As long as I can remember, the minimum RAM as stated by Microsoft would run the OS fine but leave not much room for applications. As soon as you loaded an application that required significant amounts of RAM, the system would start to do heavy paging and performance went to hell. Doubling the "minimum RAM" usually gave OK performance.
I remember the above for Windows 95, Windows 98, Windows NT 4.0 and Windows 2000. Other versions may have the same problem, but with these I have enough experience to confirm the problem.
Back on topic: I think even if IT professionals are aware of Microsoft's tendency to underspecify the hardware requirements, it is no excuse for Microsoft to do so. Especially when dealing with less experienced customers who might take those minimum requirements in good faith.
I could imagine this as a mod of Vega Strike (http://vegastrike.sourceforge.net/). But considering the handful of regular contributors, compared to the effort that goes into a typical commercial MMO, it might take another 10 years to get anywhere.
Besides, unless NASA sponsors a bit more than their logo, I see no reason for developers to dump the existing Vega Strike backstory in favor of a NASA theme. So even if Vega Strike eventually grows into a MMO on its own, the NASA "branding" might not happen.
But the bar needs to be raised a bit more than "a bit".
I think patents are far too easy to get, for far too little technical contribution in return. Some prerequisites that need to be (re-)introduced:
-Patent must significantly improve the state of the art. Must also be non-obvious (on the latter, the US Supreme court shows some encouraging tendencies - more of that please))
-Patent applications must contain instructions on how to build the item, at a level where an average engineer can do it.
-No patents for things that lack a technical contribution. In particular, no patents on business methods.
Now if Congress fails to fix the above items, I think the USA would be better off completely without a patent system.
So if ISO is no longer trustworthy, who is left to say what is a good standard. If the whole standards body has lost credibility, where can we go to find out which standards to really use?
Looks like we have to do a bit of research ourselves. As in -is the standard reasonably complete and concise? By most accounts, OOXML fails there but ODF looks better. That could be a reason to pick ODF if YOU have to support it;-) -is it actually supported? For both formats, there appears to be some support. See http://www.opendocumentfellowship.com/applications and http://en.wikipedia.org/wiki/Office_Open_XML#Application_support. Note that the ODF supporters are mostly Open Source and the OOXML supporters are from the proprietary camp. So depending on the direction your customer/organization leans to, you might not have much choice in the matter...
Spielberg did a passable to good job through most of "AI", up to the moment where the little robot gave up and dropped into the sea. The completely tacky resurrection by aliens, however, ruined the film for me. So I'm a bit skeptical about him making GITS. Riddley Scott (Blade Runner) would be perfect.
IMHO Eve has too much lag for that. For the flight-simulator kind of immersion, it is important that the shipo immediately reacts to the controls. In highly populated Eve systems it often takes a few seconds until the ship moves after klicking - too much for joystick control.
To be fair to CCP, I should add that things have much improved in the last months. Even in Jita, you can usually dock/undock or fly from A to B now without waiting for minutes for your ship to start moving. What used to be extremely frustrating, is now acceptable under the current Click&Wait interface. But before joystick control makes sense for Eve, CCP has to accelerate things a lot more.
This said, I might check out Jumpgate 2 when it comes out. A space MMO with more direct control of the ship sounds good;-)
That was the complaint in the original case against Percy Schmeisser. But according to TFA, Monsanto is threatening lots of farmers over a few Monsanto plants in their field.
I doubt that all of them are secretly aiming to breed Monsanto Plants. In particular those who believe in organic farming. For them the unwanted Monsanto plants are a contamination.
From my experience with EVE I have the impression that their QA is a bit understaffed. There are some visible bugs in the game that have been unfixed for a while, so I presume there are exploitable security bugs to match.
Going the open source route may or may not help them, depending on how much of the data available clientside has to remain hidden from the user: The deep dark secrets they don't want out could be something like players getting info on all objects in a solar system, and the client filtering out what should not bee seen. That would be immediately exploitable by a client that has the filter removed. It would also be poor design, but consistent with the general lagginess of EVE.
But then again, their behaviour indicates that they are not interested in going open source anyway.
Let's assume you replant from your own harvest. Farms use seeds in rather large quantities. For a solid estimate I'd have to ask my brother-in-law, but a metric ton per farm seems a halfway realistic guess. The seed corns are pretty small, under one gram each. So you end up with more than a million seed corns, some of them Monsanto(r) plants because of that field your neighbor planted. How do you propose to sort them out?
If there was an easy way to do it, farmers would just sieve out the Monsanto(r) seeds and use the "pure" corn for replanting.
Now if you're talking about Monsanto replacing the entire corn for replanting, that would get really expensive. Similar to that what they otherwise get paid. Would kill their business model.
So the farmers should be getting all the profits from higher yields while the people who designed the crops should be getting a one time payment? That's nothing but a tip. What if they would like to be paid for their work? You know... as in negotiate what their work is worth.
Usually I would agree, but due to Monsanto's business methods (in particular sueing farmers for having a few Monsanto plants in between due to cross-pollination) I think they deserve to have their property forfeit.
Finally, of course, there are not yet any ISO OOXML implementations in the wild. But there aren't any full implementations of ODF in the wild either. Here's a list of ODF apps, scored on their ODF functionality, and no app achieves a perfect score, not even OO.o. http://opendocumentfellowship.com/applications [opendocume...owship.com]
In summary, many here talk of how horrible OOXML is by citing problems that have been resolved in the approved ISO revision. (I'm amazed that so many here are under the illusion that the problems cited with ECMA OOXML weren't addressed at all in the final ISO version. Under that misunerstanding, I can see where one might be tempted to believe that NO votes switching to YES votes could only be the result of bribery and corruption.)
After looking at the list of ODF apps and assuming the scores are fair, we have plenty of choice in the "works pretty well" (4 out of 5) category. So it is not perfect yet but quite usable. I think this is quite normal for complex software. Even Donald Knuth's Tex system (legendary for being the only large software that is almost certainly bug-free) took years after feature freeze to debug completely.
Except OOXML already is the standard, or at least the spiritual successor to it. Microsoft Word is how 90% of the world creates their documents.
Here we have the company responsible for that 90% (if not more!) wanting to open up their file format and make it an ISO standard, giving the wider global community some sort of say in the process, for the first time ever. There is absolutely no reason to oppose OOXML's adoption as a standard. It already *IS* the standard and any attempts to block it are just idiots sticking their head in the sand.
Well, OOXML is now an ISO standard (IMHO unfortunately as I found the arguments of the critics convincing). That means an opportunity missed: Governments increasingly demand open standards, and that is the kind of power that could end Microsoft's stranglehold on document formats. If all government documents had to be done in ODF (just an example) that would create a pretty large market for word processors supporting ODF, as the vendors working for the government would have to switch as well.
And about technical flaws in the first version: It is perfectly normal that a few things slip through and are discovered in practical use. Ignoring a LARGE number of technical flaws just to complete the fast track process in time is quite irresponsible. I think ISO should have taken OOXML off the fast track and sent it through the regular approval procedure. As it is, they have damaged their reputation in the eyes of many experts.
For a variety of reasons Windows users grew accustomed to running as full administrators. Large vendors (aka customers) made assumptions when developing for Windows. These assumptions cause problems for a Windows end user (aka the customer) trying to use the large vendor's (aka the other customer's) program. If the user calls the vendor the answer is "run as admin". This conflict is only bad for Microsoft because the end user will put usability over security every day and the large vendor may get sick of dealing with "Windows bugs" and choose a different OS to develop (develop, develop, develop) for.
Microsoft was really damned if they did and damned if they didn't. It may well be their own fault (due to the original design of DOS) but unless you have a time machine nobody can change that.
So far I agree, but I wonder if they picked the right time and method to do something about it. Opportunities that come to mind:
-Around 1995, don't do Windows 95, instead bring DirectX and USB to NT. Introduce a DOS emulation on Windows NT to keep most old applications running. For applications outside the emulation, introduce Ubuntu style elevation. Yes it could have been annoying too, but at that time NT was mainly used in larger companies where people are not supposed to install their own stuff anyway. And if support has trouble with it, well... $MEGACORP complaining to the software vendor will have more effect than Joe Sixpack complaining.
-Today, stuff the old API into virtual machines. One per application. Let the applications manipulate their own install directory but nothing else. Use whatever Sci-Fi technology you want for the rest of the OS;-)
As it is, I suspect they are harder on their customers than in both the above scenarios.
In my time at university, I've participated in a group of hi-fi enthusiasts who were, among other things, building a high quality 2-way loudspeaker. During development, at one point we tried assembling the crossover with alligator clips to save time in testing.
As a result, the perceived sound quality took a nosedive. Soldering the same components together gave much better results. Lesson learned: substandard connectors will make a difference you don't like.
But then again, you should get a set of gold-plated connectors for much less than $300. I went with 4mm gold plated lab connectors for my own rig, soldered to no-name coax cables (cannibalized from some defunct PA equipment). The whole stuff cost maybe $40 in today's prices, and I'm quite happy with the results.
In general D/A conversion needs to be performed outside the computer case, in a specialized box.
I think shielding the audio circuits and filtering the power supply can go a long way. The review in TFA confirms this, as they measured a S/N ratio of more than 110 db for the Asus Xonar. This is better than what the CD audio format can give you, and for that double blind tests suggest that the inherent limitations are not audible. In particular, I'm referring to a test by the Stereoplay hi-fi magazin several years ago, where a high quality analog source was digitized, converted back to analog and then compared to the original. Unfortunately the article is not online, so i cannot give you a link:-(
So far, I have pretty good results from LCD computer monitors, at home and in the office. I get to look at five of those regularly, and the only problem so far are minor pixel defects on two of them. These screens have resolutions up to 1600 x 1200 pixels, which is in the ballpark for HD. The later models are also fast enough for viewing video, and 22'' 16:9 models with 1650 x 1050 pixels are becoming commonplace in the market. Prices start at 200 euros ($300),and some of those have sound and HDMI.
Now all that is missing are somewhat larger screen sizes, and you have a nice screen for a HD TV. The rest of the TV should not be that expensive either. I'm pretty sure that we will see 400 euro/$600 HD TV sets at a reasonable size by 2010.
You got it - I forgot to mention that the computer still spends a lot of time on Windows. But it gets rebooted into Linux for online banking. Your trojan scenario is in theory feasible but it seems far-fetched to me, as the trojan would have to do the following things: -detect that Linux is installed on the system. -bring its own ext3 file system driver, as I have none installed on Windows. -manipulate the Linux installation to intercept my connection to the bank (simply reading data from the ext3 partition won't do the job, as I type the PIN/TAN in when needed and not before). The most feasible way might be to exchange the Firefox executable, so it goes to a phishing side that completes the man-in-the-middle attack by grabbing my inputs and sending a different order to the real banking site.
I think that won't happen before phishers get really desperate about increased Windows security.
I don't think the Federal Reserve and the politicians behind it care about GP or other homeowners in too much debt. They care about a possible mass bankruptcy among banks, which could really upset the economic landscape. There are similar tendencies in Europe BTW: A few state-owned banks have are already received bailout money from the government.
But I agree about the problem with bailing out idiots. Fortunately at least some have to pay for their mistakes: Bear Stearns (http://en.wikipedia.org/wiki/Bear_Stearns#Controversy) was sold off for a small fraction of its previous value to prevent a total collapse. In this case, the stockholders paid the price for not choosing a more competent management.
No, you can't. The license does not permit you to move OEM copies to a different host. So if the machine dies, the license did just go up in a puff of smoke. I don't think I have ever seen a retail license on business owned machines, but if you did then that might leave that option open.
Depending on jurisdiction, the license may be unenforceable. Some years ago, Microsoft lost a lawsuit in Germany over unbundling of OEM versions. If I understand the reasoning of the court correctly (IANAL), the Exhaustion of rights (http://en.wikipedia.org/wiki/Exhaustion_of_rights was the deciding factor. In the US, the First Sale Doctrine could apply (ask a lawyer if it trumps the EULA, this seems to vary between states).
-SP4 and the patches since then slipstreamed into the installation CD -the latest DirectX 9 also slipstreamed into the installation CD -LBA 48 enabled by default (currently you have to edit the registry if you want to use harddisks > 128 GByte) -and as you wrote, up-to-date drivers for common hardware.
With these upgrades, I bet it would look pretty good compared to Vista. As it is, I find it a bit annoying to set up a Windows 2000 machine. But I'm still using it, and it might be my last Windows version. Linux looks nicer with every new version of Ubuntu...
I don't think the decision makers at Microsoft ever had the kind of motivation that might drive engineers. Like pride in a job done well. For them it always was about money and how to make the next flashy release with a minimum of engineering effort.
But as most programmers know, you can get away with that for a while but eventually the accumulated spaghetti code will make your product almost unmaintainable. With Vista, it seems the Windows NT platform has reached that point. It has happened before with Windows ME, but back then Microsoft had the NT platform to save them. Lets see what they will come up with now;-)
BTW I disagree on the DRMware: Product Activation in XP is similar enough that I consider it a form of DRM, and it was my main reason not to switch to XP from Windows 2000.
Google deploys Google Apps, the biggest advice is to get everything possible off of the end-point machines and onto the servers that way you don't have to worry about the above.
But make sure those servers are your own servers, not those of an external application vendor. Because, can you really trust that vendor to keep your data safe?
I guess Apple wants to limit Safari for Windows to Macs running Windows via Boot Camp (or whatever it is named now). So it does not completely ban itself.
But I still think it is a stupid decision. They are limiting the market share of their product, in an area where a popular free alternative exists with Firefox. This is not like using MacOS to push hardware sales.
Slashdot Mirror
As long as I can remember, the minimum RAM as stated by Microsoft would run the OS fine but leave not much room for applications. As soon as you loaded an application that required significant amounts of RAM, the system would start to do heavy paging and performance went to hell. Doubling the "minimum RAM" usually gave OK performance.
I remember the above for Windows 95, Windows 98, Windows NT 4.0 and Windows 2000. Other versions may have the same problem, but with these I have enough experience to confirm the problem.
Back on topic:
I think even if IT professionals are aware of Microsoft's tendency to underspecify the hardware requirements, it is no excuse for Microsoft to do so. Especially when dealing with less experienced customers who might take those minimum requirements in good faith.
I could imagine this as a mod of Vega Strike (http://vegastrike.sourceforge.net/). But considering the handful of regular contributors, compared to the effort that goes into a typical commercial MMO, it might take another 10 years to get anywhere.
Besides, unless NASA sponsors a bit more than their logo, I see no reason for developers to dump the existing Vega Strike backstory in favor of a NASA theme. So even if Vega Strike eventually grows into a MMO on its own, the NASA "branding" might not happen.
But the bar needs to be raised a bit more than "a bit".
I think patents are far too easy to get, for far too little technical contribution in return. Some prerequisites that need to be (re-)introduced:
-Patent must significantly improve the state of the art. Must also be non-obvious (on the latter, the US Supreme court shows some encouraging tendencies - more of that please))
-Patent applications must contain instructions on how to build the item, at a level where an average engineer can do it.
-No patents for things that lack a technical contribution. In particular, no patents on business methods.
Now if Congress fails to fix the above items, I think the USA would be better off completely without a patent system.
Looks like we have to do a bit of research ourselves. As in
-is the standard reasonably complete and concise? By most accounts, OOXML fails there but ODF looks better. That could be a reason to pick ODF if YOU have to support it
-is it actually supported? For both formats, there appears to be some support. See
http://www.opendocumentfellowship.com/applications and
http://en.wikipedia.org/wiki/Office_Open_XML#Application_support.
Note that the ODF supporters are mostly Open Source and the OOXML supporters are from the proprietary camp. So depending on the direction your customer/organization leans to, you might not have much choice in the matter...
Spielberg did a passable to good job through most of "AI", up to the moment where the little robot gave up and dropped into the sea. The completely tacky resurrection by aliens, however, ruined the film for me.
So I'm a bit skeptical about him making GITS. Riddley Scott (Blade Runner) would be perfect.
IMHO Eve has too much lag for that. For the flight-simulator kind of immersion, it is important that the shipo immediately reacts to the controls. In highly populated Eve systems it often takes a few seconds until the ship moves after klicking - too much for joystick control.
;-)
To be fair to CCP, I should add that things have much improved in the last months. Even in Jita, you can usually dock/undock or fly from A to B now without waiting for minutes for your ship to start moving. What used to be extremely frustrating, is now acceptable under the current Click&Wait interface. But before joystick control makes sense for Eve, CCP has to accelerate things a lot more.
This said, I might check out Jumpgate 2 when it comes out. A space MMO with more direct control of the ship sounds good
That was the complaint in the original case against Percy Schmeisser. But according to TFA, Monsanto is threatening lots of farmers over a few Monsanto plants in their field.
I doubt that all of them are secretly aiming to breed Monsanto Plants. In particular those who believe in organic farming. For them the unwanted Monsanto plants are a contamination.
From my experience with EVE I have the impression that their QA is a bit understaffed. There are some visible bugs in the game that have been unfixed for a while, so I presume there are exploitable security bugs to match.
Going the open source route may or may not help them, depending on how much of the data available clientside has to remain hidden from the user:
The deep dark secrets they don't want out could be something like players getting info on all objects in a solar system, and the client filtering out what should not bee seen. That would be immediately exploitable by a client that has the filter removed. It would also be poor design, but consistent with the general lagginess of EVE.
But then again, their behaviour indicates that they are not interested in going open source anyway.
Let's assume you replant from your own harvest.
Farms use seeds in rather large quantities. For a solid estimate I'd have to ask my brother-in-law, but a metric ton per farm seems a halfway realistic guess. The seed corns are pretty small, under one gram each. So you end up with more than a million seed corns, some of them Monsanto(r) plants because of that field your neighbor planted.
How do you propose to sort them out?
If there was an easy way to do it, farmers would just sieve out the Monsanto(r) seeds and use the "pure" corn for replanting.
Now if you're talking about Monsanto replacing the entire corn for replanting, that would get really expensive. Similar to that what they otherwise get paid. Would kill their business model.
After looking at the list of ODF apps and assuming the scores are fair, we have plenty of choice in the "works pretty well" (4 out of 5) category. So it is not perfect yet but quite usable. I think this is quite normal for complex software. Even Donald Knuth's Tex system (legendary for being the only large software that is almost certainly bug-free) took years after feature freeze to debug completely.
On the corruption theme, various observers reported rather strange proceedings in their institutions. Usually in the sense that the approval of OOXML was pushed through against the normal rules or against common sense. As a particular suspicious case, see
http://www.infoworld.com/article/08/03/31/Norway-asks-to-suspend-its-Yes-vote-on-OOXML_1.html
You can find more links at Groklaw, check those from end of march in particeular:
http://www.groklaw.net/staticpages/index.php?page=20071217022527429
Governments increasingly demand open standards, and that is the kind of power that could end Microsoft's stranglehold on document formats. If all government documents had to be done in ODF (just an example) that would create a pretty large market for word processors supporting ODF, as the vendors working for the government would have to switch as well.
And about technical flaws in the first version:
It is perfectly normal that a few things slip through and are discovered in practical use. Ignoring a LARGE number of technical flaws just to complete the fast track process in time is quite irresponsible. I think ISO should have taken OOXML off the fast track and sent it through the regular approval procedure. As it is, they have damaged their reputation in the eyes of many experts.
So far I agree, but I wonder if they picked the right time and method to do something about it. Opportunities that come to mind:
-Around 1995, don't do Windows 95, instead bring DirectX and USB to NT. Introduce a DOS emulation on Windows NT to keep most old applications running. For applications outside the emulation, introduce Ubuntu style elevation. Yes it could have been annoying too, but at that time NT was mainly used in larger companies where people are not supposed to install their own stuff anyway. And if support has trouble with it, well... $MEGACORP complaining to the software vendor will have more effect than Joe Sixpack complaining.
-Today, stuff the old API into virtual machines. One per application. Let the applications manipulate their own install directory but nothing else. Use whatever Sci-Fi technology you want for the rest of the OS
As it is, I suspect they are harder on their customers than in both the above scenarios.
In my time at university, I've participated in a group of hi-fi enthusiasts who were, among other things, building a high quality 2-way loudspeaker. During development, at one point we tried assembling the crossover with alligator clips to save time in testing.
As a result, the perceived sound quality took a nosedive. Soldering the same components together gave much better results. Lesson learned: substandard connectors will make a difference you don't like.
But then again, you should get a set of gold-plated connectors for much less than $300. I went with 4mm gold plated lab connectors for my own rig, soldered to no-name coax cables (cannibalized from some defunct PA equipment). The whole stuff cost maybe $40 in today's prices, and I'm quite happy with the results.
I think shielding the audio circuits and filtering the power supply can go a long way. The review in TFA confirms this, as they measured a S/N ratio of more than 110 db for the Asus Xonar.
This is better than what the CD audio format can give you, and for that double blind tests suggest that the inherent limitations are not audible. In particular, I'm referring to a test by the Stereoplay hi-fi magazin several years ago, where a high quality analog source was digitized, converted back to analog and then compared to the original.
Unfortunately the article is not online, so i cannot give you a link
So far, I have pretty good results from LCD computer monitors, at home and in the office. I get to look at five of those regularly, and the only problem so far are minor pixel defects on two of them.
These screens have resolutions up to 1600 x 1200 pixels, which is in the ballpark for HD. The later models are also fast enough for viewing video, and 22'' 16:9 models with 1650 x 1050 pixels are becoming commonplace in the market. Prices start at 200 euros ($300),and some of those have sound and HDMI.
Now all that is missing are somewhat larger screen sizes, and you have a nice screen for a HD TV. The rest of the TV should not be that expensive either. I'm pretty sure that we will see 400 euro/$600 HD TV sets at a reasonable size by 2010.
Correct.
Forgot to mention that.
You got it - I forgot to mention that the computer still spends a lot of time on Windows. But it gets rebooted into Linux for online banking.
Your trojan scenario is in theory feasible but it seems far-fetched to me, as the trojan would have to do the following things:
-detect that Linux is installed on the system.
-bring its own ext3 file system driver, as I have none installed on Windows.
-manipulate the Linux installation to intercept my connection to the bank (simply reading data from the ext3 partition won't do the job, as I type the PIN/TAN in when needed and not before). The most feasible way might be to exchange the Firefox executable, so it goes to a phishing side that completes the man-in-the-middle attack by grabbing my inputs and sending a different order to the real banking site.
I think that won't happen before phishers get really desperate about increased Windows security.
I don't think the Federal Reserve and the politicians behind it care about GP or other homeowners in too much debt. They care about a possible mass bankruptcy among banks, which could really upset the economic landscape. There are similar tendencies in Europe BTW:
A few state-owned banks have are already received bailout money from the government.
But I agree about the problem with bailing out idiots. Fortunately at least some have to pay for their mistakes:
Bear Stearns (http://en.wikipedia.org/wiki/Bear_Stearns#Controversy) was sold off for a small fraction of its previous value to prevent a total collapse. In this case, the stockholders paid the price for not choosing a more competent management.
Fortunately my bank accepts the use of Firefox on Linux for their online banking site (PIN/TAN based). And yes, I DO reboot the computer for this.
But if they didn't, I might change banks over this
Depending on jurisdiction, the license may be unenforceable.
Some years ago, Microsoft lost a lawsuit in Germany over unbundling of OEM versions. If I understand the reasoning of the court correctly (IANAL), the Exhaustion of rights (http://en.wikipedia.org/wiki/Exhaustion_of_rights was the deciding factor.
In the US, the First Sale Doctrine could apply (ask a lawyer if it trumps the EULA, this seems to vary between states).
-SP4 and the patches since then slipstreamed into the installation CD
-the latest DirectX 9 also slipstreamed into the installation CD
-LBA 48 enabled by default (currently you have to edit the registry if you want to use harddisks > 128 GByte)
-and as you wrote, up-to-date drivers for common hardware.
With these upgrades, I bet it would look pretty good compared to Vista. As it is, I find it a bit annoying to set up a Windows 2000 machine. But I'm still using it, and it might be my last Windows version. Linux looks nicer with every new version of Ubuntu...
I don't think the decision makers at Microsoft ever had the kind of motivation that might drive engineers. Like pride in a job done well. For them it always was about money and how to make the next flashy release with a minimum of engineering effort.
;-)
But as most programmers know, you can get away with that for a while but eventually the accumulated spaghetti code will make your product almost unmaintainable. With Vista, it seems the Windows NT platform has reached that point. It has happened before with Windows ME, but back then Microsoft had the NT platform to save them. Lets see what they will come up with now
BTW I disagree on the DRMware: Product Activation in XP is similar enough that I consider it a form of DRM, and it was my main reason not to switch to XP from Windows 2000.
But make sure those servers are your own servers, not those of an external application vendor. Because, can you really trust that vendor to keep your data safe?
I guess Apple wants to limit Safari for Windows to Macs running Windows via Boot Camp (or whatever it is named now). So it does not completely ban itself.
But I still think it is a stupid decision. They are limiting the market share of their product, in an area where a popular free alternative exists with Firefox. This is not like using MacOS to push hardware sales.