having done some research, yes, you're right that CMYK itself isn't patented. There/are/ patents on colour correction, colour management, and various other things related to handling CMY colour for preprint work - this/is/ the reason support hasn't been added to the gimp in the past.
My references: http://static.userland.com/userLandDi scussArchive/ msg005514.html (somewhere down the middle of the page) http://www.levien.com/gimp/gcmm.html
As for the implementation details, I can't comment - I'm a systems coder, not a graphics person. But I'm fairly sure the gimp's implementation of RGB graphics is/not/ sub-par - it wouldn't be so popular for web graphics otherwise.
Finally, I'm not a zealot, by a long stretch. I personally prefer free software, but I have many good reasons for this, not least of which is the fact that without free software I wouldn't have gotten interested in coding at all. Does having personal preferences make me a zealot? If so, then you're just as much of one as I am.
What is so terribly horrifically frightening about taking reasonable responsibility for your own competence? The same kind of responsibility that an engineer making a component for a car takes, or a builder building a house, or anything else like that?
It's quite simple: do the best job you can, as responsibly as you can, and taking all due care. If something goes wrong after that, then you should be safe from punitive liability. If you're negligent, then you/should/ be held responsible for your negligence. If you're incompetent, then you should be held responsible for any claims of competence you made. And if you're not willing to accept responsibility for what you do, you shouldn't be doing it.
The only people who can avoid responsibility for their actions are children - is this industry really that immature?
CMYK is patented, and licensing this patent is not at all cheap. Certainly, it's not something that's possible for a piece of software like the gimp.
Claiming that a piece of software is inadequate because the maker of the competing software uses legal means to stop competitors from implementing a piece of functionality is really quite stupid.
Using C++ funky bits makes it much harder to control exactly what code the compiler produces, regardless of whether it's more efficient or not. You also need a runtime library for a lot of things. Those are the two main reasons C++/isn't/ used.
The control thing probably sounds like a red herring, but in kernel land knowing exactly what your code is doing is extremely important - every cycle really does count. Particularly in the core kernel, you often see strange and confusing constructs that are in there simply because they've been shown to produce faster code. And this isn't some kind of he-man "I can write more gotos than you!" crap - people actually sit down and look at the asm output from bits of code and decide which version is best.
That's the main reason why the kernel is so conservative about what compilers it builds on - there's code in there that relies on specific behaviour of gcc extensions to achieve the exact results desired. This probably sounds like a bad thing, but this is an OS kernel - performance is/everything/, and when the compiler you're targetting is as portable as gcc, it's not even that big an isue for portability.
Kernel coding is/not/ like userspace coding. The code tends to look much the same, but you operate under constraints that make the details vastly different, and which make things like C++ difficult propositions.
Write your own article outlining all this and giving your analysis of what's going on, and submit/that/ to/. - you may have more luck getting heard that way . ..
Damn good point there - MS has been criminally negligent, and they've hidden behind the lack of liability clauses they put in their EULAs for too long. It's the same with other software companies, but MS's screwups are writ extra large because of their dominance.
Someone needs to take MS to court charged with negligence, and put an end to their arrogant assumption that they can do whatever the fuck they please, because they say they can.
The whole point of unions is forcing employers to give reasonable compensation (in money and in other ways) for the work their employees do for them. But it's generally not a one way street - the employees have to commit to things, too, even if it's only an implicit "Give me this standard of wages and conditions, and I'll work for you without complaint".
Unions like the AMA do more than that - they require standards from the employees, as well, standards to protect both sides. You get union protection, but only if you meet union standards. That probably sounds as terrifying and horrible as taking due responsibility for flaws in your software, but it's entirely reasonable in a skilled profession where your responsibilities are high. If you claim to be able to do something, you should be responsible for the results if your claim was false.
Historically, unions have done some screwed things. They've also made more and bigger changes to the standard of living in many countries than anything else in the last five hundred years. Get the implementation right, and they're/good/, for everyone.
The other thing about symlinks is that they can point at a target on a different filesystem - hardlinks can only refer to a file on the same filesystem.
Given the *ahem* limitations Windows in general has with handling multiple filesystems transparently, that probably doesn't come up much at all in Windows, but it's very very useful on *nix systems, where/usr,/usr/local,/home, etcetera are often seperate filesystems, and being able to link between them is useful.
The other use that comes to mind is with shared libraries - the basic file typically has a name like libfoo.so.1.2.3.4, indicating the library version. This will generally be accompanied by symlinks that point to it, with names like libfoo.so.1, libfoo.so.1.2, and so on - this facilitates having multiple versions of the same library installed and allowing programs to link to specific versions. Rather than link to libfoo.so, if it needs libfoo.so.1.2 it'll link to that pathname, and get the latest version of libfoo with that major/minor number. The details are all handled by ldconfig, which manages the links.
Symlinks are terribly nifty, and very very useful - I was rather shocked when I discovered that NTFS didn't support them, because they're simple both conceptually and implementation wise (I know - I've implemented filesystems with symlink support).
I am well aware that Linux has component models (NOW) such as Bonobo, KParts, Mono etc -- I wondered who they were copied off.
Actually, Bonobo is based around a CORBA implementation, and I believe KParts is too. If implementing a well defined and open standard is copying, then there's your answer . ..
Mono is a reimplementation of the.NET libraries - no prizes for guessing where/that/ one's copied from.
I'm also aware Linux has a kernel module system similar to device drivers in windows. Windows has had those for years. You often still have to recompile the kernel to get a certain new feature or piece of hardware working and MANY, MANY applications are still statically linked. the majority of apps on Linux are non Gnome/KDE apps which have a monolithic design (much like the Linux kernel).
Cut and pasted from the comment at the head of linux/kernel/module.c:
/*
* Originally by Anonymous (as far as I know...)
* Linux version by Bas Laarhoven
* 0.99.14 version by Jon Tombs,
* Heavily modified by Bjorn Ekwall May 1994 (C)
* Rewritten by Richard Henderson Dec 1996
* Add MOD_INITIALIZING Keith Owens Nov 1999
* Add kallsyms support, Keith Owens Apr 2000
* Add asm/module support, IA64 has special requirements. Keith Owens Sep 2000
* Fix assorted bugs in module verification. Keith Owens Sep 2000
* Fix sys_init_module race, Andrew Morton Oct 2000
* http://www.uwsg.iu.edu/hypermail/linux/kernel/0008 ./0379.html
* Replace xxx_module_symbol with inter_module_xxx. Keith Owens Oct 2000
* Add a module list lock for kernel fault race fixing. Alan Cox
*
* This source is covered by the GNU GPL, the same as all kernel sources.
*/
So you see, it predates Windows 95 by a few years.The earliest date mentioned is '94, but it's obviously older than that - could it be that it predates even Windows NT?
As for "MANY, MANY" applications still being statically linked, I'd like to see some evidence for that. Certainly, there are/very/ few applications on my Debian system that aren't dependant on libc, which is a dynamically loadable library - ie, it's/not/ statically linked. Hell, even/sbin/init is dynamically linked:
I'll accept that you may be knowledgeable about Windows, but you seem somewhat lacking in knowledge of Linux . ..
Finally, something on-topic:
Modularity doesn't mean you can remove a component without any effects.
Well, what I was taught in my CS courses was that the idea behind modularity was that it let you/replace/ components without any effects. So, while completely removing the MSHTML component might break a gazillion and one applications, replacing it with a gecko-based reimplementation of the same interface shouldn't even be noticed.
It'd be nice if someone involved with this case made a point of bringing/that/ issue up . ..
Generally speaking, they/have/ to be - they'd get screamed at insanely if they weren't. However, I'll point at various incidents like the NT service pack that caused Lotus Notes to stop working, and their skullduggery about windows 3.1 on DR-DOS. But yes, in their current position they're limited by the weight of software that they need to support, so there's a limit to what they can change.
I believe changing the OS interfaces so that Navigator wouldn't run anymore was one of the things they considered when they first looked at killing Netscape. I'm not sure if they decided they couldn't do that without breaking other things, though.
My other points stand unaltered. Microsoft/does/ have monopoly power, and it/does/ abuse it in various ways. Arguing otherwise is just plain stupid (or, in the case of Mr Reflection here, trolling).
Slashdot Mirror
That sounds like a known problem with xfs: when a client requests a ridiculously large font (and mozilla doesn't filter out such requests), xfs can go insane trying to service it.
/shouldn't/ be sending ridiculous requests to the server. After all, the server really needs to try and handle /any/ requests, whereas a client can decide that a font size it's been given is stupid, and drop it . . .
It's a bug, but it's also a bug in the client - it
There's a fix for it out, I think, but I don't know if it's been distributed.
himi
I also can't help wondering where you got the idea that I was pretending that people creating new things made no unique contribution. In fact, I explicitly stated that they add their own creativity to the mix, and that this added value. I also stated that people using other's creations added their own creativity, and that this was an endless cycle of building on other's works. Yes, each person is adding value, but each person is also deeply indebted to those they're building on - should we force anyone who wants to create something new to seek out all their possible sources of inspiration and pay them money? Only the first people in the chain? Or the first two, if they happen to be alive? Where does that end? And why should one person get something for free from the public domain, when another has to pay? Was one worth more than the other, because one author died 150 years ago and one died 50 years ago?
I'm not saying there's no value in these things, I'm arguing with the idea that the person who created them deserves some unique right to profit from them. After all, that person was building on other's work - why should he profit when
Bah. I'm disappointed - I was looking for a reasonable discussion, not a collection of straw man arguments and ad-hominem attacks. Thankyou for ignoring my argument and ruining my day.
himi
It might be useful to turn that one around, actually . . .
/I/ can't get any compensation for my creativity. But then, why should I get compensation for building on other people's works? Why should the people I built on get compensation for building on /their/ foundations? All everyone ever does is take an idea and add a bit of creativity to it - this is a valuable exercise, but it's entirely useless if you /can't/ take that idea in the first place.
/everyone/ else gains vastly more than they would have had: they can now take this information and build on it, without any limitations other than their own creativity. Disney can make animated features out of it, playwrites can create salacious parodies laced with all sorts of crap that the originator would never have thought of, writers can create books based on it . . . Anyone can do anything they want. Shakespeare's works are perfect examples of this: think how much has been built on /their/ foundations, that would not have been built if they hadn't been in the public domain.
/is/ generally a good idea to compensate people for their creative work, and no one argues otherwise. What /has/ changed is the workability of current copyright laws: there are no actual barriers remaining, which means that the originator of any piece of information is in almost the same position as everyone else when it comes to profiting from it. No amount of law-making or discussion of the rights of creators will change that - it's just something people have to get used to.
By what right does one person restrict another's ability to use and share information that they have acquired from somewhere?
That's the core of the problem with the current clash between the IP supporters and the sharers (though they may not really know it): this is pure information that people are trying to control, and once that's been released there are no natural barriers to its dissemination.
Legal barriers to copying were created specifically to make information something that could be owned, in order to encourage people to make it available (which would, in the natural order of things, make that information universally available, and hence impossible to profit from in the normal way). They don't change the fact that information has no natural barriers to dissemination, they merely create legal barriers to that end.
Personally, I'd argue that I have a right to any information I can get my hands on, and once I have that information I have a right to do whatever I want with it. Releasing information to the public is exactly the same as giving it away freely - there are no natural barriers to stop it's spreading, or to stop it's use. That's the way things work when there are no artificial barriers put in place. It's also a very nice way for things to work - it means I can take all the information I have and use it to create more information, without worrying about ownership or compensating anyone else, or anything like that, things that would normally stymie my creativity greatly. It also means that
So, what does anyone gain by making information something that can't be shared universally? Well, whoever 'owns' that information presumably gains some profit from selling it, and everyone else gains in some limited way from having the information available, even if they're not allowed to build on it without jumping through hoops.
How does this compare with the case where the information is made freely available to be built on? In this case, the creator of the information gains less, possibly no profit from it (there are definitely models where profit can be gained, and the creator can generally provide some unique qualities related to this information that may allow profits in related areas). At the same time,
Copyright laws supposedly strike a balance between these two sides: enough ownership to allow creators to profit, but not so much that other's can't build. And that's a wonderful idea (I won't go on about how current variations on copyright laws destroy the balance completely - this is more a tract on the underlying philosophy and how realities have changed).
The problem with copyright laws in the current technological climate is that the actual barriers to dissemination of information have suddenly dropped to almost nothing. Before the digitisation of everything, there were real physical and economic barriers that meant that the originator of a piece of information had a seriously big head start when it came to profiting from that information, regardless of copyright laws or anything like that. Now, though, that has disappeared. Digitised information comes damned close to matching the idealised notion of information that can be universally shared with no natural barriers. It's now just as easy to send someone a copy of a book as it is to tell them about it - two ways to disseminate information, one of which is perfectly legal, the other not, and both conceptually almost identical.
This doesn't change the fact that the concepts behind copyright are nice. It
It seems to me that creators can take one of two roads now: they can try to create an artificial scarcity of the information they create, so that the old model will still work, or they can look for some other way to profit from their creativity. That's where things are now, and as I said, no amount of laws or assertions of people's rights will change the reality.
So, to summarise the philosophical side, I don't think the question is what right do you have to someone else's work; rather, the question is what right does anyone have to limit the uses one can make of information one posesses, and what right does anyone have to demand compensation for adding their creative bit to the sources they build on.
The pragmatic side of this is simple: the compromise of copyright is a law which will only work when there are real barriers to the sharing of information, outside the legal controls. Without those barriers, some other compromise needs to be found.
Or, we can lock information up so tightly that it becomes useless to us - this seems to be the favoured approach of the commercial creators . . . I'd rather not see that happen, personally.
himi
That's about right, yes. Though I believe the currently popular terminology is more along the lines of "bush" than tree or ladder - when you look at the details of the fossil record, there tends to be a lot of experimentation which then settles down over time, so you don't get nice clean branches, you get a big messy bush. Which may have changed since I last did and anthropology unit, but hey . . .
As for the other crap the parent poster was throwing out, it's tragic - he (she? it?) could at least have done some research before creating a terribly broken argument like that. And unfortunately I'm too tired to pull it apart more . . .
himi
Most of the core ideas in Unix were developed in the 60's, actually.
Computing in the 50's was a very different thing, so limited that the idea of wasting cycles on things like memory management or protected memory would have been considered insane. It wasn't until hardware developed to the point where there were cycles and memory to spare that anything like Unix (or MULTICS, which is where most of Unix's ideas were developed) became possible.
himi
No one has ever said that the author /should/ be forced to put any time or money or anything into reviving something that falls into the public domain - the /only/ requirement is that any copies that are available become public domain, and hence usable and modifiable by anyone.
/never/ been claimed by anyone, to my knowledge, and the idea really is laughable. I mean, most of the time the author is /dead/ . . .
I have no idea where you got the idea that there was any kind of requirement on the author - it's
himi
The projected clock speed at the initial release is 1.6GHz, and it's expected to scale way beyond this. The 800MHz clock is for stepping 0 silicon - ie, this is an example of the first working silicon they've produce from this line, without any of the cleanups and optimisations that will go into it by the time it reaches release status.
These numbers are really extremely impressive - if the perfomance scales with the clock speed, an Opteron will outperform a P4 running at twice the clock speed. Given that it's also been designed to scale, it's going to be hard for Intel to keep up . . .
himi
CDs have a frequency response that goes up to ~22KHz - that (strangely enough) is about where humans stop being able to hear stuff. CD audio /cannot/ encode anything higher than that. So, you can't hide a watermark in a range above what humans can hear on a CD.
/everyone/ to replace /all/ their audio equipment - they'd need to move everyone onto DVD-audio, which handles up to 96KHz. Now, I'm sure they'd /love/ to be able to do that, but I seriously doubt that consumers would be happy with that idea. CDs are so popular because they're right about good enough for everything people want from them - they're even good enough for high quality music, though that'd probably be helped by going to DVD-audio. For the vast majority of the music sold, anything beyond CD audio is wasted - witness the prevalence of mp3s at 128kb/s, which sound shitty on any reasonable speakers.
/also/ tuned to what's useful for humans, so it'd likely go missing there, too. Are the movie studios going to mandate that all video cameras be built with detectors that can pick up some particular near-infrared frequency, so that they can see the watermark? Are they going to pay the manufacturers the difference in the price due to using that more capable detector?
/everyone/ else to jump through hoops just so that /they/ can make more money? It's ridiculous. It'd be laughable, if they didn't seem to own enough politicians to make it happen.
In order to implement that kind of thing generally they'd have to get
With visual watermarking, again, the technology is carefully designed to match up with the limitation of human senses - you could probably encode something in the ranges outside what we can see, but the response of the ADCs is
This is the thing that has me completely buggered here - how do these people think they can force
himi
WindowMaker is a bloated Windows wannabe??? Have you ever /used/ it? It's a NeXT clone, which is /vastly/ different to Windows. It's also not particularly bloated - it weighs in at 1824kB on my box at the moment.
Afterstep is, I believe, another NeXT clone (note the 'step' in the name).
Gnome/KDE I won't comment on - I don't use them, since I like wmaker.
Please at least try and know what you're talking about when you flame things.
himi
Ummmm . . . No, the type of *foo in char *foo; is a char, and foo is a pointer to a char. There's no notion of forcing anything here - you're /declaring/ the variable, not dereferencing it.
/are/ actually dereferencing, unless you explicitly cast the variable to a different type somewhere along the line the compiler will pick up the mismatched type error. Since C isn't strongly typed, this often gets you a warning and nothing else, but that reflects the fact that C lets you treat a chunk of memory however you want, because it was designed as an OS implementation language where you need direct manipulation of memory. But the language itself /does/ have a strong notion of the type of variables - strong enough that you can do a lot of static analysis. It breaks down around pointers and typecasts because you can throw away information if you choose, but again, the language's notion of types is well defined, with loopholes that make playing games with memory easier.
/is/ meaningful, and that the language properly defines that meaning.
In any case, even when you
I could compile C to Caml (a strongly typed language) if I wanted to, and the only difficult bit would be providing generalised typecasts (Caml doesn't allow casting in the general case, only for things like ints to floats or chars to ints, where there are clear, well defined conversions). Most of the type handling would pass straight through untouched.
Don't make the mistake of looking at the runtime behaviour of a language and thinking that defines it's semantics. In this case, a construct that's entirely meaningful will cause a runtime error if some other stuff isn't done, but that doesn't change the fact that it
himi
Go read K&R on this.
And use 8 space tabs - that makes code blocks more obvious than anything else, and it even warns you when you're getting ridiculously deeply nested control structures.
Another good thing to read is the Linux kernel coding style documentation (Documentation/CodingStyle). It's a good discussion of a lot of this stuff.
himi
Rather, a break in the wrong place - trying to break out of the enclosing if(){}, but actually breaking out of the enclosing switch.
;-)
A clear case where a goto would have been more appropriate
himi
And now
This demonstrates a severe lack of understanding of how real world software is developed. They did an initial implementation of the core, and it worked for what they were interested in at the time. Later experience showed that it
Anyway, if this is so terribly horriffic, maybe you should villify Microsoft - they did exactly the same thing when they started developing NT.
Well, given I get paid to develop Linux software, I really don't know how something like VC++ could ever be the best choice . . .
I
All of these are my personal preferences. I'm not standing up and shouting to the world "You must all use GNU emacs, because Xemacs is evil!" I'm simply saying
I'm not only willing to admit that open source software has a ways to go in many cases before it's as good as commercial software, I'm willing to put time and effort into helping get it there. Or, as is the case with my using BitKeeper, I simply use the superior commercial software.
As for my information not being 100% factual, you're right, I
Not that it would have made any difference - you would most likely still have been trolling, and I'd probably have ended up just ignoring you completely.
You're trolling quite well, for what it's worth. You're just a little too vehement in ignoring what I've actually said, and that's making you look like a fuckwit.
himi
Blindly defending something with a completely uninformed argument? Was I doing that?
I was off a little in the details, but half an hour's research on google came up with more information, which backed up my original claim: the gimp didn't implement CMYK because of patent problems related to it.
Would I have defended a Microsoft product this way? Perhaps, if I'd seen someone trashing it the way you were trashing the gimp. Most likely not, though - I don't use any MS products, nor do I know anything much about any of them, so I'm not likely to step in with an explanation of why something is the way it is.
I have problems with MS's approach to business, based on what I've read about them (including reading through the findings of fact from their current court case). But then, I have problems with many businesses based on my experiences with them. That doesn't make me a zealot, by any stretch of the imagination. If anything, it makes me an informed consumer - something to be proud of, I think.
I realise you're trolling, by the way, so I won't respond to any more comments you make unless there's some real substance to them.
himi
Hokay . . . Having done some research, yes, you're right that CMYK itself isn't patented. There /are/ patents on colour correction, colour management, and various other things related to handling CMY colour for preprint work - this /is/ the reason support hasn't been added to the gimp in the past.
i scussArchive/ msg005514.html
/not/ sub-par - it wouldn't be so popular for web graphics otherwise.
My references:
http://static.userland.com/userLandD
(somewhere down the middle of the page)
http://www.levien.com/gimp/gcmm.html
As for the implementation details, I can't comment - I'm a systems coder, not a graphics person. But I'm fairly sure the gimp's implementation of RGB graphics is
Finally, I'm not a zealot, by a long stretch. I personally prefer free software, but I have many good reasons for this, not least of which is the fact that without free software I wouldn't have gotten interested in coding at all. Does having personal preferences make me a zealot? If so, then you're just as much of one as I am.
himi
Mod the parent up, please.
/should/ be held responsible for your negligence. If you're incompetent, then you should be held responsible for any claims of competence you made. And if you're not willing to accept responsibility for what you do, you shouldn't be doing it.
What is so terribly horrifically frightening about taking reasonable responsibility for your own competence? The same kind of responsibility that an engineer making a component for a car takes, or a builder building a house, or anything else like that?
It's quite simple: do the best job you can, as responsibly as you can, and taking all due care. If something goes wrong after that, then you should be safe from punitive liability. If you're negligent, then you
The only people who can avoid responsibility for their actions are children - is this industry really that immature?
himi
CMYK is patented, and licensing this patent is not at all cheap. Certainly, it's not something that's possible for a piece of software like the gimp.
Claiming that a piece of software is inadequate because the maker of the competing software uses legal means to stop competitors from implementing a piece of functionality is really quite stupid.
himi
Using C++ funky bits makes it much harder to control exactly what code the compiler produces, regardless of whether it's more efficient or not. You also need a runtime library for a lot of things. Those are the two main reasons C++ /isn't/ used.
/everything/, and when the compiler you're targetting is as portable as gcc, it's not even that big an isue for portability.
/not/ like userspace coding. The code tends to look much the same, but you operate under constraints that make the details vastly different, and which make things like C++ difficult propositions.
The control thing probably sounds like a red herring, but in kernel land knowing exactly what your code is doing is extremely important - every cycle really does count. Particularly in the core kernel, you often see strange and confusing constructs that are in there simply because they've been shown to produce faster code. And this isn't some kind of he-man "I can write more gotos than you!" crap - people actually sit down and look at the asm output from bits of code and decide which version is best.
That's the main reason why the kernel is so conservative about what compilers it builds on - there's code in there that relies on specific behaviour of gcc extensions to achieve the exact results desired. This probably sounds like a bad thing, but this is an OS kernel - performance is
Kernel coding is
himi
See here
.
Not in depth theory, but an excellent explanation . .
himi
Write your own article outlining all this and giving your analysis of what's going on, and submit /that/ to /. - you may have more luck getting heard that way . . .
;-)
Lots of interesting quotes, by the way - kudos
himi
Damn good point there - MS has been criminally negligent, and they've hidden behind the lack of liability clauses they put in their EULAs for too long. It's the same with other software companies, but MS's screwups are writ extra large because of their dominance.
Someone needs to take MS to court charged with negligence, and put an end to their arrogant assumption that they can do whatever the fuck they please, because they say they can.
himi
The whole point of unions is forcing employers to give reasonable compensation (in money and in other ways) for the work their employees do for them. But it's generally not a one way street - the employees have to commit to things, too, even if it's only an implicit "Give me this standard of wages and conditions, and I'll work for you without complaint".
/good/, for everyone.
Unions like the AMA do more than that - they require standards from the employees, as well, standards to protect both sides. You get union protection, but only if you meet union standards. That probably sounds as terrifying and horrible as taking due responsibility for flaws in your software, but it's entirely reasonable in a skilled profession where your responsibilities are high. If you claim to be able to do something, you should be responsible for the results if your claim was false.
Historically, unions have done some screwed things. They've also made more and bigger changes to the standard of living in many countries than anything else in the last five hundred years. Get the implementation right, and they're
himi
The other thing about symlinks is that they can point at a target on a different filesystem - hardlinks can only refer to a file on the same filesystem.
/usr, /usr/local, /home, etcetera are often seperate filesystems, and being able to link between them is useful.
Given the *ahem* limitations Windows in general has with handling multiple filesystems transparently, that probably doesn't come up much at all in Windows, but it's very very useful on *nix systems, where
The other use that comes to mind is with shared libraries - the basic file typically has a name like libfoo.so.1.2.3.4, indicating the library version. This will generally be accompanied by symlinks that point to it, with names like libfoo.so.1, libfoo.so.1.2, and so on - this facilitates having multiple versions of the same library installed and allowing programs to link to specific versions. Rather than link to libfoo.so, if it needs libfoo.so.1.2 it'll link to that pathname, and get the latest version of libfoo with that major/minor number. The details are all handled by ldconfig, which manages the links.
Symlinks are terribly nifty, and very very useful - I was rather shocked when I discovered that NTFS didn't support them, because they're simple both conceptually and implementation wise (I know - I've implemented filesystems with symlink support).
himi
Actually, Bonobo is based around a CORBA implementation, and I believe KParts is too. If implementing a well defined and open standard is copying, then there's your answer . .
Mono is a reimplementation of the
Cut and pasted from the comment at the head of linux/kernel/module.c:
/*
* Originally by Anonymous (as far as I know...)
* Linux version by Bas Laarhoven
* 0.99.14 version by Jon Tombs
* Heavily modified by Bjorn Ekwall May 1994 (C)
* Rewritten by Richard Henderson Dec 1996
* Add MOD_INITIALIZING Keith Owens Nov 1999
* Add kallsyms support, Keith Owens Apr 2000
* Add asm/module support, IA64 has special requirements. Keith Owens Sep 2000
* Fix assorted bugs in module verification. Keith Owens Sep 2000
* Fix sys_init_module race, Andrew Morton Oct 2000
* http://www.uwsg.iu.edu/hypermail/linux/kernel/000
* Replace xxx_module_symbol with inter_module_xxx. Keith Owens Oct 2000
* Add a module list lock for kernel fault race fixing. Alan Cox
*
* This source is covered by the GNU GPL, the same as all kernel sources.
*/
So you see, it predates Windows 95 by a few years.The earliest date mentioned is '94, but it's obviously older than that - could it be that it predates even Windows NT?
As for "MANY, MANY" applications still being statically linked, I'd like to see some evidence for that. Certainly, there are
simon@caccini:~/hacking/linux/kernel$ ldd
libc.so.6 =>
I'll accept that you may be knowledgeable about Windows, but you seem somewhat lacking in knowledge of Linux . .
Finally, something on-topic:
Well, what I was taught in my CS courses was that the idea behind modularity was that it let you
It'd be nice if someone involved with this case made a point of bringing
himi
Generally speaking, they /have/ to be - they'd get screamed at insanely if they weren't. However, I'll point at various incidents like the NT service pack that caused Lotus Notes to stop working, and their skullduggery about windows 3.1 on DR-DOS. But yes, in their current position they're limited by the weight of software that they need to support, so there's a limit to what they can change.
/does/ have monopoly power, and it /does/ abuse it in various ways. Arguing otherwise is just plain stupid (or, in the case of Mr Reflection here, trolling).
I believe changing the OS interfaces so that Navigator wouldn't run anymore was one of the things they considered when they first looked at killing Netscape. I'm not sure if they decided they couldn't do that without breaking other things, though.
My other points stand unaltered. Microsoft
himi