If your physical security is so weak that anybody can just go up to your server and plug a card in, *or take one out* does it matter much about the IT security aspect?
Actually, yes!
Most people will not plug a card in. But it is not that unusual for people to probe random ip-addresses and infect you with trojans, or other forms of remote attacks. If you want to keep the server running, you'd better have some security there.
Remember that most servers aren't completely unsecured physically, they generally are inside some room in some building where most people don't just happen to walk by (and most of those passing by will be employees or people associated with the company in other ways, so they can for the most part be trusted to some degree). On the internet, the server is available to every person on the planet! So even if it isn't really secured physically, there is at least less chance of a physical than remote attack.
Compare a house in the city with a house on the countryside. Now, anyone can still get to the house on the country, but there will not be so many random bypassers, so the security is higher, even though it doesn't have better locks or alarm-systems.
This won't let me run MacOS X graphical applications.
Yes it will. GnuSTEP is an attempt at recreating the API's from NeXT. Mac OS X's Cocoa is also based on the NeXT API's. (Erm, actually, both are based on the Openstep spec, but that would never have existed, nor would anyone care about it if it wasn't for the NeXT). No doubt GnuSTEP will include whatever extensions to Openstep Apple has made once it get's a little bit more stable. (Quartz is just a fancy name for Display Postscript, also an idea from the NeXT, and also part of the GnuSTEP project).
wouldn't it be interesting for people like Trolltech and the like to emulate Quartz/Cocoa/Carbon?
I'm not so sure why it would be beneficial for a TrollTech to make Quartz/Cocoa/Carbon libraries for the x86 (or other platforms). It would be a lot of work, it competes directly with their own product, and let's face it: There are not that many apps for the new Mac OS X yet.
Quark and Adobe wouldn't need much more than a recompile on their apps to port them over to Linux/FreeBSD, or am I missing something?
Yes, it could be beneficial for Quark or Adobe (or other application vendors) to make Quark/Cocoa (and maybe even Carbon) libraries on other platforms, if they choose to tie their new applications to Apple OS X. If OS X ever gets popular, maybe some bigname companies will choose to sponsor the GnuSTEP project to get their apps ported. As for now, I wouldn't hold my breath.
While technically true, the difference is really not that much important in practice, as there is a lot of other problems involved in building a good garbage collector (and for any practical program and garbage collector, the size of the live objects would be of the same order as the size of memory requested from the operating system).
A copying garbage collector is actually guaranteed to allocate at least twice as much memory from the operating system then is currently occupied by live objects. Otherwise, it couldn't copy them. Now, imagine what that does for locality of memory, and cache efficiency (This is why Appel's paper comparing gc and stack allocation haven't made most people abandoning the stack just yet).
On the other hand, a mark-sweep garbage collector never moves objects, resulting in external fragmentation (also true of malloc/free), which is also bad for locality of memory and cache efficiency.
You can also use a compacting garbage collector, which works mostly like a mark-sweep garbage collector, but occasionally compacts fragmented memory as well. The downside is complexity (and thus also performance of the compacting phase).
In practice, the problem of external fragmentation isn't really much of a problem with a good allocator using the buddy-system. At least, almost any non-contrived application or benchmark should have less than 10% fragmentation using a good allocator. But that still leaves the problem of having related objects occupy nearby memory locations open (which is needed for good cache efficiency).
Copying garbage collection can be used for that purpose. But if you are using the simple two-space algorithm, you'll find that you are doing the exact opposite of that, the two-space algorithm tends to put related objects far away from each other, due to the depth-first search it uses to find live objects.
So we need another more complex copying collector... But that will again slow down the performance of the collector.
One possible performance neck of a mark-sweep collector is that it needs two passes: one to mark live memory, and one to collect garbage. But the copying collector is no better, first you need to traverse live objects, then you need to copy them somewhere else.
While I still haven't touched on the subject of generational and incremental garbage collection, I think you will already see that writing a good garbage collector is far from trivial. There are no simple cookbook approaches for achieving optimal performance. The only reasonable approach is experimentation, testing, and fine-tuning of parameters. Most real-life garbage collectors use a hybrid of different techniques for different types of objects (distinguished by such attributes as age, size, mutability, etc...)
Modifying a car requires tools and skills very different in nature from modifying software
Modifying software requires tools and skills very different in nature from moidyfing a car
Basically, to modify software, you need a basic knowledge of programming, common algorithms and datastructures, and a programming language, although you will get better with more knowledge. The tools you need are a computer, text-editor and a compiler/interpreter, but other tools can also come in helpful.
To modify a car, you need basic knowledge of how a car works, some knowledge of common problems and symptoms, and some skill and technique to apply this to fixing the car, although you will get better with more knowledge. Useful tools would be a set of wrenches and some spare parts, although other tools can also come in helpful.
I hardly see any difference at all. Saying that authors of software that comes with source code should not be liable for faults with the software is just like saying that producers of cars coming with a repair-manual, should not be liable for faults with the car.
No, it is definitely not easy to make a reasonable compromise between protecting stupid customers against the irresponsible faults of the worlds largest software vendor, and protecting individual programmers and small software companies from legal harassment, but the line has to be drawn somewhere entirely elsewhere than whether you as a customer can fix the problem yourself. A good first approximation would be a money-back warranty for problems deemed worse enough by some committee, but even there worms pop up as you try to settle the details.
Well, I find it quite reasonable that airline companies are not held liable for the horrors of september 11.
This was something very few people expected happening, and the costs and inconveniences for customers would have made the airline bankrupt before you could say "bankrupt", if it wasn't mandatory for everyone. And I certainly don't remember any outbursts from the public wanting more security at the airports and airlines.
Does that mean that I think the government should have made it mandatory? Maybe, but I don't have the numbers (or enough interest to get them) that was available to security analysts before the incident to see whether this was really something one could expect happening (I suspect it wasn't, though, and that securing against plains crashing for non-terrorist reasons was seen as more cost-efficient).
Is what the US government has done in retrospect sufficient? Well, they have secured airports and airlines better, so i'd have to say yes. Is it required? Well, they have started a war, made a secret, military, no-appeals, orwellian, right-to-judge-you-and-take-your-life-anywhere-in- the-world-"court", and made inhumane prisoner camps violating the geneva convention for prisoners of war. I fail to see how this will not generate more terrorists, but that might be just me...
Now, I do like Solaris, but never found the x86 implementation to be worth the trouble. Since Solaris is mostly source compatible with Linux or *BSD anyway, the point is moot.
If you want a real machine, buy a sparc, otherwise just keep using the free unices on x86...
Thanks to the heavy ocean cover and relatively sparse city covering of the land, odds are we'll get hit in a nice relatively non-fatal location before a city-buster earns its name.
Well, from what I can imagine, we would be much more lucky if the asteroid hit a big city, than if it hit the ocean. Imagine the waves! Now consider the fact that most people live somewhat close to the sea...
I think a large unpopulated land-based area such as Antarctica might be the best. An additional benefit might be that most of the stuff that would be thrown into the atmoshpere would be ice instead of dirt, thus not totally blocking the sun.
But of course, I'm just speculating, I have neither the time, interest or knowledge to do real calculations.
...And that is you!
It is fortunately not uncommon that different people have different preferences. That's why asking others will often result in different people giving you different answers.
Whether you prefer "full spectrum lighting", "blacklight", normal lightbulbs, "warm" lightbulbs, darkness, daylight through the window, stroboscope, or any other form of lighting is up to you.
If you think "full spectrum lighting" will make you happy, try it, and see what you think. End of story.
Not really.
I've seen a lot of "embedded" webservers and webservers running on old hardware really. And when you think of it, well how hard can a minimal tcp/ip-stack and minimal implementation of http be?
Besides, the Apple Lisa has more than enough RAM for such a task (512kB), there's room for both a real tcp/ip stack and a real webserver without having to wrestle for space. And I am sure it has a serial-port you can run PPP over (which is a really simple protocol, if you choose to implement only what you need).
And, as someone already has mentioned earlier in this thread, the Lisa mentioned is so upgraded that it is no longer really a Lisa. Which makes it even less impressive.
Obviously he did. You, however have obviously not read the article. Go read it again, and find out why he is talking about an iMac. Then try to think before you post next time.
Correct me if I'm wrong, but wasn't the Open Source software model founded by that hippie, shoeless slob Richard Stallman? If open source didn't exist (don't know how anyone can claim they "invented" OSS anyway), how could have Linus copylefted the Linux kernel in the first place?
You are wrong so I correct you.
The open source software model was founded by that limpy, gun-fanatic, geek Eric Steven Raymond, when he wrote the paper "The Cathedral and the Bazaar". While Eric didn't invent the open source model, he put into words and clarified what had by that time already become a large philosophy for a large group of programmers.
The sholess slob hippie Richard Stallman is the creator of the Free Software Movement, which has a completely different philosophy. Of course, free software existed long before Richard Stallman started his evangelizing, but Richard put into words what many people felt, and decided to do something about it, and thus was the gnu project started and the free software foundation founded.
Since copyleft is a phrase invented by Richard Stallman, it has absolutely nothing to do with open source, so the fact that Linus copylefted his linux-kernel is not of particular interest in discussing the origins of open source.
But the fact that linux is freely available with source code, and that it created a large following of programmers and users, is relevant. The fact that linux was free and popular eventually resulted in what is now known as the open source movement, first observed and described in detail by Eric Raymond.
As you will remember, Fermat claimed in the margin of some notes that he had a wonderful proof, but unfortunately it wouldn't fit there.
By the time this was written, it would count as proof by intuition and proof by obviousness, but this view has shifted through the times, and today the lost-reference and authority proof-techniques are viewed as the most strikingly elegant touches in this proof.
Short-sightedness on the part of our story poster here. I think natural language processing, or at least work in its direction is "real work".
Nope, this has nothing to do with natural language processing. If you had read the manual you would understand that as well. It is a fairly simple language quite similar to the original basic-dialects, with a lot of syntactic sugar. Adding syntactic sugar has very little to do with natural language processing.
But I'm sure it was fun, and that they learned something in the process of creating it, so I wouldn't call it worthless...
Well, technically, a tool that translates from one language to another (such as C to x86 object code) should really be called a translator. Compiling is the craft of putting those peaces together, so we should really start calling our linkers compilers, and our compilers translators.
But if we are to accept common (albeit faulty) terminology, I will agree that something that translates language foo to C is definitely worthy of being called a compiler...
Doesn't matter at all. If Dolby has patented it, there is nothing you can do. A cleanroom implementation is used to avoid problems with copyright, not patents.
Unfortunately, I don't know much about either what AC3 is all about, or what the patents claim, but a quick patent search on delphion showed two patents related to it.
Actually, flash isn't that bad, but it fails in the sense of being more complicated than HTML for the beginner.
I rather like my semantic web, thank you very much.
So do I, I'm sure there are some useful uses for it. I just don't think that it should be the only solution for everyone.
Did it ever occur to you that maybe the web isn't useful to anybody because nobody has anything useful to say?;)
Yes. Did I stick with the notion? No, I find that people often have interesting viewpoints in real-life, so a general planet-wide lack of brain-activity cannot be the reason for the apparent lack of intelligent content on the web.
While I agree with what you are saying in theory, we have presentation languages (ie. postscript, pdf) that would do quite well for transmitting your data over the web.
Yes, postscript and pdf are excellent page-description languages for printers. But they are not intended for dynamic interactive web-content. Flash and Curl is, however.
Why don't you go bark at adobe for implementing linking in a braindead way in their PDF plugin, since that alone is why it is unsuitable as a web presentation language?
Well, I don't tend to bark at people for random reasons, and can't understand why anyone would want me to. Besides, the major reason pdf is unsuitable as a web presentation language is because it is a paper presentation language, and those two are different notions. Imho, the linking was an ill-thought-out feature that could only have come from the marketing department anyway. I'm still surprised somebody actually tries to use it.
Nope, it's what we have been wanting all along. While some content (e.g. books, manuals, newspaper articles, etc) is suitable for viewing in different formats, on paper, on the web, or read loud, there also exists lot's of reasons for creating media-specific content. Do you want every new movie to be available in MOVXML so that you can automatically generate actor scripts, lighting instructions, animated scenes, and what else, but never actually enjoy the actors playing on the scene, because the format the movie was delivered in was content-based instead of presentation based (after all, actors are just presentation)?
If there is one thing the web needs right now, it is freedom from the vision of the semantic web, and all those other idiotic visions keeping us from progressing further. There is nothing wrong with the semantic web per se, but it is not, and should not be the solution for everybody's needs. If the people at W3C had realized this when they started, instead of making overly complicated standards trying to catering for everyone's need, we might even have had a useful web by now. A simple presentation language for people wanting to describe layout precisely, and a still-simple HTML-variant for describing content. If you were worried about both, you would be fucked of course, but very few people are.
If you are still not convinced, let's look at a few examples:
Slashdot and other discussion-based sites:
Presentation language is needed, nobody prints out slashdot anyway.
Shopping sites, online banks, intranet usage with forms, etc...:
ditto
Online comics:
ditto
Gutenberg project:
HTML
Newspapers:
presentation based, with HTML access for a subscription
...etc. The layout/content separation theory is good in theory, but I think it's on time to start liberating ourselves a bit from it now. It has been over 10 years with the web, and still very few are happy with the way it works.
Well, why a client-side web-language should need to offer database-connectivity in order to seem mature is beyond me. Would you use database-connectivity in javascript, if it had any? These things should be done on the server side, and you know it.
Whether or not it's mature? Well, I don't know about stability (probably not much worse than most other shit available on the web), but the features it offers seems to be just the right thing.
The web today does still consist of static pages. I do not consider applets or animated gifs an exception to this rule, because their interaction with their environment is close to nada. And javascript plainly suck, we all know it is good for nothing. Have you ever tried to write tick-tack-toe in javascript to be displayed in a browser window?
Curl does the right thing, it is a simple to use, easy to program, presentation language for the web. All those standards from W3C may be fine and dandy, but they still have some problems when it comes to the real world:
They are for describing content, not presentation.
They have lot's of historical baggage.
There are too many of them, and it's starting to get really complicated.
None of them has implementations that are even remotely capable of doing what Curl do.
I'd say curl is a great idea, and probably something it would be worthwile for the free software community to look into. But this fucking corporate licensing "give your first unborn son to us" stuff, makes anyone interested go away. Considering that curl originally started with a DARPA-grant, it's really sad to see it going down for no reason except corporate greediness.
Wake up. One single open source project is not going to kill all other open source projects. The fact that.NET still makes one hell of a development environment remains, whether we will have passport or not. True, passport remains important, but it's not going to be the end of all e-commerce solutions, and even if it is, it's not going to kill all open source (or free software) applications.
Yes, but regex pattern matching has nothing to do with pattern matching when speaking of (functional) computer languages.
Instead it refers to a convenient mechanism for building switch-like statements over algebraic datatypes. Since the language in the task is described by a grammar, it can be represented as an algebraic datatype in a functional language.
In Java-programmer-friendly lingo, this would mean: Make one constructor (differently named) for each element in the grammar. That way you can build a tree from the text you parse. Then you can make recursive functions, either modifying or returning a new tree, where you will typically use switch-statements on the data, differentiating each elemeent by the constructor used to create it.
Most functional languages are quite convenient when working with data represented in a tree-form, because algebraic datatypes are so extremely convenient. In Java, one would have to use classes and subclasses instead of one simple definition of an algebraic datatype, and lot's of casting and if-statements instead of pattern matching. It's not that it can't be done, it is just a bit more to write (which happens to be somewhat important when having limited time available, but certainly not all-important, if you type fast).
Having something that senses if I start to get drowsy and warning me wouldn't be to bad.
It doesn't even have to be annoying. A loud buzz if you close your eyes for half a second or more. That should be enough to remind you that taking a break and stretching your legs a little would be a smart move.
However, the idea of a "virtual passenger" really offends me. I don't want another Eliza to offend my stupidity (or intelligence for that matter).
They should make it simple, and something that works for responsible drivers. If they have to splash someone in the eyes to wake him up, it's already far too late, and no safety system on the planet, except perhaps something taking control of the car, could help.
And for those who don't know what lutefisk is all about. It is a traditional scandinavian dish which is eaten at most once a year, but everyone praises like it was the best gourmet food ever.
It consists of fish, that is first dried, then put into lye for a few months, and finally put inside a stream to remove the lye (and any remaining nutricients), and finally boiled. It looks pretty much like jelly, feels pretty much like jelly, and tastes pretty much like jelly, except that it doesn't have strawberry flavour added.
It is usually eaten somewhere in the christmas holiday, and is served with potatoes, bacon, peas, mustard, pepper, and anything else that has some taste. Few people really knows what it tastes like, as they have never really tasted it without all the add-ons and mix-ins. Yet they continue to praise it and still refuse to eat it more than once in a year.
Personally, I'm quite fond of jelly, and don't see what all the fuss people make about this fish is all about. It is neither as bad or as good as some people would have it.
They'll
help Mono until.NET is widely adapted, at that time,
they drop any support for non-Microsoft OSes and add
proprietary undocumented functions, just like they
always did -- and if you intend to continue using.net stuff, you have to "upgrade" to Windows.
Of course, but does it matter? No, the open source community will in the meantime have gained most of the functionality of the.NET platform, and can happily move on in what direction pleases them. There is no fundamental reason for Mono to be exactly like.NET (although it would be useful), but having something mostly like.NET will be a huge gain for open source developers everywhere. Just like gcc and emacs is. IMHO, this is what would benefit both most.
Slashdot Mirror
On the other hand, I've never felt the need to see the movie I rented twice. Instead I watch another movie, so I don't care much about it...
Actually, yes!
Most people will not plug a card in. But it is not that unusual for people to probe random ip-addresses and infect you with trojans, or other forms of remote attacks. If you want to keep the server running, you'd better have some security there.
Remember that most servers aren't completely unsecured physically, they generally are inside some room in some building where most people don't just happen to walk by (and most of those passing by will be employees or people associated with the company in other ways, so they can for the most part be trusted to some degree). On the internet, the server is available to every person on the planet! So even if it isn't really secured physically, there is at least less chance of a physical than remote attack.
Compare a house in the city with a house on the countryside. Now, anyone can still get to the house on the country, but there will not be so many random bypassers, so the security is higher, even though it doesn't have better locks or alarm-systems.
Yes it will. GnuSTEP is an attempt at recreating the API's from NeXT. Mac OS X's Cocoa is also based on the NeXT API's. (Erm, actually, both are based on the Openstep spec, but that would never have existed, nor would anyone care about it if it wasn't for the NeXT). No doubt GnuSTEP will include whatever extensions to Openstep Apple has made once it get's a little bit more stable. (Quartz is just a fancy name for Display Postscript, also an idea from the NeXT, and also part of the GnuSTEP project).
wouldn't it be interesting for people like Trolltech and the like to emulate Quartz/Cocoa/Carbon?
I'm not so sure why it would be beneficial for a TrollTech to make Quartz/Cocoa/Carbon libraries for the x86 (or other platforms). It would be a lot of work, it competes directly with their own product, and let's face it: There are not that many apps for the new Mac OS X yet.
Quark and Adobe wouldn't need much more than a recompile on their apps to port them over to Linux/FreeBSD, or am I missing something?
Yes, it could be beneficial for Quark or Adobe (or other application vendors) to make Quark/Cocoa (and maybe even Carbon) libraries on other platforms, if they choose to tie their new applications to Apple OS X. If OS X ever gets popular, maybe some bigname companies will choose to sponsor the GnuSTEP project to get their apps ported. As for now, I wouldn't hold my breath.
A copying garbage collector is actually guaranteed to allocate at least twice as much memory from the operating system then is currently occupied by live objects. Otherwise, it couldn't copy them. Now, imagine what that does for locality of memory, and cache efficiency (This is why Appel's paper comparing gc and stack allocation haven't made most people abandoning the stack just yet).
On the other hand, a mark-sweep garbage collector never moves objects, resulting in external fragmentation (also true of malloc/free), which is also bad for locality of memory and cache efficiency.
You can also use a compacting garbage collector, which works mostly like a mark-sweep garbage collector, but occasionally compacts fragmented memory as well. The downside is complexity (and thus also performance of the compacting phase).
In practice, the problem of external fragmentation isn't really much of a problem with a good allocator using the buddy-system. At least, almost any non-contrived application or benchmark should have less than 10% fragmentation using a good allocator. But that still leaves the problem of having related objects occupy nearby memory locations open (which is needed for good cache efficiency).
Copying garbage collection can be used for that purpose. But if you are using the simple two-space algorithm, you'll find that you are doing the exact opposite of that, the two-space algorithm tends to put related objects far away from each other, due to the depth-first search it uses to find live objects.
So we need another more complex copying collector... But that will again slow down the performance of the collector.
One possible performance neck of a mark-sweep collector is that it needs two passes: one to mark live memory, and one to collect garbage. But the copying collector is no better, first you need to traverse live objects, then you need to copy them somewhere else.
While I still haven't touched on the subject of generational and incremental garbage collection, I think you will already see that writing a good garbage collector is far from trivial. There are no simple cookbook approaches for achieving optimal performance. The only reasonable approach is experimentation, testing, and fine-tuning of parameters. Most real-life garbage collectors use a hybrid of different techniques for different types of objects (distinguished by such attributes as age, size, mutability, etc...)
Basically, to modify software, you need a basic knowledge of programming, common algorithms and datastructures, and a programming language, although you will get better with more knowledge. The tools you need are a computer, text-editor and a compiler/interpreter, but other tools can also come in helpful.
To modify a car, you need basic knowledge of how a car works, some knowledge of common problems and symptoms, and some skill and technique to apply this to fixing the car, although you will get better with more knowledge. Useful tools would be a set of wrenches and some spare parts, although other tools can also come in helpful.
I hardly see any difference at all. Saying that authors of software that comes with source code should not be liable for faults with the software is just like saying that producers of cars coming with a repair-manual, should not be liable for faults with the car.
No, it is definitely not easy to make a reasonable compromise between protecting stupid customers against the irresponsible faults of the worlds largest software vendor, and protecting individual programmers and small software companies from legal harassment, but the line has to be drawn somewhere entirely elsewhere than whether you as a customer can fix the problem yourself. A good first approximation would be a money-back warranty for problems deemed worse enough by some committee, but even there worms pop up as you try to settle the details.
This was something very few people expected happening, and the costs and inconveniences for customers would have made the airline bankrupt before you could say "bankrupt", if it wasn't mandatory for everyone. And I certainly don't remember any outbursts from the public wanting more security at the airports and airlines.
Does that mean that I think the government should have made it mandatory? Maybe, but I don't have the numbers (or enough interest to get them) that was available to security analysts before the incident to see whether this was really something one could expect happening (I suspect it wasn't, though, and that securing against plains crashing for non-terrorist reasons was seen as more cost-efficient).
Is what the US government has done in retrospect sufficient? Well, they have secured airports and airlines better, so i'd have to say yes. Is it required? Well, they have started a war, made a secret, military, no-appeals, orwellian, right-to-judge-you-and-take-your-life-anywhere-in- the-world-"court", and made inhumane prisoner camps violating the geneva convention for prisoners of war. I fail to see how this will not generate more terrorists, but that might be just me...
Ok, this was probably off-topic.
If you want a real machine, buy a sparc, otherwise just keep using the free unices on x86...
Well, from what I can imagine, we would be much more lucky if the asteroid hit a big city, than if it hit the ocean. Imagine the waves! Now consider the fact that most people live somewhat close to the sea...
I think a large unpopulated land-based area such as Antarctica might be the best. An additional benefit might be that most of the stuff that would be thrown into the atmoshpere would be ice instead of dirt, thus not totally blocking the sun.
But of course, I'm just speculating, I have neither the time, interest or knowledge to do real calculations.
Whether you prefer "full spectrum lighting", "blacklight", normal lightbulbs, "warm" lightbulbs, darkness, daylight through the window, stroboscope, or any other form of lighting is up to you.
If you think "full spectrum lighting" will make you happy, try it, and see what you think. End of story.
Besides, the Apple Lisa has more than enough RAM for such a task (512kB), there's room for both a real tcp/ip stack and a real webserver without having to wrestle for space. And I am sure it has a serial-port you can run PPP over (which is a really simple protocol, if you choose to implement only what you need).
And, as someone already has mentioned earlier in this thread, the Lisa mentioned is so upgraded that it is no longer really a Lisa. Which makes it even less impressive.
Obviously he did. You, however have obviously not read the article. Go read it again, and find out why he is talking about an iMac. Then try to think before you post next time.
You are wrong so I correct you.
The open source software model was founded by that limpy, gun-fanatic, geek Eric Steven Raymond, when he wrote the paper "The Cathedral and the Bazaar". While Eric didn't invent the open source model, he put into words and clarified what had by that time already become a large philosophy for a large group of programmers.
The sholess slob hippie Richard Stallman is the creator of the Free Software Movement, which has a completely different philosophy. Of course, free software existed long before Richard Stallman started his evangelizing, but Richard put into words what many people felt, and decided to do something about it, and thus was the gnu project started and the free software foundation founded.
Since copyleft is a phrase invented by Richard Stallman, it has absolutely nothing to do with open source, so the fact that Linus copylefted his linux-kernel is not of particular interest in discussing the origins of open source.
But the fact that linux is freely available with source code, and that it created a large following of programmers and users, is relevant. The fact that linux was free and popular eventually resulted in what is now known as the open source movement, first observed and described in detail by Eric Raymond.
This page will give you a clue...
As you will remember, Fermat claimed in the margin of some notes that he had a wonderful proof, but unfortunately it wouldn't fit there.
By the time this was written, it would count as proof by intuition and proof by obviousness, but this view has shifted through the times, and today the lost-reference and authority proof-techniques are viewed as the most strikingly elegant touches in this proof.
Nope, this has nothing to do with natural language processing. If you had read the manual you would understand that as well. It is a fairly simple language quite similar to the original basic-dialects, with a lot of syntactic sugar. Adding syntactic sugar has very little to do with natural language processing.
But I'm sure it was fun, and that they learned something in the process of creating it, so I wouldn't call it worthless...
But if we are to accept common (albeit faulty) terminology, I will agree that something that translates language foo to C is definitely worthy of being called a compiler...
Unfortunately, I don't know much about either what AC3 is all about, or what the patents claim, but a quick patent search on delphion showed two patents related to it.
So I don't think a cleanroom implementation would help at all. But there could be a way to work around the patents...
Actually, flash isn't that bad, but it fails in the sense of being more complicated than HTML for the beginner.
I rather like my semantic web, thank you very much.
So do I, I'm sure there are some useful uses for it. I just don't think that it should be the only solution for everyone.
Did it ever occur to you that maybe the web isn't useful to anybody because nobody has anything useful to say? ;)
Yes. Did I stick with the notion? No, I find that people often have interesting viewpoints in real-life, so a general planet-wide lack of brain-activity cannot be the reason for the apparent lack of intelligent content on the web.
While I agree with what you are saying in theory, we have presentation languages (ie. postscript, pdf) that would do quite well for transmitting your data over the web.
Yes, postscript and pdf are excellent page-description languages for printers. But they are not intended for dynamic interactive web-content. Flash and Curl is, however.
Why don't you go bark at adobe for implementing linking in a braindead way in their PDF plugin, since that alone is why it is unsuitable as a web presentation language?
Well, I don't tend to bark at people for random reasons, and can't understand why anyone would want me to. Besides, the major reason pdf is unsuitable as a web presentation language is because it is a paper presentation language, and those two are different notions. Imho, the linking was an ill-thought-out feature that could only have come from the marketing department anyway. I'm still surprised somebody actually tries to use it.
If there is one thing the web needs right now, it is freedom from the vision of the semantic web, and all those other idiotic visions keeping us from progressing further. There is nothing wrong with the semantic web per se, but it is not, and should not be the solution for everybody's needs. If the people at W3C had realized this when they started, instead of making overly complicated standards trying to catering for everyone's need, we might even have had a useful web by now. A simple presentation language for people wanting to describe layout precisely, and a still-simple HTML-variant for describing content. If you were worried about both, you would be fucked of course, but very few people are.
If you are still not convinced, let's look at a few examples:
...etc. The layout/content separation theory is good in theory, but I think it's on time to start liberating ourselves a bit from it now. It has been over 10 years with the web, and still very few are happy with the way it works.
Whether or not it's mature? Well, I don't know about stability (probably not much worse than most other shit available on the web), but the features it offers seems to be just the right thing.
The web today does still consist of static pages. I do not consider applets or animated gifs an exception to this rule, because their interaction with their environment is close to nada. And javascript plainly suck, we all know it is good for nothing. Have you ever tried to write tick-tack-toe in javascript to be displayed in a browser window?
Curl does the right thing, it is a simple to use, easy to program, presentation language for the web. All those standards from W3C may be fine and dandy, but they still have some problems when it comes to the real world:
I'd say curl is a great idea, and probably something it would be worthwile for the free software community to look into. But this fucking corporate licensing "give your first unborn son to us" stuff, makes anyone interested go away. Considering that curl originally started with a DARPA-grant, it's really sad to see it going down for no reason except corporate greediness.
Wake up. One single open source project is not going to kill all other open source projects. The fact that .NET still makes one hell of a development environment remains, whether we will have passport or not. True, passport remains important, but it's not going to be the end of all e-commerce solutions, and even if it is, it's not going to kill all open source (or free software) applications.
Instead it refers to a convenient mechanism for building switch-like statements over algebraic datatypes. Since the language in the task is described by a grammar, it can be represented as an algebraic datatype in a functional language.
In Java-programmer-friendly lingo, this would mean: Make one constructor (differently named) for each element in the grammar. That way you can build a tree from the text you parse. Then you can make recursive functions, either modifying or returning a new tree, where you will typically use switch-statements on the data, differentiating each elemeent by the constructor used to create it.
Most functional languages are quite convenient when working with data represented in a tree-form, because algebraic datatypes are so extremely convenient. In Java, one would have to use classes and subclasses instead of one simple definition of an algebraic datatype, and lot's of casting and if-statements instead of pattern matching. It's not that it can't be done, it is just a bit more to write (which happens to be somewhat important when having limited time available, but certainly not all-important, if you type fast).
It doesn't even have to be annoying. A loud buzz if you close your eyes for half a second or more. That should be enough to remind you that taking a break and stretching your legs a little would be a smart move.
However, the idea of a "virtual passenger" really offends me. I don't want another Eliza to offend my stupidity (or intelligence for that matter).
They should make it simple, and something that works for responsible drivers. If they have to splash someone in the eyes to wake him up, it's already far too late, and no safety system on the planet, except perhaps something taking control of the car, could help.
And for those who don't know what lutefisk is all about. It is a traditional scandinavian dish which is eaten at most once a year, but everyone praises like it was the best gourmet food ever.
It consists of fish, that is first dried, then put into lye for a few months, and finally put inside a stream to remove the lye (and any remaining nutricients), and finally boiled. It looks pretty much like jelly, feels pretty much like jelly, and tastes pretty much like jelly, except that it doesn't have strawberry flavour added.
It is usually eaten somewhere in the christmas holiday, and is served with potatoes, bacon, peas, mustard, pepper, and anything else that has some taste. Few people really knows what it tastes like, as they have never really tasted it without all the add-ons and mix-ins. Yet they continue to praise it and still refuse to eat it more than once in a year.
Personally, I'm quite fond of jelly, and don't see what all the fuss people make about this fish is all about. It is neither as bad or as good as some people would have it.
Then you must have a really short memory, or be very young. Because that was what all websites were saying back before IE became popular.
Of course, but does it matter? No, the open source community will in the meantime have gained most of the functionality of the .NET platform, and can happily move on in what direction pleases them. There is no fundamental reason for Mono to be exactly like .NET (although it would be useful), but having something mostly like .NET will be a huge gain for open source developers everywhere. Just like gcc and emacs is. IMHO, this is what would benefit both most.