No mention is made of increased fan base, or of the likelyhood that increased exposure = more likely sales. [...] Result? I'm actually considering buying a couple of CD's.
Somehow I doubt the RIAA is impressed by "considering buying". Granted, some people might decide to buy a CD after hearing an MP3 - but they would also buy it after hearing the CD.
Copyright law is a good law. You might not like it in all cases; but it's the same law on which the GPL is based. If you don't like what the record industry is doing - vote with your money. Spend it on something else.
We must not accept anything that is not under a free license, preferably GPL. If we accept BeOS, or even Sun, we could very well end up with different masters, though we remain slaves.
Can I see the GPL for the hardware you are using? Or, are you by any chance, using propriatery hardware? "Free" is nice, but it isn't the end-all be-all.
Don't forget that Sun think that they're in a difficult position because of the Linux vs Solaris issue, so inevitably there will be internal pressure from some of their divisions *not* to help out the Linux scene too much. In contrast, adding value to little ol' Be holds no danger for them.
I find it extremely unlikely that Linux had anything to do with Suns decision to port Java to Be. The world doesn't revolve around Linux, you know. Sun promotes Java. It's promoted as "write once, run everywhere". That's more than enough reason to port Java to any serious platform; including BeOS.
In reality of course, assuming that they really make no money from Solaris as the rumours suggest, their best bet is probably to GPL the entirety of Solaris and to support all free operating systems equally on their nice hardware.
That's absurd, for various reasons. First of all, their OS wasn't 100% written from the ground up, but contains enough BSD and AT&T copyrighted code to make GPL impossible. Secondly, Sun is a commercial organisation, they provide commercial support for their OS. Making sure their OS works well with their hardware is their business - supporting more OSses would only make it more difficult from them; and for which gain?
Unfortunately, internal politics may not allow that
If you replace "internal politics" with "commercial viability", I would agree with you.
They may understand the medium, but they are coding for most of their target market.
Uhm, yeah, sure. Lots of companies have "MSIE users" as their target market. Just like all those companies who have "people driving in a black sedan" as their target market.
I wonder if the guy in the cage feels OK after the fact. It's like using thousands of mobile phones at the same time, in terms of electromagnetic radiation hammering your brain.
No, it's not. The cage is made of metal, and hence it is a Faraday cage. There is *NO* electromagnetic force inside the cage. Very simplified explaination: Metal is a conductor; electricity is carried by electrons; the sign of the charge of all electrons is the same, hence, all electrons repel each other. Because they repel, they want to be as far as possible from each other. Being as far away from each other as possible means that electricity flows on the outside of conductors.
I disagree - the Web was designed to be a hyperlinked multimedia delivery system. That doesn't mean that it is a multi-service delivery system. Essentially, it's a resource request mechanism: you ask for a resource, the server gives it up. What the requesting client does with it is completely up to implementation. The idea of standards provides a consistant representation of what the client should do with that resource.
That's an description of *HTTP*. HTTP isn't the same as the web, it's *part* of the web. One bases of the web is the underlaying URI/URN/URL addressing scheme. HTTP is just one of the schemes, FTP, GOPHER, NEWS, etc, are others.
SMTP/POP/IMAP/NNTP were designed to provide completely different types of services. These types of transactions are not request/response based, but instead are dialog based (ie, LIST/UIDL/GET[[UIDL/GET]...] for POP3). The underlying mechanisms are completely different paradigms, therefore they should use different tools.
Why? What does it matter? Why should a POP server care if it was called by a piece of program that has its own entry in the process table, or by a piece of program that shares its entry in the process table with a piece of program that shows bouncing gifs? You'd think the POP server behaves differently?
What matters is that the user has the choice, and if (s)he prefers to seaminglessly jump from an HTML page delivered by HTTP to a Usenet posting delivered by NNTP, more power to him/her. I prefer having a car that drives me everywhere - I wouldn't want to switch cars when entering a freeway, or for driving a gravel road.
I'm basically begging the rest of the world to wake up to what the unix world realized a long time ago: 'do one thing and do it well - then connect the tools'
That's an implementation detail. If it works fine for you, good for you. I use mutt and slrn for my email and news, but I'm not going to start an FTP client if I find a link in an HTML page that happens to use the ftp:// scheme, or use a separate gopher client.
I assume you don't inline images either, but use specialized tools for that? (One for GIFs, one of JPGs, one for PNGs....)
First, they fucked up the standards. Since Netscape did not follow the standards, IETF released HTML 2.0, which was an attempt for a standard to follow Netscape. But they still managed to screw it up, thus making standard violation acceptable, and "enhancements" hard to detect.
Well, it's certainly true that Netscape ignored the standards to gain marketshare, but that's not the reason for HTML 2.0. HTML 2.0 describes "state of the art" of early 1994. That's 6 months before the first Netscape alphas became available.
However, Netscape's ass-whiping behaviour of ignoring the proposed HTML 3.0 table syntax lead to a large rewrite, eventually leading to RFC 1942. We're now 4.5 years past the introduction of tables in Netscape, and there hasn't been much improvement yet - yeah, you can now have background colours in your tables - big deal. They are nowhere implementing RFC 1946. Which was formalized in May 1996, but essentially finished (and available) in fall 1995.
Up to this very day, Netscape doesn't fully implement HTML 2.0. Heck, they can't even parse a comment.
Netscape has access to the Linux kernel and the source for X...why can NS make a Linux specific version of Navigator?
Oh, it's not a matter of can. M$ could make a Linux specific version of MSIE as well. And Sun a Linux specific version of Hot Java. But they are all commercial companies. Netscape compiles on a whole bunch of Unix platforms. Why on earth would they invest money in making a platform specific version, if there's no reason to think there's any gain?
this will only aid the acceptance of Mozilla as the standard WWW browser.
That's funny. What's next? Microsoft releasing its source code for Windows, so people can build the standard OS?
Netscape has always been utter crap in all aspects. Never been able to implement a standard, always bloated, severely crippled news and mail readers. Only people with time to waste would build something on top of that. And only fools would believe something useful coming out of it.
That's why for a number of years web designers went to great pains to ensure that webpages were well-tested on both browsers
"What kind of music do you usually have here?" "Oh, we got both kinds. We got country *and* western."
People "designing" web pages or web sites for a specific set of browsers are utter lusers who do not understand the media they are working with. 2 digit IQs - in octal.
Either you make something for the World Wide Web, or you don't. There's no middle ground. Regardless of what Netscape and MSIE are making you believe.
They[W3C]want everyone to follow the standard, yet they purvey reference implementations that can be molded into whatever proprietary shape that the Microsofts or Netscapes of this world care to dream up.
Well, duh. W3C isn't a standards body. It's an industrial consortium, with Netscape and Microsoft (and others) as members.
A long-standing problem with HTML is its non-turing-completeness.
Yeah, and it doesn't breathe underwater either. The ceiling in my living room isn't Turing complete either, but that isn't a problem. And neither is the fact that HTML isn't. HTML isn't a programming language, so, it wouldn't have a property that applies to programming languages.
Is XML turing-complete?
No. XML is just SGML-lite. SGML and XML are systems to describe grammars of markup languages. HTML is an SGML application. HTML *describes* the structure of a document, like all SGML and XML applications. It doesn't calculate, so talking about Turing completeness is nonsense.
There is a good reason not to allow everyone "extend" their documents with new elements. Suppose I would "extend" my document with a FLUIP element. How are you, as a browser author, going to deal with that? Mind, you have to take care of the fact that the users of your product might have special viewing requirements.
Having said that, you can extend HTML. Just write your own DTD, and use the appropriate doctype. Of course, you will have a hard time finding a browser that will actually fetch and parse DTDs, but the principle is there.
But there's this nagging problem. Now you have your new elements BRIFFLE, KRUIT, STAVOR and HINKE. You have a DTD that describes the syntax. But what do those elements mean? How should they be displayed? Stylesheets, you might argue. But can you make stylesheets for all possible platforms? Do you even know all possible platforms? What about people with special needs, like the use of large fonts, or a high contrast, who will have their own stylesheets overruling any document attached ones. How are they going to deal with those unknown elements?
Extending HTML without standards is a really bad idea.
The OS needs to be usable to a new user - on the same level as Windows.
No, it doesn't. There is absolutely no reason to gain a large marketshare. Let's focus on usability for the computer minded person. Do you think Ferrari, or Rolls-Royce have the urge to change their product such that it would appeal to the masses? No - they just make quality products.
Linux is a Unix. Unices come with many benefits; but the flip side is that's it's harder to use. It's not suitable for Sally Q. User. Big fucking deal. Why should that matter?
Single file which can contain multiple objects. These objects could be anything including HTML, XML, images, Java, etc. I suggest the file format be identical to the Java JAR format. Let's call it a WAR file for Web ARchive.
That would be real bad. It doesn't gain anything, and you lose a lot. Browser can already download everything in a single connection using the 'Keep-Alive' option of HTTP/1.1, dating from 1995 or 1996.
And you would lose the ability of doing content-negotiation; be it by using HTTP's Accept headers, nested OBJECTs in HTML, or even simple HEAD requests. On top of losing flexibility, you increase network traffic. If the text of a page has changed, but the 400k image it inlines hasn't, you would have to download the entire image again, just to see the updated text. Indexers like Lycos and Altavista now suddenly have to download a multitude of what they are downloading currently; meaning that it would take even longer before your pages get indexed.
The cache is a red herring. There are no advantages over keeping a huge single file in a cache than a bunch of smaller ones, which combined make the huge one. There are disadvantages though.
There might be some benefits on the server side, but frankly, I fail to see them.
I am getting sick and tired of blaming MSIE of the downfall of the web.
It's Netscape's fault. The web was ruined by Netscape's way of ignoring standards long before M$ realized what the web was. If it wasn't for M$ and MSIE, the "fast forward a year" would be already have happened. Except it would have been Netscape monopolizing the market, and not M$.
From the begining, all Netscape was interested in was making money, by dominating the market. Remember the early years, where you actually had to pay for your Netscape browser? (Ok, hands up, who was not a software pirate and downloaded Netscape and paid for it? - and don't come with "non-commercial use wasn't free", because it wasn't.)
As for third-party, platform dependent plugins, you cannot blame Microsoft for that. Plugins were a Netscape invention; and third-party plugins are, well, third-party plugins. Of course, while the cubicles of Netscape housed some programmers that were interested in porting Netscape to as many platforms as possible; the company itself mostly looked at the Windows market, as that's the only platform of interest when you want to dominate.
The browser war is over, and there isn't a clear winner. Which is a good thing.
And there are some non-Unix OSses that do that as well. In a previous job, we had a Netapp machine. If it crashed, you could dump out the kernel - which was cool because then you could ftp 80 Mb (or something like that) of core dumps to Netapps, and they would have a look. Which sometimes resulted in kernel patches a couple of weeks later.
And you're playing into it perfectly. They've positioned themselves so Joe Public will say "Well everyone's using Windows so if i use Linux my freinds can't help me."
Ah, yes, and the Unix world is so much better.
wavy lines Suppose I have a problem with my Linux machine...
Friend 1: I am sorry, I run OpenBSD, and that works fine.
Friend 2: Well, on my Solaris x86, I would do it like this...
Friend 3: Sure, you can borrow my HP-UX manuals.
Friend 4: Ah, in IRIX, we do....
Friend 5: Cool, do you also like AIX?
Friend 6: If you don't do SCO, I cannot help you.
Friend 7: I've been waiting for the Hurd for 12 years now.
Friend 8: Yeah, I run Linux. But all I do is install Red Hat RPMs. Oh, and you run Debian? Sorry...
Now, I'm out of friends, and still have this problem... wavy lines
You defend Microsoft by saying that everyone should use Windows because Linux is hard. Don't you realize that if it weren't for the monopoly then Linux might have the commercial support to make it easier?
Huh? Linux isn't "hard" because there isn't commercial support. It's "hard" because it's a Unix. There has been commercial Unices even before Bill Gates first uttered "windows".
Unices are perfectly fine OSses, suitable for many tasks, and many people. It's not suitable as a desktop machine for Sally Q. User.
Ferrari's and 24 wheeler trucks are very good cars. They are however, unsuitable for most people who just need to pick up their kids from school.
Oil tankers and miles of copper wiring aren't amenable to easy division and distrobution, but this is INFORMATION, guys.
Code and information alone does not make a big company. Being a huge multinational means having a lot of infrastructure as well.
Care for a little "what if"? Suppose M$ does get broken into pieces. Or even have all its code and other IP be put in the public domain. Just imagine for now, there would be no M$ anymore. There still will be a couple of hundred million people out there needing software.
Ok, so we get a whole queue of new companies writing software. OSses, applications, etc, etc. Of course, each new little company wants to protect its stuff, so we get lots and lots of new, propriatery OSses and applications. This time, fully backed up by US law.
Who gains by this? The consumers? No, not at all. Currently, if you are on a windows system, you have a reasonable chance your applications work on your neighbours computer as well - as he's likely to have windows as well. Same for documents; word documents might be a propriatery format, and not very compatible with older versions - half the world does run the current version of word and is able to read your document. Perhaps the programmers? Nah. Whether they write for platform X, or platform Y, they'll get paid the same on average. All the incompatabilities make life just more frustrating. There will be a bunch of suits and investors that'll profit from many new companies though. And lawyers.
Oh, you might say, that will never happen. Well, I think it will - it has happened in the computer world so often. Look at the Unix world; a twisty little maze of OSses, all different. I cannot take my Solaris program and run it on HP. Even with the source code, it needs to be "ported". Attempts of creating languages so that you can "write once, run everywhere" don't quite work. Java, for instance, has various, incompatible virtual machines - and not just because of M$. And for a language like Perl, that has one implementation on Windows, Unix and VMS, there are too many dependencies on the underlaying OS to truly "write once, run everywhere".
I find it very cynical that the DoJ case stems from a browser war. Remember Tim's original idea? Make information accessable to all - by having open standards: HTML and HTTP. That went all fine, untill we got a player in the market that wanted to just like Microsoft: Netscape communications. They ignored the standards as much as they could, invented their own "extensions" instead of implementing proposed standards. All with one goal, gain money and marketshare. Microsoft goals with Microsoft techniques.
I am very grateful Microsoft decided to produce its own browser, and compete with Netscape. Hadn't it been for Microsoft, Netscape would have monopolized the HTML/HTTP market, and it would now be all propriatery.
Don't get me wrong. I'm not at all a Microsoft fan. I might have used a computer running an OS from Redmond perhaps 10 times in my life. Probably less. But the computer world outside Microsoft is a big, incompatible mess; even if you only look at Unix. In fact, I doubt M$ would have ever grown so big if it wasn't for the Unix mess. Users aren't helped with incompatible systems. For them, life without Microsoft would be scary.
Slashdot Mirror
Somehow I doubt the RIAA is impressed by "considering buying". Granted, some people might decide to buy a CD after hearing an MP3 - but they would also buy it after hearing the CD.
Copyright law is a good law. You might not like it in all cases; but it's the same law on which the GPL is based. If you don't like what the record industry is doing - vote with your money. Spend it on something else.
-- Abigail
Uhm, lightyears? It's nice to chat dogma's, but have you any reason to claim so?
-- Abigail
Can I see the GPL for the hardware you are using? Or, are you by any chance, using propriatery hardware? "Free" is nice, but it isn't the end-all be-all.
-- Abigail
I find it extremely unlikely that Linux had anything to do with Suns decision to port Java to Be. The world doesn't revolve around Linux, you know. Sun promotes Java. It's promoted as "write once, run everywhere". That's more than enough reason to port Java to any serious platform; including BeOS.
In reality of course, assuming that they really make no money from Solaris as the rumours suggest, their best bet is probably to GPL the entirety of Solaris and to support all free operating systems equally on their nice hardware.
That's absurd, for various reasons. First of all, their OS wasn't 100% written from the ground up, but contains enough BSD and AT&T copyrighted code to make GPL impossible. Secondly, Sun is a commercial organisation, they provide commercial support for their OS. Making sure their OS works well with their hardware is their business - supporting more OSses would only make it more difficult from them; and for which gain?
Unfortunately, internal politics may not allow that
If you replace "internal politics" with "commercial viability", I would agree with you.
-- Abigail
Uhm, yeah, sure. Lots of companies have "MSIE users" as their target market. Just like all those companies who have "people driving in a black sedan" as their target market.
-- Abigail
No, it's not. The cage is made of metal, and hence it is a Faraday cage. There is *NO* electromagnetic force inside the cage. Very simplified explaination: Metal is a conductor; electricity is carried by electrons; the sign of the charge of all electrons is the same, hence, all electrons repel each other. Because they repel, they want to be as far as possible from each other. Being as far away from each other as possible means that electricity flows on the outside of conductors.
-- Abigail
That's an description of *HTTP*. HTTP isn't the same as the web, it's *part* of the web. One bases of the web is the underlaying URI/URN/URL addressing scheme. HTTP is just one of the schemes, FTP, GOPHER, NEWS, etc, are others.
SMTP/POP/IMAP/NNTP were designed to provide completely different types of services. These types of transactions are not request/response based, but instead are dialog based (ie, LIST/UIDL/GET[[UIDL/GET]...] for POP3). The underlying mechanisms are completely different paradigms, therefore they should use different tools.
Why? What does it matter? Why should a POP server care if it was called by a piece of program that has its own entry in the process table, or by a piece of program that shares its entry in the process table with a piece of program that shows bouncing gifs? You'd think the POP server behaves differently?
What matters is that the user has the choice, and if (s)he prefers to seaminglessly jump from an HTML page delivered by HTTP to a Usenet posting delivered by NNTP, more power to him/her. I prefer having a car that drives me everywhere - I wouldn't want to switch cars when entering a freeway, or for driving a gravel road.
I'm basically begging the rest of the world to wake up to what the unix world realized a long time ago: 'do one thing and do it well - then connect the tools'
That's an implementation detail. If it works fine for you, good for you. I use mutt and slrn for my email and news, but I'm not going to start an FTP client if I find a link in an HTML page that happens to use the ftp:// scheme, or use a separate gopher client.
I assume you don't inline images either, but use specialized tools for that? (One for GIFs, one of JPGs, one for PNGs....)
-- Abigail
Well, it's certainly true that Netscape ignored the standards to gain marketshare, but that's not the reason for HTML 2.0. HTML 2.0 describes "state of the art" of early 1994. That's 6 months before the first Netscape alphas became available.
However, Netscape's ass-whiping behaviour of ignoring the proposed HTML 3.0 table syntax lead to a large rewrite, eventually leading to RFC 1942. We're now 4.5 years past the introduction of tables in Netscape, and there hasn't been much improvement yet - yeah, you can now have background colours in your tables - big deal. They are nowhere implementing RFC 1946. Which was formalized in May 1996, but essentially finished (and available) in fall 1995.
Up to this very day, Netscape doesn't fully implement HTML 2.0. Heck, they can't even parse a comment.
-- Abigail
Having to use all those different tools was one of the reasons "the Web" was invented. Multi-scheme browsers predate Netscape Communications.
-- Abigail
Oh, it's not a matter of can. M$ could make a Linux specific version of MSIE as well. And Sun a Linux specific version of Hot Java. But they are all commercial companies. Netscape compiles on a whole bunch of Unix platforms. Why on earth would they invest money in making a platform specific version, if there's no reason to think there's any gain?
-- Abigail
That's funny. What's next? Microsoft releasing its source code for Windows, so people can build the standard OS?
Netscape has always been utter crap in all aspects. Never been able to implement a standard, always bloated, severely crippled news and mail readers. Only people with time to waste would build something on top of that. And only fools would believe something useful coming out of it.
-- Abigail
Who the frig would care whether your site renders fast or not? Tables for layout means you've no fucking clue about the media you're working with.
-- Abigail
"What kind of music do you usually have here?"
"Oh, we got both kinds. We got country *and* western."
People "designing" web pages or web sites for a specific set of browsers are utter lusers who do not understand the media they are working with. 2 digit IQs - in octal.
Either you make something for the World Wide Web, or you don't. There's no middle ground. Regardless of what Netscape and MSIE are making you believe.
-- Abigail
Given that's based on the shit Netscape produced, can Mozilla parse HTML 2? Netscape was never able to do that.
-- Abigail
Well, duh. W3C isn't a standards body. It's an industrial consortium, with Netscape and Microsoft (and others) as members.
-- Abigail
Yeah, and it doesn't breathe underwater either. The ceiling in my living room isn't Turing complete either, but that isn't a problem. And neither is the fact that HTML isn't. HTML isn't a programming language, so, it wouldn't have a property that applies to programming languages.
Is XML turing-complete?
No. XML is just SGML-lite. SGML and XML are systems to describe grammars of markup languages. HTML is an SGML application. HTML *describes* the structure of a document, like all SGML and XML applications. It doesn't calculate, so talking about Turing completeness is nonsense.
There is a good reason not to allow everyone "extend" their documents with new elements. Suppose I would "extend" my document with a FLUIP element. How are you, as a browser author, going to deal with that? Mind, you have to take care of the fact that the users of your product might have special viewing requirements.
Having said that, you can extend HTML. Just write your own DTD, and use the appropriate doctype. Of course, you will have a hard time finding a browser that will actually fetch and parse DTDs, but the principle is there.
But there's this nagging problem. Now you have your new elements BRIFFLE, KRUIT, STAVOR and HINKE. You have a DTD that describes the syntax. But what do those elements mean? How should they be displayed? Stylesheets, you might argue. But can you make stylesheets for all possible platforms? Do you even know all possible platforms? What about people with special needs, like the use of large fonts, or a high contrast, who will have their own stylesheets overruling any document attached ones. How are they going to deal with those unknown elements?
Extending HTML without standards is a really bad idea.
-- Abigail
No, it doesn't. There is absolutely no reason to gain a large marketshare. Let's focus on usability for the computer minded person. Do you think Ferrari, or Rolls-Royce have the urge to change their product such that it would appeal to the masses? No - they just make quality products.
Linux is a Unix. Unices come with many benefits; but the flip side is that's it's harder to use. It's not suitable for Sally Q. User. Big fucking deal. Why should that matter?
-- Abigail
Very true. Ditto for Netscape.
-- Abigail
That would be real bad. It doesn't gain anything, and you lose a lot. Browser can already download everything in a single connection using the 'Keep-Alive' option of HTTP/1.1, dating from 1995 or 1996.
And you would lose the ability of doing content-negotiation; be it by using HTTP's Accept headers, nested OBJECTs in HTML, or even simple HEAD requests.
On top of losing flexibility, you increase network traffic. If the text of a page has changed, but the 400k image it inlines hasn't, you would have to download the entire image again, just to see the updated text. Indexers like Lycos and Altavista now suddenly have to download a multitude of what they are downloading currently; meaning that it would take even longer before your pages get indexed.
The cache is a red herring. There are no advantages over keeping a huge single file in a cache than a bunch of smaller ones, which combined make the huge one. There are disadvantages though.
There might be some benefits on the server side, but frankly, I fail to see them.
-- Abigail
Yes. I used to run MSIE under Solaris, and Solaris also runs on Intel platforms. However, I do not know whether MSIE also runs under Solaris x86.
-- Abigail
It's Netscape's fault. The web was ruined by Netscape's way of ignoring standards long before M$ realized what the web was. If it wasn't for M$ and MSIE, the "fast forward a year" would be already have happened. Except it would have been Netscape monopolizing the market, and not M$.
From the begining, all Netscape was interested in was making money, by dominating the market. Remember the early years, where you actually had to pay for your Netscape browser? (Ok, hands up, who was not a software pirate and downloaded Netscape and paid for it? - and don't come with "non-commercial use wasn't free", because it wasn't.)
As for third-party, platform dependent plugins, you cannot blame Microsoft for that. Plugins were a Netscape invention; and third-party plugins are, well, third-party plugins. Of course, while the cubicles of Netscape housed some programmers that were interested in porting Netscape to as many platforms as possible; the company itself mostly looked at the Windows market, as that's the only platform of interest when you want to dominate.
The browser war is over, and there isn't a clear winner. Which is a good thing.
-- Abigail
-- Abigail
Ah, yes, and the Unix world is so much better.
wavy lines
Suppose I have a problem with my Linux machine...
- Friend 1: I am sorry, I run OpenBSD, and that works fine.
- Friend 2: Well, on my Solaris x86, I would do it like this...
- Friend 3: Sure, you can borrow my HP-UX manuals.
- Friend 4: Ah, in IRIX, we do....
- Friend 5: Cool, do you also like AIX?
- Friend 6: If you don't do SCO, I cannot help you.
- Friend 7: I've been waiting for the Hurd for 12 years now.
- Friend 8: Yeah, I run Linux. But all I do is install Red Hat RPMs. Oh, and you run Debian? Sorry...
Now, I'm out of friends, and still have this problem... wavy linesUnix isn't for Joe Public.
-- Abigail
Huh? Linux isn't "hard" because there isn't commercial support. It's "hard" because it's a Unix. There has been commercial Unices even before Bill Gates first uttered "windows".
Unices are perfectly fine OSses, suitable for many tasks, and many people. It's not suitable as a desktop machine for Sally Q. User.
Ferrari's and 24 wheeler trucks are very good cars. They are however, unsuitable for most people who just need to pick up their kids from school.
-- Abigail
Code and information alone does not make a big company. Being a huge multinational means having a lot of infrastructure as well.
Care for a little "what if"? Suppose M$ does get broken into pieces. Or even have all its code and other IP be put in the public domain. Just imagine for now, there would be no M$ anymore. There still will be a couple of hundred million people out there needing software.
Ok, so we get a whole queue of new companies writing software. OSses, applications, etc, etc. Of course, each new little company wants to protect its stuff, so we get lots and lots of new, propriatery OSses and applications. This time, fully backed up by US law.
Who gains by this? The consumers? No, not at all. Currently, if you are on a windows system, you have a reasonable chance your applications work on your neighbours computer as well - as he's likely to have windows as well. Same for documents; word documents might be a propriatery format, and not very compatible with older versions - half the world does run the current version of word and is able to read your document.
Perhaps the programmers? Nah. Whether they write for platform X, or platform Y, they'll get paid the same on average. All the incompatabilities make life just more frustrating.
There will be a bunch of suits and investors that'll profit from many new companies though. And lawyers.
Oh, you might say, that will never happen. Well, I think it will - it has happened in the computer world so often. Look at the Unix world; a twisty little maze of OSses, all different. I cannot take my Solaris program and run it on HP. Even with the source code, it needs to be "ported". Attempts of creating languages so that you can "write once, run everywhere" don't quite work. Java, for instance, has various, incompatible virtual machines - and not just because of M$. And for a language like Perl, that has one implementation on Windows, Unix and VMS, there are too many dependencies on the underlaying OS to truly "write once, run everywhere".
I find it very cynical that the DoJ case stems from a browser war. Remember Tim's original idea? Make information accessable to all - by having open standards: HTML and HTTP. That went all fine, untill we got a player in the market that wanted to just like Microsoft: Netscape communications. They ignored the standards as much as they could, invented their own "extensions" instead of implementing proposed standards. All with one goal, gain money and marketshare. Microsoft goals with Microsoft techniques.
I am very grateful Microsoft decided to produce its own browser, and compete with Netscape. Hadn't it been for Microsoft, Netscape would have monopolized the HTML/HTTP market, and it would now be all propriatery.
Don't get me wrong. I'm not at all a Microsoft fan. I might have used a computer running an OS from Redmond perhaps 10 times in my life. Probably less. But the computer world outside Microsoft is a big, incompatible mess; even if you only look at Unix. In fact, I doubt M$ would have ever grown so big if it wasn't for the Unix mess. Users aren't helped with incompatible systems. For them, life without Microsoft would be scary.
-- Abigail