Opera was first publicly released in 1996, a year after IE, 2 years after Netscape, and 3 years after Mosaic. That version does not bear much resemblance to Opera of today, and certainly was not CSS aware.
The first version of Opera to support CSS was 3, which was around 98, though it wasn't until about 3.5 that Opera really had good CSS which was late 98/early 99, roughly the same time as IE4 which also had CSS support.
The real point, however, is that Firefox and IE have more similariteis in CSS rendering than either of them have with opera.
Opera is a problem, because there are relatively few hacks that can be used to get around it's bugs, largely because it tries to support the whole standard (same with Mozilla, btw).
Ironically, IE is easier to to code around because it doesn't even try to support parts of the standard. This makes it relatively easy to add some code IE doesn't recognize, but Mozilla does. Opera becomes a PITA to work around.
The problem with Opera's CSS support is that it tends to break in ways that both IE and Firefox do not, creating yet a third series of hacks you must do to compensate. While Opera's CSS support is good, it's just different enough for it to be a headache.
First, IIS is not installed on Windows 2000 Pro machines by default. It's only installed on Windows 2000 Server machines, and that's largely because much of the server documentation is stored accessed on the local web server.
Second, Netcraft doesn't count random machines with servers on them, it onl counts machines that have registered DNS hostnames, which the vast majority of systems will not have. If you've got a hostname running at www.domain.com, and a server at that IP, then chances are you know it's there.
You are correct that popularity is not the sole reason. Other reasons include: A bias against them in the hacker/cracker community, a single hardware architecture (it's easier to target 100 million machines all running x86 than 100 million running x86, sparc, PPC, MIPS, Alpha, etc..
If all those other factors were not there, I would guess that you would see roughly similar attach rates. Remember that the vast majority of the worms, viruses, attacks, etc.. are taking advantage of a very small number of vulnerabilities.
Oh come off it. We know that nearly all web sites running on Windows are running IIS or an IIS derivative (ie, personal web server, etc...). Sure, Domino and a number of others also run on Windows, but their market share is miniscule (as the survey itself shows).
For nearly all intents and purposes, Windows = IIS.
But, and this is key, none of your arguments really adress the point I waas making, which is that the argument that popularity = Target is invalid because Apache has the largest market share. IIS is the single most populous web server out there on physical machines. That makes it the largest target, especially considering you can count on all those machines running on x86 for buffer overflow exploits.
Look under the section titled Operating Systems used by Computers running public Internet Web Sites, June 2001
While it's 3 years old, the regular survey numbers are not that different from what they are today, so it's probably still relatively accurate.
Granted, this shows servers running Windows, and Apache does run on Windows, but at the time this survey was taken Apache on Windows was not a solid product and was not used much. Pretty much everything is either IIS or an IIS derivitive.
While this survey is from 2001, the number of hosts on apache and IIS are not too different from what they are today. The important part of the survey though, is the section titled: Operating Systems used by Computers running public Internet Web Sites, June 2001. Windows runs > 50% of the web servers.
People always misinterpret the Netcraft numbers. They list HOSTNAMES, not servers. Since Apache is used far more often on multi-site hosting servers, it has a higher percentage, but not a larger market share.
I would find it difficult to believe that you've *NEVER* had a buffer overflow error in any program you've written unless:
1) You've not written any programs (or programs of any complexity) 2) You've only used scripting, interpreted or runtime languages (ie Perl, Java, etc..) 3)... I can't think of any other reason
I would tend to believe that you did have vulnerabilities in your code, and were simply unaware of them. Buffer overflows can sometimes be very difficult to spot, since you must also know the inner workings of libraries and other code which you pass pointers to.
You're right, it's not difficult to avoid the vast majority of buffer overflows, but there are whole classes of subtle overflows that can go undetected in code for decades (for example, not too long a number of such bugs were uncovered in BIND that had been there for 10+ years.)
Sorry, but IIS doesn't have a smaller market share. Considering that a server is vulnerable, not a host, there are more servers that run IIS than run Apache according to a (dated, but probably still relatively accurate) Netcraft study.
IIS runs on about 50% of the physical servers out there.
Further, IIS can be run as a non adminstrator as well, and defaults to this configuration in IIS6, which, btw, has only had 1 moderate vulnerability in it's > 1 year on the market.
OpenDoc was a very viable product, and Novell was charged with porting it to Windows, something they never actually did (well, at least they never finished it). The fact that OpenDoc wouldn't run on the majority platform killed it, thus Novell was directly responsible for OD's death.
Apple and IBM gave up the ghost after it became obvious Novell would not complete it's side of the deal.
As for WordPerfect, Novell sat on their laurels with the product and did almost NOTHING to improve it once they owned it. It withered on the vine, so to speak.
Unfortunately, neither of those cases are actually what you claim they are.
In the first, there was no legal threat. Notice that he wasn't contacted by MS's legal staff, but rather someone from the Windows Media Group. There was no cease and desist, no official MS action. Someone from MS's development team sent an unofficial email to Avery Lee. He should have waited for an official C&D, which probably never would have come, since MS doesn't seem to actually do that sort of thing.
In the second, MS wasn't using the FAT patent against software developers, ie.. their competitors. They were using it against camera manufacturers and other other embedded devices. Even then, it was entirely voluntary, as MS never sued anyone over it.
In order to show that MS, as an organization (not just some people in it that take it upon themselves to police MS's IP) are using IP offensively, you must show C&D's and Lawsuits, or official threats of them.
I hate to defend MS on this, but they have not once ever sued anyone or threatened them offensively with patents. Sure, they've tried to make money on voluntary licensing, but that's a far cry from offensive patent use.
Unfortunately, Novell doesn't have a very good history of making good decisions by itself. Remember that Novell was about to die before they brought in Stone to change things.
The fact of the matter is, Novell has killed EVERYTHING it has ever touched. Everything. WordPerfect - All but Dead. OpenDoc - Dead. USL/Unixware - Dead. Etc.. etc.. etc..
I was rightly concerned when Novell bought Ximian, and even more concerned when they bought SuSE. Apart from the utter stupidity of an american company running the show at a german one (and a brazillian one), Novell doesn't know how to grow a company, how to change, how to adapt.
The best that can be hoped for is that Novell see's how incompetant it is and sell those divisions off again. At least that's what it did with WordPerfect and USL, but it was too late.
Unfortunately, the fallacy is yours. Apache doesn't have the largest market share. Netcraft only counts hostnames, not market share. Apache is used much more heavily in large hosting environments, which means a high hostname to server ratio. IIS is used more for single sites.
According to an older study by netcraft, IIS runs on at least 50% of the servers out there.
Hmm.. by your own admission, the PC you configured was almost 50% faster than the iMac you configured (you claimed the 2Ghz G5 = 3Ghz P4). You configured a 1.6Ghz G5 and a 3Ghz P4.
You also have a 128MB high end video card on the PC and a mid-grade 64MB video card on the Mac. Unfortunately, they don't make a 2Ghz iMac, but the 1.8 costs about $200 more. One might expect a 2Ghz model, if there was such a beast, to be at least another $200 more as well, plus throw in another $100 to compensate for the difference in video cards and you're at least $500 more expensive. You also configured the dell with a Gigabit network adapter, which would add another $50-100 to the cost, so more like $600.
Further, the iMac isn't anywhere near as expandable as the PC is. That costs money, as the PowerMac illustrates.
You are incorrect. Microsoft never disallowed any OEM from shipping Netscape with their computers, and quite a few did. I bought Toshiba laptops with Netscape pre-installed, and there were several others as well.
You are basing your argument on an invalid assumption. You are probably one of those people that completely misinterpreted the (admitedly poorly worded) news stories about MS canceling Compaq's license for Windows because they chose to ship Netscape *instead* of IE. They were free to ship Netscape *IN ADDITION TO* IE, they just couldn't replace IE with Netscape.
Now, whether or not it was right or wrong for MS to do that is largely irrelevant to this argument, as it completely changes the assumptions you are basing your argument on.
If you were a developer, then you must have recognized how limited BeOS's developer support was. You *HAD* to develop in C++, since the OS was written in C++ and based on a C++ object model that was fundamentally tied to the compiler.
You also had to realize how poor the hardware support of BeOS was. It only supported a very small subset of the hardware available didn't seem to have much modern hardware support at all. Further, Be had so few resources to devote to it that even mediocre support was a pipe dream.
It was a damn fine OS, what little of it there was, that's true. But one of the things that makes an OS like BeOS (or AmigaOS, or a whole slew of other OS's) is that they have so few features that they can do what they do really well. As it grows, it bloats, and gets slower. That's the just the way things are. By the time BeOS would have reached a modern Linux or Windows or Apple OS level, it would have performed about the same, at least in my opinion.
Actually, that's not true. No OEM was ever prevented from putting Opera or Netscape on the PC's they sold (In fact, many did include Netscape, including Toshiba and Compaq). Where they ran into trouble was when they tried to remove IE. They were fine as long as they shipped both.
You know, it's this kind of argument that completely kills the credibility of those arguing for anti-trust.
On one hand, you argue that MS artificially keeps the cost of the OS high, and this causes harm to the consumer because they are "extorted" into paying these high prices. Then, you turn around and argue that MS makes the OS too cheap, and thus competitors can't match prices enough to get OEM's to ship their product.
The arguments are largely mutually exclusive, and they don't do anyone arguing against monopolies any favors.
Actually, yes. OLE is covered by this system, and MS settled with Wang for over for $90 million in 1995 (two years before Wang sold the patents to Kodak).
My guess is that.NET may be immune from that earlier lawsuit. Also,.NET and Java work in subtly different, but quite possibly significantly important ways. The Sun patents are specific to two seperate applications working in unison.
MS's.NET (not counting remoting) is largely a single application that loads a runtime shared library, while Sun's java is a standalone virtual machine that communicates with a seperate application (yes, it can work otherwise, but that doesn't change that the current JVM does work as a "plurality" of programs as the patent claims).
I know you're just trolling anyways, but I'll answer your points.
First, in a domain, anyone can log in to any other machine on that domain (unless specifically configured not to). As such, someone can connect with RDC to another computer and log in. It's supposed to do that.
Second, RDC is disabled by default, so either you enabled it, or your admins have set up a policy to enable it (probably so they can remotely administrate it themselves). When the service is enabled, the firewall opens an incoming port. This is part of the UPnP specification, and many home hardware firewalls will do the same thing.
Third, why can another user log you off? They can't, at least not exactly. What they do is force you out of the console, however the user at the console is given a dialog box asking for permission to do so which will time out after a certain period of time (if you're not using the computer, it doesn't matter if someone kicks you out of the console).
You're not logged off, your tasks continue to run. When you come back and unlock the machine, it kicks off the remote user. Only one user is allowed by license to use the PC at a time.
There was a period of SP2 when multiple users could use the PC simultaneously, but MS decided to prevent that.
So the answer is, SP2's not to blame for what you or your admins deliberately configured it for.
Slashdot Mirror
Opera was first publicly released in 1996, a year after IE, 2 years after Netscape, and 3 years after Mosaic. That version does not bear much resemblance to Opera of today, and certainly was not CSS aware.
The first version of Opera to support CSS was 3, which was around 98, though it wasn't until about 3.5 that Opera really had good CSS which was late 98/early 99, roughly the same time as IE4 which also had CSS support.
The real point, however, is that Firefox and IE have more similariteis in CSS rendering than either of them have with opera.
Opera is a problem, because there are relatively few hacks that can be used to get around it's bugs, largely because it tries to support the whole standard (same with Mozilla, btw).
Ironically, IE is easier to to code around because it doesn't even try to support parts of the standard. This makes it relatively easy to add some code IE doesn't recognize, but Mozilla does. Opera becomes a PITA to work around.
The problem with Opera's CSS support is that it tends to break in ways that both IE and Firefox do not, creating yet a third series of hacks you must do to compensate. While Opera's CSS support is good, it's just different enough for it to be a headache.
You might want to look at:
Opera CSS Issues
Actually, no.
First, IIS is not installed on Windows 2000 Pro machines by default. It's only installed on Windows 2000 Server machines, and that's largely because much of the server documentation is stored accessed on the local web server.
Second, Netcraft doesn't count random machines with servers on them, it onl counts machines that have registered DNS hostnames, which the vast majority of systems will not have. If you've got a hostname running at www.domain.com, and a server at that IP, then chances are you know it's there.
You are correct that popularity is not the sole reason. Other reasons include: A bias against them in the hacker/cracker community, a single hardware architecture (it's easier to target 100 million machines all running x86 than 100 million running x86, sparc, PPC, MIPS, Alpha, etc..
If all those other factors were not there, I would guess that you would see roughly similar attach rates. Remember that the vast majority of the worms, viruses, attacks, etc.. are taking advantage of a very small number of vulnerabilities.
Oh come off it. We know that nearly all web sites running on Windows are running IIS or an IIS derivative (ie, personal web server, etc...). Sure, Domino and a number of others also run on Windows, but their market share is miniscule (as the survey itself shows).
For nearly all intents and purposes, Windows = IIS.
But, and this is key, none of your arguments really adress the point I waas making, which is that the argument that popularity = Target is invalid because Apache has the largest market share. IIS is the single most populous web server out there on physical machines. That makes it the largest target, especially considering you can count on all those machines running on x86 for buffer overflow exploits.
Your link does not show that. What it shows is that Apache hosts more sites than IIS, not that there are more Apache servers.
This survey is from June 2001, and shows physical server counts and what OS they run:
http://www.netcraft.com/Survey/index-200109.html
Look under the section titled Operating Systems used by Computers running public Internet Web Sites, June 2001
While it's 3 years old, the regular survey numbers are not that different from what they are today, so it's probably still relatively accurate.
Granted, this shows servers running Windows, and Apache does run on Windows, but at the time this survey was taken Apache on Windows was not a solid product and was not used much. Pretty much everything is either IIS or an IIS derivitive.
Actually, yes. Look at them:
http://www.netcraft.com/Survey/index-200109.html
While this survey is from 2001, the number of hosts on apache and IIS are not too different from what they are today. The important part of the survey though, is the section titled: Operating Systems used by Computers running public Internet Web Sites, June 2001. Windows runs > 50% of the web servers.
People always misinterpret the Netcraft numbers. They list HOSTNAMES, not servers. Since Apache is used far more often on multi-site hosting servers, it has a higher percentage, but not a larger market share.
If there's really such a difference, then why has Mozilla recently had a bunch of similar flaws to IE flaws?
I would find it difficult to believe that you've *NEVER* had a buffer overflow error in any program you've written unless:
... I can't think of any other reason
1) You've not written any programs (or programs of any complexity)
2) You've only used scripting, interpreted or runtime languages (ie Perl, Java, etc..)
3)
I would tend to believe that you did have vulnerabilities in your code, and were simply unaware of them. Buffer overflows can sometimes be very difficult to spot, since you must also know the inner workings of libraries and other code which you pass pointers to.
You're right, it's not difficult to avoid the vast majority of buffer overflows, but there are whole classes of subtle overflows that can go undetected in code for decades (for example, not too long a number of such bugs were uncovered in BIND that had been there for 10+ years.)
Sorry, but IIS doesn't have a smaller market share. Considering that a server is vulnerable, not a host, there are more servers that run IIS than run Apache according to a (dated, but probably still relatively accurate) Netcraft study.
IIS runs on about 50% of the physical servers out there.
Further, IIS can be run as a non adminstrator as well, and defaults to this configuration in IIS6, which, btw, has only had 1 moderate vulnerability in it's > 1 year on the market.
OpenDoc was a very viable product, and Novell was charged with porting it to Windows, something they never actually did (well, at least they never finished it). The fact that OpenDoc wouldn't run on the majority platform killed it, thus Novell was directly responsible for OD's death.
Apple and IBM gave up the ghost after it became obvious Novell would not complete it's side of the deal.
As for WordPerfect, Novell sat on their laurels with the product and did almost NOTHING to improve it once they owned it. It withered on the vine, so to speak.
Unfortunately, neither of those cases are actually what you claim they are.
In the first, there was no legal threat. Notice that he wasn't contacted by MS's legal staff, but rather someone from the Windows Media Group. There was no cease and desist, no official MS action. Someone from MS's development team sent an unofficial email to Avery Lee. He should have waited for an official C&D, which probably never would have come, since MS doesn't seem to actually do that sort of thing.
In the second, MS wasn't using the FAT patent against software developers, ie.. their competitors. They were using it against camera manufacturers and other other embedded devices. Even then, it was entirely voluntary, as MS never sued anyone over it.
In order to show that MS, as an organization (not just some people in it that take it upon themselves to police MS's IP) are using IP offensively, you must show C&D's and Lawsuits, or official threats of them.
I hate to defend MS on this, but they have not once ever sued anyone or threatened them offensively with patents. Sure, they've tried to make money on voluntary licensing, but that's a far cry from offensive patent use.
Unfortunately, Novell doesn't have a very good history of making good decisions by itself. Remember that Novell was about to die before they brought in Stone to change things.
The fact of the matter is, Novell has killed EVERYTHING it has ever touched. Everything. WordPerfect - All but Dead. OpenDoc - Dead. USL/Unixware - Dead. Etc.. etc.. etc..
I was rightly concerned when Novell bought Ximian, and even more concerned when they bought SuSE. Apart from the utter stupidity of an american company running the show at a german one (and a brazillian one), Novell doesn't know how to grow a company, how to change, how to adapt.
The best that can be hoped for is that Novell see's how incompetant it is and sell those divisions off again. At least that's what it did with WordPerfect and USL, but it was too late.
Unfortunately, the fallacy is yours. Apache doesn't have the largest market share. Netcraft only counts hostnames, not market share. Apache is used much more heavily in large hosting environments, which means a high hostname to server ratio. IIS is used more for single sites.
According to an older study by netcraft, IIS runs on at least 50% of the servers out there.
While it may not be rigid by itself, if it attached to two ends and stretched it might be.
Hmm.. by your own admission, the PC you configured was almost 50% faster than the iMac you configured (you claimed the 2Ghz G5 = 3Ghz P4). You configured a 1.6Ghz G5 and a 3Ghz P4.
You also have a 128MB high end video card on the PC and a mid-grade 64MB video card on the Mac. Unfortunately, they don't make a 2Ghz iMac, but the 1.8 costs about $200 more. One might expect a 2Ghz model, if there was such a beast, to be at least another $200 more as well, plus throw in another $100 to compensate for the difference in video cards and you're at least $500 more expensive. You also configured the dell with a Gigabit network adapter, which would add another $50-100 to the cost, so more like $600.
Further, the iMac isn't anywhere near as expandable as the PC is. That costs money, as the PowerMac illustrates.
You are incorrect. Microsoft never disallowed any OEM from shipping Netscape with their computers, and quite a few did. I bought Toshiba laptops with Netscape pre-installed, and there were several others as well.
You are basing your argument on an invalid assumption. You are probably one of those people that completely misinterpreted the (admitedly poorly worded) news stories about MS canceling Compaq's license for Windows because they chose to ship Netscape *instead* of IE. They were free to ship Netscape *IN ADDITION TO* IE, they just couldn't replace IE with Netscape.
Now, whether or not it was right or wrong for MS to do that is largely irrelevant to this argument, as it completely changes the assumptions you are basing your argument on.
If you were a developer, then you must have recognized how limited BeOS's developer support was. You *HAD* to develop in C++, since the OS was written in C++ and based on a C++ object model that was fundamentally tied to the compiler.
You also had to realize how poor the hardware support of BeOS was. It only supported a very small subset of the hardware available didn't seem to have much modern hardware support at all. Further, Be had so few resources to devote to it that even mediocre support was a pipe dream.
It was a damn fine OS, what little of it there was, that's true. But one of the things that makes an OS like BeOS (or AmigaOS, or a whole slew of other OS's) is that they have so few features that they can do what they do really well. As it grows, it bloats, and gets slower. That's the just the way things are. By the time BeOS would have reached a modern Linux or Windows or Apple OS level, it would have performed about the same, at least in my opinion.
Actually, that's not true. No OEM was ever prevented from putting Opera or Netscape on the PC's they sold (In fact, many did include Netscape, including Toshiba and Compaq). Where they ran into trouble was when they tried to remove IE. They were fine as long as they shipped both.
You know, it's this kind of argument that completely kills the credibility of those arguing for anti-trust.
On one hand, you argue that MS artificially keeps the cost of the OS high, and this causes harm to the consumer because they are "extorted" into paying these high prices. Then, you turn around and argue that MS makes the OS too cheap, and thus competitors can't match prices enough to get OEM's to ship their product.
The arguments are largely mutually exclusive, and they don't do anyone arguing against monopolies any favors.
Actually, yes. OLE is covered by this system, and MS settled with Wang for over for $90 million in 1995 (two years before Wang sold the patents to Kodak).
.NET may be immune from that earlier lawsuit. Also, .NET and Java work in subtly different, but quite possibly significantly important ways. The Sun patents are specific to two seperate applications working in unison.
.NET (not counting remoting) is largely a single application that loads a runtime shared library, while Sun's java is a standalone virtual machine that communicates with a seperate application (yes, it can work otherwise, but that doesn't change that the current JVM does work as a "plurality" of programs as the patent claims).
My guess is that
MS's
Have you tried places like this one?
http://www.xtremenotebooks.com/
You can get an Athlon 64 3700+ with 2GB of memory, a 128MB Radeon 9700 Pro and 15.4" display for $1800. No OS.
I wonder if this is going to change their policy in regards to selling Clie's in the US?
I know you're just trolling anyways, but I'll answer your points.
First, in a domain, anyone can log in to any other machine on that domain (unless specifically configured not to). As such, someone can connect with RDC to another computer and log in. It's supposed to do that.
Second, RDC is disabled by default, so either you enabled it, or your admins have set up a policy to enable it (probably so they can remotely administrate it themselves). When the service is enabled, the firewall opens an incoming port. This is part of the UPnP specification, and many home hardware firewalls will do the same thing.
Third, why can another user log you off? They can't, at least not exactly. What they do is force you out of the console, however the user at the console is given a dialog box asking for permission to do so which will time out after a certain period of time (if you're not using the computer, it doesn't matter if someone kicks you out of the console).
You're not logged off, your tasks continue to run. When you come back and unlock the machine, it kicks off the remote user. Only one user is allowed by license to use the PC at a time.
There was a period of SP2 when multiple users could use the PC simultaneously, but MS decided to prevent that.
So the answer is, SP2's not to blame for what you or your admins deliberately configured it for.