Let's say I'm running AMD, and in the impartial benchmarks, our chips beat equivalent-speed Intel chips by 31%. I'd be a fool _not_ to use those benchmarks in my ads--and crow about how I win using impartial, open benchmarks.
In the same way, it Adobe starts to worry about GIMP and releases a linux version of Photoshop, and it blows away GIMP on the same machine using impartial, open benchmarks, they'll use those benchmarks in their ads.
It's like when the PowerPC G4 did so ridiculously well on Byte's benchmarks, Apple's ads make very clear that the benchmarks came from Byte, and were impartial (hell, if anything, Byte was biased _against_ Apple).
In the end, I'm not sure open benchmarks will improve things all that much. In the examples above, Intel can just make sure that the new PIII Argon improves over the PIII Xeon in the specific areas that will matter.
And if there are too many benchmark suites, it just gets confusing. AMD can advertise that they win on benchmarks A, B, and C, and Intel that they win on D, E, and F--and then who do you believe?
I'll assume this isn't a troll, even though it probably is, and try to answer very simply:
1. Your credit card number may be "information," but you don't want everyone to have it.
2. If you send an email to CNN descibing misdeeds of the repressive regime in your country, you probably don't want the local regime to be able to intercept and modify it.
Just because international waters are free doesn't mean pirates can intercept any ship more than 10 miles from shore with impunity.
Start Menu: Apple Menu Spell checking as you type: at least two third-party apps for Word for Macintosh OLE/DDE/D&D: OpenDoc/Publish&Subscribe/D&D DirectX: Um... What's new here, exactly? The Mac had a way to do full-screen drawing, DOS allowed you to write directly to the hardware, OpenGL allowed hardware 3D support, etc. DCOM: Corba ActiveDirectory: NDS and LDAP
Some of these are more borderline. The basic concepts behind the Intellimouse and the Natural Keyboard already existed, but only in specialized applications. So they deserve some credit there. Also, OLE's linking was much better than P&S's.
There are some much better examples of Microsoft innovation. First, two killer apps for the Macintosh: Word, which had a GUI and WYSIWYG layout at least as good as MacWrite's but was powerful enough to compete with the DOS standards WordStar and WordPerfect, and Excel, a spreadsheet nearly as good as Lotus with a much simpler learning curve (plus better printing support and other improvements).
And it wasn't just in the early days. Win95 and Office95's pervasive, consistent context menus finally made that second mouse button useful for consumer users. ODBC allowed middleware to be ported from Microsoft SQL Server to Oracle in a matter of days. The auto-install/self-repair features in Office98 and IE4.5/OE5.0 for MacOS is the first good implementation of that idea.
In other words, Microsoft has made some innovations. Not as many as some of their competitors, not as many as they'd like you to believe, but they have made some.
As far as "think of any feature Linux does not have," that's a little silly. While linux doesn't have any of those specific technologies, it has other technologies that provide the same benefits (OpenGL, GGI, SDL, Corba, LDAP, etc.).
Yeah, but other than drivers, the only crappy software that gets to change core OS DLLs is IE and Office....
And as far as drivers, well, if Microsoft certifies a driver for Win98, and it doesn't work for you in Win98, then either (a) it's Microsoft's fault, or (b) the certification is BS.
In general, the idea is that the judge can do anything if you offer an alternative which is legal, but much worse.
Let's say the judge studies the issues, and decides that he can't legally compel a breakup--but he can produce an injunction against any sales of Win98 or Win2k or any other OS derived from the same sources. Or suspend their corporate charter. Or fine them $300 trillion.
Whatever, he can offer them the choice between the draconian sentence and the breakup, and it's up to them to choose.
I think the breakup would be well within his power--but then I'm not a lawyer.
Also, remember that they can appeal, and the Supreme Court is not guaranteed to be infallible (much less to agree with the currently presiding judge), which makes things even more complicated.
"OTOH - installing IIS, and SQL Server on NT or W2K is easier than building, configuring, and installing apache, php, mysql, etc."
Having done both myself, let's compare. In the first case, you install NT Server 4.0, then install BackOffice 4.0. In the second case, you install RedHat 6.2.
You don't have to build or install any of those packages. But it's nice that you have the option. If I want to recompile apache to use PHP4 instead of PHP3, or remove PHP support, I can (and have). If I want to recompile IIS to remove ASP....
Also, I've found that it's much easier to turn off apache and install, say, Roxen on my linux box than to turn off IIS and install another webserver.
Many of the world's "industrialized democracies" have an explicit concept of a "permitted monopoly." In some industries, there may be a good reason for a monopoly--so the government decides that there will be one, awards it to someone, and regulates them differently than they regulate non-monopolistic companies.
America has no such concept. We still have de facto permitted monopolies, and we do come up with special rules for them, but it's much more capricious. I'm not sure which is better--probably each is better in some ways, worse in others--but it's an interesting distinction.
Everything you say about corporations is true, but it's really a side issue. The big thing corporations provide is limited liability. The people making the decisions don't get punished; the people making the money don't get taxed (at least not as high as they would otherwise).
Instead, the corporation, as a fictitious person, takes the liability. For example, when a corporation commits a crime and gets caught, stockholders and execs may lose money, but they don't go to jail.
Well then, what we really ought to do is make things as bad as possible, to bring on the revolution faster.
This isn't entirely sarcastic--I have a friend who actually believes this. He campaigned for Pat Robertson, helped with the effort to infiltrate the Sierra Club with anti-immigration right-wingers, hired out to harass striking workers, etc., in the hopes that it'll bring closer the day when the people see what's going on and rise up.
I don't agree with him, but it's at least arguable.
One small comment: Globalisation is even less new that you think. It's exactly the same thing that happened a little over a century ago, when corporations began growing too large for individual states to deal with them, and there was no interstate regulatory structure to turn to.
Eventually, the federal government (and national unions, etc.) got into the game, and the problem subsided. So now corporations are growing too large for nations to deal with them, and there's no international regulatory structure to turn to.
Exactly the same thing.
Hell, in Adam Smith's day, the "free market" was entirely within the borders of a single town or village. So we've been through this more than once before....
Cheaper? Let's see, $100 for Win98 vs. $80 for MacOS 9, $80 for BeOS, $50 for RedHat linux on CD, or $0 for RedHat linux by FTP. $1000 for NT4 vs. $400 for AppleShare IP or MacOS X Server, $50 for RedHat linux on CD, or $0 for RedHat linux by FTP. $1000 for IIS (you have to buy NT Server to get it...) vs. $400 or so for a Netscape server vs. $150 for WinApache vs. $0 for apache by FTP. $200 for Visual Basic vs. $100 for RealBasic.
Anyone starting to see a pattern here? Even forgetting about free software, they have the most expensive of the major consumer OS's, the second most expensive of the major small-business server OS (after NetWare), the most expensive Win32 webserver, the most expensive RAD Basic....
I don't know about the prices for Office vs. IBM and Corel's offerings for Windows, or Claris's for MacOS, but I'll offer a guess: Microsoft isn't winning there because of price either.
So why are they winning in those markets? Well, it's not quality, and it's not price. That isn't proof that they're doing something illegal, but it definitely shoots holes in your argument.
Speaking of which, "... never let themselves be reliant on any other corporation or technology." Right, you don't need an Intel-compatible CPU, chipset, or PCI implementation to run Windows; we all know how well-supported the Alpha, MIPS, and PowerPC versions of NT4 are....
"But the browser is part of the O/S now, and not just in Windows."
Really? It's not part of MacOS, or MacOS X, or linux, or BSD, or Solaris, or BeOS, or any other real OS I can think of (maybe JavaOS and WebTV are exceptions?).
True, MacOS 9.0 does install IE 4.5 by default, and RedHat 6.2 does install Netscape 4.7 by default. But in both cases, I can choose not to install--or to remove it later. (Not to mention the fact that MacOS 9.0 also installs Netscape 4.6, and if you're upgrading from 8.5 it even asks you which of the two you'd like as your default browser, or whether you'd like to continue to use your old browser.)
One less relevant comment: "I'm not sure what is so amazing about this [money oiling the world in favor of big companies], since this is how things have worked for a few thousand years." No, it's only been since the industrial revolution--which was only a few _hundred_ years ago. In feudal times, companies were under the iron rule of the king (or local lord)--who could set taxes, revoke your charter, etc. on a whim.
Plus, the modern corporate fiction didn't exist--if your company pissed off the government, they had an easy remedy--chop off your head, or throw you in prison.
Capitalists financed most of the democratic revolutions of the past 300 years, so it may not be surprising that the resulting governments were more capitalist than democratic. But it's still a sad result, and one that has no precedent before industrialism.
Are you sure you're not reminded of one in LA just a few months ago? The LA Times' Times Magazine did a big fluff feature on the Staples Center, and split the ad profits with the Staples Center, and allowed the ad execs to kill (or at least postpone) any stories in the magazine they didn't like.
There was a huge scandal which still wasn't yet resolved when the Chicago Tribune bought out the Times....
The only scandal I remember from a couple of years ago was when the bosses at the paper declared their intentions to "break down the walls" between advertising and "content," and the only result of that "scandal" (other than a lot of hand-wringing by pundits) was an increase in stock prices....
Anyway, back on topic computer magazines make a huge percentage of their profits from advertising (as opposed to magazine sales). So maybe what _you're_ paying for is objectivity, but the advertisers are paying just as much for just the opposite....
Long, long ago. Yeah, Windows98 was, what, 18 months ago? I mean, it was already 3 months old when the trial started--and if the statute of limitations is more than two weeks, I say it's too long.
Let's just forget the whole thing; it was probably just a "youthful indiscretion" that'll never happen again.
One thing: I don't think IE and Outlook should be part of the same company as IIS and Exchange. If the IIS team wants to put in some new non-standard feature, they have to tell all of the browser developers of the world how to use it, and hope the most important ones do--not just sneak support into the next version of IE.
Half a dozen version numbers to check? Nah, it doesn't matter whether you're using IE4 preview release 1 or IE 5.01; they're all version 4.70 to Microsoft.
Yeah, that'd be terrible. It'd be just as bad as it is today! Let's look at the requirements (paraphrased) for the Win32 version of an app I worked on recently:
Windows95 OSR/2 or later with DirectX 3 or later and IE4 or later, Windows98 with the bundled IE4 or IE5.01 or later (but not IE5.0), Windows2000, Windows NT4 with Service Pack 4 or later and either a version of IE4 from SP4 or later or IE4.02 or later. It will work with Windows95A if you install WinSock2 (which we can't distribute, or tell you where to get, and it's not documented, but if you can find it, it'll work) with IE4 or later. It may work with Windows NT 3.51 or Windows 3.1 with Win32s with appropriate additions, but we don't support that.
By comparison, here are the MacOS and linux requirements:
PowerMac: MacOS (7.5 to 9.1alpha), or any version of MacOS X. Requires OpenTransport 1.1 or later (available from this URL), and any web browser configured with any version of Internet Config.
x86 Linux: kernel 2.1 or later, with glibc 2.1. Requires SDL 1.1 (available at this URL). The RPM installation works with RedHat 6.0 or later, and Mandrake 7.0 or later, but not SUSE. Sound support requires any version of OSS, or ALSA 0.2 or later.
Um... Last I checked, Neoplanet just used the IE ActiveX control (the one built in to the operating system).
But more importantly, every _does_ have to use IE if they use Win98/NT/2000. Why? Because it's integrated into the Windows Explorer (which is always running), and it's got tendrils reaching into the kernel, networking, gdi, and user/windows DLLs. Even if I use Netscape to browse the web, IE is always still running. It's still eating memory and CPU. It can still bring down Netscape, or the entire OS.
Now, on my Mac system, IE was installed by default with OS 9.0. But when I'm running Netscape, or Mozilla, or iCab (my default browser), IE isn't running. All it's taking up is drive space. And I can remove it just by deleting a single folder, without affecting anything else.
The same thing goes for my linux boxes, of course, except that IE isn't even an option.
Plus, if you've ever written an Internet app for Win32, you know the horrors that appear the first time QA finally gets a Win95A box to test against....
Of course you can just write "requires WinSock2" in your specs--except that Win95A users can't get WinSock2 except by installing a massive app (like Office 2000) that upgrades behind their backs....
And speaking of Microsoft's TCP stack, it's _still_ the only implementation in the world that makes you explicitly configure your MTU and RWIN to values appropriate for your connection rather (and doesn't even give you a documented way to do it)?
It'd be more courageous to yank the corporate charter, but let them keep their business license, running as a partnership. Then, next time Microsoft breaks the law, they can put the execs in jail.
Even better, yank everyone's corporate charter, and hold the decision-makers accountable rather than corporate fictions. All of the laws would get much simpler.
The US government is never going to pick a single OS. Looking just at.gov and.mil web servers, you'll find linux, BSD, Solaris, AIX, MacOS, MacOS X Server, and probably anything other current OS you can think of. In fact, forget current systems--I know of government software running on VAX/VMS, Amiga 1000s, pre-RISC IBM AS/* systems, etc.
And this is a good thing. If the government mandated "linux everywhere for all federal computers" I'd be just as unhappy as if they mandated "NT everywhere for all federal computers."
Well, not _quite_ as unhappy, but you get the point.
Various different offices are actively migrating away from Windows (to linux, BSD, MacOS, whatever). There are also offices actively migrating to Windows. Pretty much the same as it is in the corporate world.
Yeah, I used to hand-code all my Z80 and 6502 code--hell, I even hand-assembled it and typed it into the computer in hex because the assemblers were either too slow, incomplete, or non-existent. But I was also writing programs that had to fit into 8KB of RAM. Naturally, they didn't do all that much.
Even forgetting about the much larger scale of today's projects, have you ever tried hand-optimizing code for a modern CPU--say, a PowerPC 604e or an Alpha? I guarantee you you'll screw up on the scheduling and end up slowing the whole thing down. And then when you have to port (say, from a 604e to a G3), you have to do the whole thing all over again.
"There is only one datatype, and that is the short integer." Yeah, the 16-bit integer, the datatype that's handled slowest by most modern processors. Good choice. I'd much rather make 100 non-aligned 16-bit moves than 50 aligned 32-bit moves--how else are we going to encourage people to keep upgrading to faster, more expensive machines? And as for floats, pointers, etc.--well, if you treat them as integers, that means we'll never be out of a job....
I know you're exaggerating a little to make a point, but I think the point is wrong-headed. You will not do as good a job if you reinvent the wheel every time. Someone's already done large chunks of it better. And not even necessarily because they're smarter, or have more time. For example, if you want 3D graphics, and you use OpenGL, there's 3D hardware acceleration out there on many video cards; if you project to 2D in your own code and call no functions but setpixel(x,y,r,g,b) you won't even get 2D acceleration....
The funny thing is, the fact that they have access to internal APIs and can even rewrite central pieces of the OS to improve their applications seems to hurt their development efforts, rather than helping. Compare Office98, IE 4.5, and Outlook Express 5.0 on the Mac to the equivalent products on Windows. The Mac versions were all developed faster, by smaller teams. And they're smaller, faster, and more stable. They can be installed just by putting a single file anywhere you want on the hard drive (or even running right off the CD), without going through a major installation process (involving two reboots). And they have all of the important features of the Windows versions plus a few genuinely useful extra features. Why is this? Surely not because Apple's 15-year-old APIs are so much better than Microsoft's 10-year-old APIs. And although the fact that there is a coherent, consistent way to go about writing MacOS apps that everyone follows may help a little, it can't explain everything. I think the major reason is that they're writing to a stable, published API. If they want to, say, draw an animation in a child window, they have to figure out how to do something like that in MacOS--they can't just change the kernel, GUI, and windowing code in the MacOS system file they way they replaced kernel.dll, gdi.dll, and user.dll in Windows. If they had to work the same way within Windows, their Windows products would probably turn out to be better, and cheaper to develop. More benefit for stockholders, employees, and consumers....
Slashdot Mirror
Who will use it? Whoever wins, of course.
Let's say I'm running AMD, and in the impartial benchmarks, our chips beat equivalent-speed Intel chips by 31%. I'd be a fool _not_ to use those benchmarks in my ads--and crow about how I win using impartial, open benchmarks.
In the same way, it Adobe starts to worry about GIMP and releases a linux version of Photoshop, and it blows away GIMP on the same machine using impartial, open benchmarks, they'll use those benchmarks in their ads.
It's like when the PowerPC G4 did so ridiculously well on Byte's benchmarks, Apple's ads make very clear that the benchmarks came from Byte, and were impartial (hell, if anything, Byte was biased _against_ Apple).
In the end, I'm not sure open benchmarks will improve things all that much. In the examples above, Intel can just make sure that the new PIII Argon improves over the PIII Xeon in the specific areas that will matter.
And if there are too many benchmark suites, it just gets confusing. AMD can advertise that they win on benchmarks A, B, and C, and Intel that they win on D, E, and F--and then who do you believe?
I'll assume this isn't a troll, even though it probably is, and try to answer very simply:
1. Your credit card number may be "information," but you don't want everyone to have it.
2. If you send an email to CNN descibing misdeeds of the repressive regime in your country, you probably don't want the local regime to be able to intercept and modify it.
Just because international waters are free doesn't mean pirates can intercept any ship more than 10 miles from shore with impunity.
That joke is so old that the first time I heard it, I had a MillionInstructionsPerSecond processor and the name sounded impressive....
Start Menu: Apple Menu
Spell checking as you type: at least two third-party apps for Word for Macintosh
OLE/DDE/D&D: OpenDoc/Publish&Subscribe/D&D
DirectX: Um... What's new here, exactly? The Mac had a way to do full-screen drawing, DOS allowed you to write directly to the hardware, OpenGL allowed hardware 3D support, etc.
DCOM: Corba
ActiveDirectory: NDS and LDAP
Some of these are more borderline. The basic concepts behind the Intellimouse and the Natural Keyboard already existed, but only in specialized applications. So they deserve some credit there. Also, OLE's linking was much better than P&S's.
There are some much better examples of Microsoft innovation. First, two killer apps for the Macintosh: Word, which had a GUI and WYSIWYG layout at least as good as MacWrite's but was powerful enough to compete with the DOS standards WordStar and WordPerfect, and Excel, a spreadsheet nearly as good as Lotus with a much simpler learning curve (plus better printing support and other improvements).
And it wasn't just in the early days. Win95 and Office95's pervasive, consistent context menus finally made that second mouse button useful for consumer users. ODBC allowed middleware to be ported from Microsoft SQL Server to Oracle in a matter of days. The auto-install/self-repair features in Office98 and IE4.5/OE5.0 for MacOS is the first good implementation of that idea.
In other words, Microsoft has made some innovations. Not as many as some of their competitors, not as many as they'd like you to believe, but they have made some.
As far as "think of any feature Linux does not have," that's a little silly. While linux doesn't have any of those specific technologies, it has other technologies that provide the same benefits (OpenGL, GGI, SDL, Corba, LDAP, etc.).
Yeah, but other than drivers, the only crappy software that gets to change core OS DLLs is IE and Office....
And as far as drivers, well, if Microsoft certifies a driver for Win98, and it doesn't work for you in Win98, then either (a) it's Microsoft's fault, or (b) the certification is BS.
Good question.
In general, the idea is that the judge can do anything if you offer an alternative which is legal, but much worse.
Let's say the judge studies the issues, and decides that he can't legally compel a breakup--but he can produce an injunction against any sales of Win98 or Win2k or any other OS derived from the same sources. Or suspend their corporate charter. Or fine them $300 trillion.
Whatever, he can offer them the choice between the draconian sentence and the breakup, and it's up to them to choose.
I think the breakup would be well within his power--but then I'm not a lawyer.
Also, remember that they can appeal, and the Supreme Court is not guaranteed to be infallible (much less to agree with the currently presiding judge), which makes things even more complicated.
"OTOH - installing IIS, and SQL Server on NT or W2K is easier than building, configuring, and installing apache, php, mysql, etc."
Having done both myself, let's compare. In the first case, you install NT Server 4.0, then install BackOffice 4.0. In the second case, you install RedHat 6.2.
You don't have to build or install any of those packages. But it's nice that you have the option. If I want to recompile apache to use PHP4 instead of PHP3, or remove PHP support, I can (and have). If I want to recompile IIS to remove ASP....
Also, I've found that it's much easier to turn off apache and install, say, Roxen on my linux box than to turn off IIS and install another webserver.
Many of the world's "industrialized democracies" have an explicit concept of a "permitted monopoly." In some industries, there may be a good reason for a monopoly--so the government decides that there will be one, awards it to someone, and regulates them differently than they regulate non-monopolistic companies.
America has no such concept. We still have de facto permitted monopolies, and we do come up with special rules for them, but it's much more capricious. I'm not sure which is better--probably each is better in some ways, worse in others--but it's an interesting distinction.
Everything you say about corporations is true, but it's really a side issue. The big thing corporations provide is limited liability. The people making the decisions don't get punished; the people making the money don't get taxed (at least not as high as they would otherwise).
Instead, the corporation, as a fictitious person, takes the liability. For example, when a corporation commits a crime and gets caught, stockholders and execs may lose money, but they don't go to jail.
Well then, what we really ought to do is make things as bad as possible, to bring on the revolution faster.
This isn't entirely sarcastic--I have a friend who actually believes this. He campaigned for Pat Robertson, helped with the effort to infiltrate the Sierra Club with anti-immigration right-wingers, hired out to harass striking workers, etc., in the hopes that it'll bring closer the day when the people see what's going on and rise up.
I don't agree with him, but it's at least arguable.
One small comment: Globalisation is even less new that you think. It's exactly the same thing that happened a little over a century ago, when corporations began growing too large for individual states to deal with them, and there was no interstate regulatory structure to turn to.
Eventually, the federal government (and national unions, etc.) got into the game, and the problem subsided. So now corporations are growing too large for nations to deal with them, and there's no international regulatory structure to turn to.
Exactly the same thing.
Hell, in Adam Smith's day, the "free market" was entirely within the borders of a single town or village. So we've been through this more than once before....
Cheaper? Let's see, $100 for Win98 vs. $80 for MacOS 9, $80 for BeOS, $50 for RedHat linux on CD, or $0 for RedHat linux by FTP. $1000 for NT4 vs. $400 for AppleShare IP or MacOS X Server, $50 for RedHat linux on CD, or $0 for RedHat linux by FTP. $1000 for IIS (you have to buy NT Server to get it...) vs. $400 or so for a Netscape server vs. $150 for WinApache vs. $0 for apache by FTP. $200 for Visual Basic vs. $100 for RealBasic.
Anyone starting to see a pattern here? Even forgetting about free software, they have the most expensive of the major consumer OS's, the second most expensive of the major small-business server OS (after NetWare), the most expensive Win32 webserver, the most expensive RAD Basic....
I don't know about the prices for Office vs. IBM and Corel's offerings for Windows, or Claris's for MacOS, but I'll offer a guess: Microsoft isn't winning there because of price either.
So why are they winning in those markets? Well, it's not quality, and it's not price. That isn't proof that they're doing something illegal, but it definitely shoots holes in your argument.
Speaking of which, "... never let themselves be reliant on any other corporation or technology." Right, you don't need an Intel-compatible CPU, chipset, or PCI implementation to run Windows; we all know how well-supported the Alpha, MIPS, and PowerPC versions of NT4 are....
"But the browser is part of the O/S now, and not just in Windows."
Really? It's not part of MacOS, or MacOS X, or linux, or BSD, or Solaris, or BeOS, or any other real OS I can think of (maybe JavaOS and WebTV are exceptions?).
True, MacOS 9.0 does install IE 4.5 by default, and RedHat 6.2 does install Netscape 4.7 by default. But in both cases, I can choose not to install--or to remove it later. (Not to mention the fact that MacOS 9.0 also installs Netscape 4.6, and if you're upgrading from 8.5 it even asks you which of the two you'd like as your default browser, or whether you'd like to continue to use your old browser.)
One less relevant comment: "I'm not sure what is so amazing about this [money oiling the world in favor of big companies], since this is how things have worked for a few thousand years." No, it's only been since the industrial revolution--which was only a few _hundred_ years ago. In feudal times, companies were under the iron rule of the king (or local lord)--who could set taxes, revoke your charter, etc. on a whim.
Plus, the modern corporate fiction didn't exist--if your company pissed off the government, they had an easy remedy--chop off your head, or throw you in prison.
Capitalists financed most of the democratic revolutions of the past 300 years, so it may not be surprising that the resulting governments were more capitalist than democratic. But it's still a sad result, and one that has no precedent before industrialism.
Are you sure you're not reminded of one in LA just a few months ago? The LA Times' Times Magazine did a big fluff feature on the Staples Center, and split the ad profits with the Staples Center, and allowed the ad execs to kill (or at least postpone) any stories in the magazine they didn't like.
There was a huge scandal which still wasn't yet resolved when the Chicago Tribune bought out the Times....
The only scandal I remember from a couple of years ago was when the bosses at the paper declared their intentions to "break down the walls" between advertising and "content," and the only result of that "scandal" (other than a lot of hand-wringing by pundits) was an increase in stock prices....
Anyway, back on topic computer magazines make a huge percentage of their profits from advertising (as opposed to magazine sales). So maybe what _you're_ paying for is objectivity, but the advertisers are paying just as much for just the opposite....
Long, long ago. Yeah, Windows98 was, what, 18 months ago? I mean, it was already 3 months old when the trial started--and if the statute of limitations is more than two weeks, I say it's too long.
Let's just forget the whole thing; it was probably just a "youthful indiscretion" that'll never happen again.
One thing: I don't think IE and Outlook should be part of the same company as IIS and Exchange. If the IIS team wants to put in some new non-standard feature, they have to tell all of the browser developers of the world how to use it, and hope the most important ones do--not just sneak support into the next version of IE.
Half a dozen version numbers to check? Nah, it doesn't matter whether you're using IE4 preview release 1 or IE 5.01; they're all version 4.70 to Microsoft.
Yeah, that'd be terrible. It'd be just as bad as it is today! Let's look at the requirements (paraphrased) for the Win32 version of an app I worked on recently:
Windows95 OSR/2 or later with DirectX 3 or later and IE4 or later, Windows98 with the bundled IE4 or IE5.01 or later (but not IE5.0), Windows2000, Windows NT4 with Service Pack 4 or later and either a version of IE4 from SP4 or later or IE4.02 or later. It will work with Windows95A if you install WinSock2 (which we can't distribute, or tell you where to get, and it's not documented, but if you can find it, it'll work) with IE4 or later. It may work with Windows NT 3.51 or Windows 3.1 with Win32s with appropriate additions, but we don't support that.
By comparison, here are the MacOS and linux requirements:
PowerMac: MacOS (7.5 to 9.1alpha), or any version of MacOS X. Requires OpenTransport 1.1 or later (available from this URL), and any web browser configured with any version of Internet Config.
x86 Linux: kernel 2.1 or later, with glibc 2.1. Requires SDL 1.1 (available at this URL). The RPM installation works with RedHat 6.0 or later, and Mandrake 7.0 or later, but not SUSE. Sound support requires any version of OSS, or ALSA 0.2 or later.
Um... Last I checked, Neoplanet just used the IE ActiveX control (the one built in to the operating system).
But more importantly, every _does_ have to use IE if they use Win98/NT/2000. Why? Because it's integrated into the Windows Explorer (which is always running), and it's got tendrils reaching into the kernel, networking, gdi, and user/windows DLLs. Even if I use Netscape to browse the web, IE is always still running. It's still eating memory and CPU. It can still bring down Netscape, or the entire OS.
Now, on my Mac system, IE was installed by default with OS 9.0. But when I'm running Netscape, or Mozilla, or iCab (my default browser), IE isn't running. All it's taking up is drive space. And I can remove it just by deleting a single folder, without affecting anything else.
The same thing goes for my linux boxes, of course, except that IE isn't even an option.
Plus, if you've ever written an Internet app for Win32, you know the horrors that appear the first time QA finally gets a Win95A box to test against....
Of course you can just write "requires WinSock2" in your specs--except that Win95A users can't get WinSock2 except by installing a massive app (like Office 2000) that upgrades behind their backs....
And speaking of Microsoft's TCP stack, it's _still_ the only implementation in the world that makes you explicitly configure your MTU and RWIN to values appropriate for your connection rather (and doesn't even give you a documented way to do it)?
It'd be more courageous to yank the corporate charter, but let them keep their business license, running as a partnership. Then, next time Microsoft breaks the law, they can put the execs in jail.
Even better, yank everyone's corporate charter, and hold the decision-makers accountable rather than corporate fictions. All of the laws would get much simpler.
The US government is never going to pick a single OS. Looking just at .gov and .mil web servers, you'll find linux, BSD, Solaris, AIX, MacOS, MacOS X Server, and probably anything other current OS you can think of. In fact, forget current systems--I know of government software running on VAX/VMS, Amiga 1000s, pre-RISC IBM AS/* systems, etc.
And this is a good thing. If the government mandated "linux everywhere for all federal computers" I'd be just as unhappy as if they mandated "NT everywhere for all federal computers."
Well, not _quite_ as unhappy, but you get the point.
Various different offices are actively migrating away from Windows (to linux, BSD, MacOS, whatever). There are also offices actively migrating to Windows. Pretty much the same as it is in the corporate world.
Some of us aren't coding for 6510's anymore.
Yeah, I used to hand-code all my Z80 and 6502 code--hell, I even hand-assembled it and typed it into the computer in hex because the assemblers were either too slow, incomplete, or non-existent. But I was also writing programs that had to fit into 8KB of RAM. Naturally, they didn't do all that much.
Even forgetting about the much larger scale of today's projects, have you ever tried hand-optimizing code for a modern CPU--say, a PowerPC 604e or an Alpha? I guarantee you you'll screw up on the scheduling and end up slowing the whole thing down. And then when you have to port (say, from a 604e to a G3), you have to do the whole thing all over again.
"There is only one datatype, and that is the short integer." Yeah, the 16-bit integer, the datatype that's handled slowest by most modern processors. Good choice. I'd much rather make 100 non-aligned 16-bit moves than 50 aligned 32-bit moves--how else are we going to encourage people to keep upgrading to faster, more expensive machines? And as for floats, pointers, etc.--well, if you treat them as integers, that means we'll never be out of a job....
I know you're exaggerating a little to make a point, but I think the point is wrong-headed. You will not do as good a job if you reinvent the wheel every time. Someone's already done large chunks of it better. And not even necessarily because they're smarter, or have more time. For example, if you want 3D graphics, and you use OpenGL, there's 3D hardware acceleration out there on many video cards; if you project to 2D in your own code and call no functions but setpixel(x,y,r,g,b) you won't even get 2D acceleration....
The funny thing is, the fact that they have access to internal APIs and can even rewrite central pieces of the OS to improve their applications seems to hurt their development efforts, rather than helping. Compare Office98, IE 4.5, and Outlook Express 5.0 on the Mac to the equivalent products on Windows. The Mac versions were all developed faster, by smaller teams. And they're smaller, faster, and more stable. They can be installed just by putting a single file anywhere you want on the hard drive (or even running right off the CD), without going through a major installation process (involving two reboots). And they have all of the important features of the Windows versions plus a few genuinely useful extra features. Why is this? Surely not because Apple's 15-year-old APIs are so much better than Microsoft's 10-year-old APIs. And although the fact that there is a coherent, consistent way to go about writing MacOS apps that everyone follows may help a little, it can't explain everything. I think the major reason is that they're writing to a stable, published API. If they want to, say, draw an animation in a child window, they have to figure out how to do something like that in MacOS--they can't just change the kernel, GUI, and windowing code in the MacOS system file they way they replaced kernel.dll, gdi.dll, and user.dll in Windows. If they had to work the same way within Windows, their Windows products would probably turn out to be better, and cheaper to develop. More benefit for stockholders, employees, and consumers....