This is true of KDE as much as QT. In hindsight, the KDE developers would have probably chosen a different license, but now that it's popular, the QT and KDE license don't work so well together, so goes the argument.
The real problem is that one apparently needs a savvy lawyer right from the start, but none of these projects have the money to make this happen. Why should I need a lawyer to do free software? This is the real problem with the GPL and its intricacies. It needs to be either clearer or more flexible.
As other commenters have notes, the GPL is based on copyright law, and so whether software is distributed will depend on how copyright law determines whether a person or business made a copy from another person or business.
But the linking clause is really hard to figure out. Any code depends on abstractions (e.g., specifed by a header file, admittedly partially), and the code doesn't really care whether the implementation of that abstraction (e.g., a library) is GPLed or not. Actually, the header file might be GPLed, too, but the code doesn't care about that either. foo.h might or might not be GPLed.
And, of course, a script links programs together in a different way.
My view is that the LGPL is a viable way of looking at the software world. The GPL wants to limit usage only by other GPL code, but again code doesn't really care about that. I guess RMS sure cares and that is why not much change can be expected.
Re:This could be the most important net developmen
on
ICANN Board Election
·
· Score: 1
PacketMaster states the case a little too strongly, but there is no reason why every/.er shouldn't register to participate in how the internet will be run.
Currently, ICANN is thinking about adding new genetic TLDs (in addition to.com, maybe add.web,.xxx, etc.). ICANN is also trying to figure out how to elect at-large members. Although the article doesn't state it, the most onerous of the proposed rules is a 10% requirement for putting a person on the ballot. Almost every public comment states that 10% is too high. The current board meets in mid-July. We'll see what happens.
So join and participate in how the internet is going to look like in the future. My feeling is that addresses like word.com will eventually fall by the wayside, and we will need to access web sites by more conventional information (name of company, country, product, etc.). This would like some combination of the post office and the yellow pages.
At my university, the CS dept. switched to ssh/scp instead of telnet/ftp when we had a cracker on our networks sniffing out passwords. At a university, there is always going to be some machines which are poorly administrated. So you have to protect yourself from external and internal sources.
Bill Joy had a better we're-going-to-screw-it-up-royally article. The/. discussion is here.
A lot of comments essentially have a "the future is like the past" argument. Something like "We haven't killed ourselves yet with previous technologies, so we won't kill ourselves with this one." The problem with this argument is that it glosses over the differences in the new technology compared to old technologies.
Genetic engineering is going to be exactly what it says: constructing genetic makeup according to specifications. If it is possible to request a kid with specifications, why not anything else? Who is going to control things so that bad or even disasterous specifications are not allowed? Who can tell whether a spec is bad anyway? History is littered with unintended effects of introducing species to new environments.
The potential problem with genetic engineering is the lack of control. Once something new gets out of the lab, it is difficult to limit its reproduction, and thus limit its effect. Also, the effect outside of the lab is very difficult to predict. I don't think we can afford a wild, wild west phase of generic engineering, but appears to be what we are entering.
... the true owner of the Internet's open architecture has never been either the government or the telecom companies. The true owner was us -- the hackers. We composed the Request for Comments. We defined the standards. We wrote the software. We designed the routers. And we now run the ISPs (the suits may think they run them, but we know better). We have successfully fought off multiple attempts to lock up the network (including, for example, the voice-traffic oligopoly's concerted attempt to replace TCP/IP with ATM, and Microsoft's efforts to displace us with MSN and ActiveX.).
Since 1969 we have owned the Internet architecture, because none of the government or corporate power groups that wanted to control it or extract money from it could run it without us -- and despite our beards and Birkenstocks, we are very, very good at co-opting the system. For 30 years, we have been pushing toward a conscious, shared goal: to build the ultimate open-access network. We have successfully transmitted that dream across both corporate and national boundaries and across generations of engineers.
This policy debate is framed by technology we created and that only we understand -- because as people grok [deeply understand] the technology they become us; they absorb our values and goals and our cultural logic as part of the learning process, they become part of our freemasonry. Like the Chinese empire, we have defeated or assimilated every wave of barbarians to come at our borders.
We didn't always know how we would do it, and we almost never knew what the critical breakthroughs would be in advance or who would make them, but we knew where we were headed. So, for example, the World Wide Web was a surprise -- but Tim Berners-Lee is one of us, and his disposal of his invention was not.
While inspiring, I have a few quibbles. I think it would be more precise to say that "we run the internet", not that we own it. And what is meant by "we" anyway?
Certainly, computer-savvy people ("hackers") are required to keep the Internet going. To the extent that we hackers believe in an open Internet and implement that belief, the Internet will stay open. However, do we as a group really believe in an open Internet? And will we as a group remain steadfast in the face of government and corporate coercion? E.g., there appears to be some number of us who work for corporations (such as Microsoft) and governments (such as China) with different visions of the Internet. The benefits of working for Microsoft and the penalties for disobeying oppressive governments can be very persuasive.
Certainly, those of us who believe in an open Internet need to lobby politicians and persuade our managers that this is the right thing to do. The government can sometimes be very helpful. For example, I think that the USA's action against Microsoft has given a big boost to open source.
But our primary goal must be to expand our culture to all hackers. After all, we "own" the Internet.
How to pick a beginning programming language
on
Who's Afraid Of C++?
·
· Score: 1
In my opinion, Pascal remains the best intro language. The syntax is relatively simple, and you have some chance of learning the whole language. You also get structured programming and strong typing along the way. The big disadvantage (besides no OO) is that knowing Pascal is not very useful on your resume.
If you want object-oriented programming, I would think Java would maybe be the best choice. Unlike C++, there is at least a small chance of learning the whole language (though not all those libraries). And students can play with windows pretty much right away. Also, automatic garbage collection is a wonderful thing. Java is far from perfect, but OO+garbage collection is a great combination for a high-level language. I would agree that Smalltalk and Scheme are nice languages, but Java beats them by being more useful.
As far as Perl and Python is concerned, they have their uses, but I think you want strong typing for beginning programmers. There is no need to encourage bad habits from the start.
The problem is that the current naming system is ambiguous. For example, icann.org could just as well refer to the International Conference on Artificial Neural Networks. Maybe there is also a group of poorly-spelling optimists called the ICANN club (as opposed to the ICANNOT club).
The solution is to adopt an unambiguous naming system. Maybe calling it a "location" system is a better name. Just like the post office requires an unambiguous address to process the mail, the user needs to provide enough information to single out the desired site. The post office will not deliver mail with just a name on it (even if it is famous).
There are of course a lot of details to be worked out, but my idea is that companies (or people) provide a classified list of names that refer to them, e.g., company name, physical location, products, trademarks, abbreviations, etc. The user fills out a form or uses a shorthand to describe the desired site. If not enough info is provided, the user can select from a list of matches. Once disambiguated, the browser can save the match.
The disadvantages are that no one will get a short name and there would be an enormous hue and cry to make such a change. The advantages are that no one will be left out and it would be a more rational system.
Apparently, we needed something this summer to keep us from getting bored, but is the old KDE licensing issue the best that/. can do?
More seriously, from the tone of Dr. Bechly's diatribe, the whole thing sounds like a setup. It is easy to guess in advance that the KDE developers would do their best to ignore his offer. When they did, the good Doctor lists all the KDE sins he can think of or make up. His offer was obviously not very gracious.
I'm not quite what the advantage is for MS if the appeals drag out. MS's future will be uncertain until all the appeals are finished. If MS still ends up in one piece, they will still be subject to constant lawsuits. I also recall that other countries (in Europe?) were waiting to see what will happen in the US before taking any action.
Meanwhile, MS has to behave at least a little bit, and that has been and will be the open door for Linux and the BSDs to get a foothold in the OS market, and for Mozilla, KDE, and Gnome to get a foothold in the applications market.
You have probably gotten a lot of information from a large number of copyrighted works. I would think that you need to respect the copyright of GPLed code the same way as any copyrighted work. You can look at GPLed code all you want, but if you don't copy any of the code, you haven't violated the copyright.
With all advances in computers, why is a secure computer so difficult to do? To me, it appears to be a fundamental design error in the way that operating systems are put together. But where is the error and how can it be fixed?
Yes, I know there is OpenBSD and other more-or-less secure OSes. But it is still very easy to create security holes, and it is a lot of work to keep a system secure. The millions of ordinary users soon to come on cable modem and ADSL won't appreciate doing this sort of work.
I think it would be useful to standardize at least a subset of Perl. The syntax is a non-issue, I think; after all, Perl does have a parser. For the semantics, most of it boils down to numbers and strings (and POSIX). Regular expressions are a mess. Maybe pragmas, too.
I don't think software has become so expensive, but that less comes with it anymore. It used to be that when you bought software, you would get a semi-useful manual that you could learn from, and you would get some support in case you had trouble.
Now whenever you buy software, you have to head to Amazon or Borders or... to buy a book that explains it (the online manual being next to useless) and you have to pay for support. Throw in the bugs, the bloat, the installation headaches, and the licenses on top of that, and it is not surprising that the loyalty of the customer is lost.
This oxymoron "recovery CD" and other "piracy protection" will just make it worse. The company essentially accuses me of being a pirate, and insists on making my computer harder to use unless I become a "pirate" by circumventing their protection. This will unwittingly encourage their customers to become criminals.
Re: The Bible is copyrighted, you idiot
on
At The Crossroads
·
· Score: 1
Any modern version of the Bible is copyrighted. Why do you think you can only (easily) find older version (e.g., King James) online?
I would like a tool that would learn to extract headlines (or other info) from a variety of sites and display them for me. In the training phase, I would select the headlines, and the tool would learn the patterns to extract them. After training, I would get all the headlines from all the pages plus links to the article. Each person could create individualized headline services from the sites of their choice.
Maybe there is a lot more GUI work than AI here, but it would be nice tool to have.
Perhaps the best thing for/. is to meekly give in after, of course, they have thoroughly studied the situation with their lawyers for a few days/weeks/months. Who looks at these comments anyway after a few days? Meanwhile, the best thing for/. readers is to continue to post copies and links to copies. Then all Microsoft can do to/. is to sue concerning the speed that/. gets rid of the allegedly illegal posts. Meanwhile,/. can publish all the correspondence along the way and continue to give a black eye to Microsoft.
When is anybody, Microsoft in particular, going to learn anything from these types of attacks? Adding macros to documents or scripts to email can be useful, but it is a security nightmare unless the allowable operations are severely restricted, e.g., Java applets.
Also, a computer system should have some semblance of security against stupid actions by ordinary users. After all, we all do stupid things at one time or another. At the very minimum, the OS should differentiate between superuser and ordinary mode. Even if it doesn't require a password (I am thinking single-user Macs or Windows), at least the user would get a warning before something happened.
Finally, why are we stupid enough to put up with this stuff? We should demand better.
Living with more truth than we want
on
The Eroded Self
·
· Score: 1
I think we are going through a phase transition where privacy will disappear and we will manage to "grow up" about what kinds of people we really are and about the distortions of facts out of context. In the past, it was less likely that we would have to face up to our "private" sins and strange behaviors. Now this information will be more likely to be known by anybody. We will learn that ordinary people do all sorts of strange and disgusting things. We probably didn't want to know the truth, but "information wants to be free".
We will also have to learn not to take facts out of context. We are starting to see this now for famous people. How important are Clinton's affairs to his presidency? The people who hate Clinton see this as more evidence of his incompetency. Another group sees these as irrelevant compared to his politics.
How important is George Bush's party animal past to his run for President? Again, many people will see as evidence of incompetency, and others will see these as irrelevant compared to what policies he supports.
Everyone has a skeleton in their closets. In the future, all the closets will be open for everyone to see. We will live with it, and it will become a normal part of life.
This article discusses a joint venture of Microsoft and Xerox into "copyright protection software". Does anyone have any idea what kind of technology is involved here? If I can see it on the screen or hear it on the speakers, can you really make it impossible to copy? Inconvenient, yes, but I don't see how you can do better (or worse, I should say).
To oversimplify, there are (at least) four reasons why open source has become popular.
1. The hackers. You can't say enough about the efforts of the BSDers, RMS, Linus, etc., to make it happen and to make it work well. Even ESR gets some credit here, too.
2. The internet. The US government started it, let the geeks design it, funded it for a while, and opened it up to the world. The ARPANET is one the best things that the US government did in the past century. This is not to say that the US is not trying to screw things up in other ways.
3. Wintel (and Moore's law). For all their faults, Intel and Microsoft have popularized the cheap, powerful machines that make it possible to run Unix.
4. The government prosecution of Microsoft. This has forced Microsoft to behave just as open source appeared on MS's radar screen. Without government pressure, MS would have tried to crush open source using all possible means. Maybe it would have been more fun that way with open source being the perennial underdog and us screaming bloody murder all the time.
So by this count, the actions of the big, bad US government is crucial on two points, and big, bad business led to the machines that make it possible to run the OSes and software we know and love. Maybe it wasn't their intent, but it is hard to see this as the work of the devil(s).
How to maintain open source? Keep writing code and make your voice heard. Keep writing code that is open both in license and in making the power of the computer and computer networks available to everyone.
The/. guys must be really desperate for people to interview if need to resort to somebody as low as Jordan Pollack.
Hey, I'm just kidding you, Jordan, and I get to insult you while I hide behind my/. nick. I should get a hold of all of your old students. They would know what questions would deflate you, you old pompous windbag.
Anyway, I guess I am supposed to have a question instead of having fun. I know you are a big fan of doing AI by modeling the brain (e.g., neural networks and the like) rather than modeling the world (e.g., logic). What seems to be lost in all this is how can computers have fun like we do? [Yes, this the old, old problem of qualia.] A lot of what we humans do is not because we want to process information, but because we want to have fun. Having fun seems to be why a lot of people are writing free software and presumably why a lot of academics do the research they do (however, no one has any idea why you do the research you do, least of all yourself). And I don't want an answer in terms of your dynamical nonsense.
Gee, this really is fun insulting you. Maybe I'll have to think of another question.
Just an old Ohio State buddy giving you a hard time.
Endless Appeals are Good for Free Software
on
Microsoft Loses
·
· Score: 4
As long as this case is open, Microsoft has to behave, at least behave better than they would otherwise. Free software has been helped a lot by the fact that Microsoft couldn't squash it using all possible means because that would have hurt their case in the courts. The appeals mean that Microsoft has to be careful about their bullying. Whatever the final result is, Microsoft will not be the same. The IBM antitrust case is a good example here; IBM eventually won the court battle, but they lost a lot in the market.
The lack of a settlement means that this case will stay in the courts for years. Microsoft will have to behave in the meantime if it wants any chance at winning. Why have a settlement when not having a settlement will even better ensure Microsoft's behavior.
... is nearly impossible. In the past, even under oppressive governments, many underground organizations have been able to reach the faithful. The internet allows information to move faster and much more freely, and it makes it easier to broadcast information. If dictatorships were unable to control information flow before the internet, how can democracies expect to control information flow with the internet?
Of course, governments and corporations can adopt oppressive policies and make examples of a few people. This will force the information flow to go underground, but won't stop it very much where there is a demand.
The illicit drug trade is a good example. Drugs should be easier to stop because they are a physical quantity. But the ease of transport and consumer demand continues to overcome law enforcement.
Slashdot Mirror
This is true of KDE as much as QT. In hindsight, the KDE developers would have probably chosen a different license, but now that it's popular, the QT and KDE license don't work so well together, so goes the argument.
The real problem is that one apparently needs a savvy lawyer right from the start, but none of these projects have the money to make this happen. Why should I need a lawyer to do free software? This is the real problem with the GPL and its intricacies. It needs to be either clearer or more flexible.
But the linking clause is really hard to figure out. Any code depends on abstractions (e.g., specifed by a header file, admittedly partially), and the code doesn't really care whether the implementation of that abstraction (e.g., a library) is GPLed or not. Actually, the header file might be GPLed, too, but the code doesn't care about that either. foo.h might or might not be GPLed.
And, of course, a script links programs together in a different way.
My view is that the LGPL is a viable way of looking at the software world. The GPL wants to limit usage only by other GPL code, but again code doesn't really care about that. I guess RMS sure cares and that is why not much change can be expected.
Currently, ICANN is thinking about adding new genetic TLDs (in addition to .com, maybe add .web, .xxx, etc.). ICANN is also trying to figure out how to elect at-large members. Although the article doesn't state it, the most onerous of the proposed rules is a 10% requirement for putting a person on the ballot. Almost every public comment states that 10% is too high. The current board meets in mid-July. We'll see what happens.
So join and participate in how the internet is going to look like in the future. My feeling is that addresses like word.com will eventually fall by the wayside, and we will need to access web sites by more conventional information (name of company, country, product, etc.). This would like some combination of the post office and the yellow pages.
At my university, the CS dept. switched to ssh/scp instead of telnet/ftp when we had a cracker on our networks sniffing out passwords. At a university, there is always going to be some machines which are poorly administrated. So you have to protect yourself from external and internal sources.
A lot of comments essentially have a "the future is like the past" argument. Something like "We haven't killed ourselves yet with previous technologies, so we won't kill ourselves with this one." The problem with this argument is that it glosses over the differences in the new technology compared to old technologies.
Genetic engineering is going to be exactly what it says: constructing genetic makeup according to specifications. If it is possible to request a kid with specifications, why not anything else? Who is going to control things so that bad or even disasterous specifications are not allowed? Who can tell whether a spec is bad anyway? History is littered with unintended effects of introducing species to new environments.
The potential problem with genetic engineering is the lack of control. Once something new gets out of the lab, it is difficult to limit its reproduction, and thus limit its effect. Also, the effect outside of the lab is very difficult to predict. I don't think we can afford a wild, wild west phase of generic engineering, but appears to be what we are entering.
Certainly, computer-savvy people ("hackers") are required to keep the Internet going. To the extent that we hackers believe in an open Internet and implement that belief, the Internet will stay open. However, do we as a group really believe in an open Internet? And will we as a group remain steadfast in the face of government and corporate coercion? E.g., there appears to be some number of us who work for corporations (such as Microsoft) and governments (such as China) with different visions of the Internet. The benefits of working for Microsoft and the penalties for disobeying oppressive governments can be very persuasive.
Certainly, those of us who believe in an open Internet need to lobby politicians and persuade our managers that this is the right thing to do. The government can sometimes be very helpful. For example, I think that the USA's action against Microsoft has given a big boost to open source.
But our primary goal must be to expand our culture to all hackers. After all, we "own" the Internet.
If you want object-oriented programming, I would think Java would maybe be the best choice. Unlike C++, there is at least a small chance of learning the whole language (though not all those libraries). And students can play with windows pretty much right away. Also, automatic garbage collection is a wonderful thing. Java is far from perfect, but OO+garbage collection is a great combination for a high-level language. I would agree that Smalltalk and Scheme are nice languages, but Java beats them by being more useful.
As far as Perl and Python is concerned, they have their uses, but I think you want strong typing for beginning programmers. There is no need to encourage bad habits from the start.
The solution is to adopt an unambiguous naming system. Maybe calling it a "location" system is a better name. Just like the post office requires an unambiguous address to process the mail, the user needs to provide enough information to single out the desired site. The post office will not deliver mail with just a name on it (even if it is famous).
There are of course a lot of details to be worked out, but my idea is that companies (or people) provide a classified list of names that refer to them, e.g., company name, physical location, products, trademarks, abbreviations, etc. The user fills out a form or uses a shorthand to describe the desired site. If not enough info is provided, the user can select from a list of matches. Once disambiguated, the browser can save the match.
The disadvantages are that no one will get a short name and there would be an enormous hue and cry to make such a change. The advantages are that no one will be left out and it would be a more rational system.
More seriously, from the tone of Dr. Bechly's diatribe, the whole thing sounds like a setup. It is easy to guess in advance that the KDE developers would do their best to ignore his offer. When they did, the good Doctor lists all the KDE sins he can think of or make up. His offer was obviously not very gracious.
I nominate Dr. Bechly for chump-of-the-month.
Meanwhile, MS has to behave at least a little bit, and that has been and will be the open door for Linux and the BSDs to get a foothold in the OS market, and for Mozilla, KDE, and Gnome to get a foothold in the applications market.
You have probably gotten a lot of information from a large number of copyrighted works. I would think that you need to respect the copyright of GPLed code the same way as any copyrighted work. You can look at GPLed code all you want, but if you don't copy any of the code, you haven't violated the copyright.
Yes, I know there is OpenBSD and other more-or-less secure OSes. But it is still very easy to create security holes, and it is a lot of work to keep a system secure. The millions of ordinary users soon to come on cable modem and ADSL won't appreciate doing this sort of work.
So what is really the problem?
I think it would be useful to standardize at least a subset of Perl. The syntax is a non-issue, I think; after all, Perl does have a parser. For the semantics, most of it boils down to numbers and strings (and POSIX). Regular expressions are a mess. Maybe pragmas, too.
Now whenever you buy software, you have to head to Amazon or Borders or ... to buy a book that explains it (the online manual being next to useless) and you have to pay for support. Throw in the bugs, the bloat, the installation headaches, and the licenses on top of that, and it is not surprising that the loyalty of the customer is lost.
This oxymoron "recovery CD" and other "piracy protection" will just make it worse. The company essentially accuses me of being a pirate, and insists on making my computer harder to use unless I become a "pirate" by circumventing their protection. This will unwittingly encourage their customers to become criminals.
Any modern version of the Bible is copyrighted. Why do you think you can only (easily) find older version (e.g., King James) online?
Maybe there is a lot more GUI work than AI here, but it would be nice tool to have.
Perhaps the best thing for /. is to meekly give in after, of course, they have thoroughly studied the situation with their lawyers for a few days/weeks/months. Who looks at these comments anyway after a few days? Meanwhile, the best thing for /. readers is to continue to post copies and links to copies. Then all Microsoft can do to /. is to sue concerning the speed that /. gets rid of the allegedly illegal posts. Meanwhile, /. can publish all the correspondence along the way and continue to give a black eye to Microsoft.
Also, a computer system should have some semblance of security against stupid actions by ordinary users. After all, we all do stupid things at one time or another. At the very minimum, the OS should differentiate between superuser and ordinary mode. Even if it doesn't require a password (I am thinking single-user Macs or Windows), at least the user would get a warning before something happened.
Finally, why are we stupid enough to put up with this stuff? We should demand better.
We will also have to learn not to take facts out of context. We are starting to see this now for famous people. How important are Clinton's affairs to his presidency? The people who hate Clinton see this as more evidence of his incompetency. Another group sees these as irrelevant compared to his politics.
How important is George Bush's party animal past to his run for President? Again, many people will see as evidence of incompetency, and others will see these as irrelevant compared to what policies he supports.
Everyone has a skeleton in their closets. In the future, all the closets will be open for everyone to see. We will live with it, and it will become a normal part of life.
This article discusses a joint venture of Microsoft and Xerox into "copyright protection software". Does anyone have any idea what kind of technology is involved here? If I can see it on the screen or hear it on the speakers, can you really make it impossible to copy? Inconvenient, yes, but I don't see how you can do better (or worse, I should say).
1. The hackers. You can't say enough about the efforts of the BSDers, RMS, Linus, etc., to make it happen and to make it work well. Even ESR gets some credit here, too.
2. The internet. The US government started it, let the geeks design it, funded it for a while, and opened it up to the world. The ARPANET is one the best things that the US government did in the past century. This is not to say that the US is not trying to screw things up in other ways.
3. Wintel (and Moore's law). For all their faults, Intel and Microsoft have popularized the cheap, powerful machines that make it possible to run Unix.
4. The government prosecution of Microsoft. This has forced Microsoft to behave just as open source appeared on MS's radar screen. Without government pressure, MS would have tried to crush open source using all possible means. Maybe it would have been more fun that way with open source being the perennial underdog and us screaming bloody murder all the time.
So by this count, the actions of the big, bad US government is crucial on two points, and big, bad business led to the machines that make it possible to run the OSes and software we know and love. Maybe it wasn't their intent, but it is hard to see this as the work of the devil(s).
How to maintain open source? Keep writing code and make your voice heard. Keep writing code that is open both in license and in making the power of the computer and computer networks available to everyone.
Hey, I'm just kidding you, Jordan, and I get to insult you while I hide behind my /. nick. I should get a hold of all of your old students. They would know what questions would deflate you, you old pompous windbag.
Anyway, I guess I am supposed to have a question instead of having fun. I know you are a big fan of doing AI by modeling the brain (e.g., neural networks and the like) rather than modeling the world (e.g., logic). What seems to be lost in all this is how can computers have fun like we do? [Yes, this the old, old problem of qualia.] A lot of what we humans do is not because we want to process information, but because we want to have fun. Having fun seems to be why a lot of people are writing free software and presumably why a lot of academics do the research they do (however, no one has any idea why you do the research you do, least of all yourself). And I don't want an answer in terms of your dynamical nonsense.
Gee, this really is fun insulting you. Maybe I'll have to think of another question.
Just an old Ohio State buddy giving you a hard time.
As long as this case is open, Microsoft has to behave, at least behave better than they would otherwise. Free software has been helped a lot by the fact that Microsoft couldn't squash it using all possible means because that would have hurt their case in the courts. The appeals mean that Microsoft has to be careful about their bullying. Whatever the final result is, Microsoft will not be the same. The IBM antitrust case is a good example here; IBM eventually won the court battle, but they lost a lot in the market.
The lack of a settlement means that this case will stay in the courts for years. Microsoft will have to behave in the meantime if it wants any chance at winning. Why have a settlement when not having a settlement will even better ensure Microsoft's behavior.
Of course, governments and corporations can adopt oppressive policies and make examples of a few people. This will force the information flow to go underground, but won't stop it very much where there is a demand.
The illicit drug trade is a good example. Drugs should be easier to stop because they are a physical quantity. But the ease of transport and consumer demand continues to overcome law enforcement.