On the other hand, a sysadmin that runs a publically accessible FTP server without being aware of the inherent risks like flaky daemons (wu-ftpd anyone?) or having to rely on root-privileged daemons exposed to the outside, not the mention the horrid mess FTP can make of a firewall ruleset, is just plain looking for a quick trip to unemployment.
Thank goodness the major distros no longer ship with services on by default, like they still did about a year ago. Nowadays whoever enables FTP ought to know what they're doing.
It was put in a rather long message that gets displayed when mkreiserfs is done creating a filesystem.
Sounds like exactly the right place to put it. First time I saw it I got a bit of a start, as I usually suspect an error message when I see such amounts of text scrolling by, but it makes sense to put it there.
See any SUID bits there? I don't. BUT! I hear you object: 'it needs root privileges to bind to port 80, so obviously it must run as root'.
To which I can just say that you know fsck all about the modern state of *nix software development. Recent network applications that need root privilege drop it as soon as they have bound to their port. An application doesn't need root to run on a priviledged port, just to bind to it. Apache is the prime example of this practice, and it is by far not the only one. All major daemons do it like this.
Compare this to a well-known webserver that runs as SYSTEM by default and by all accounts is a pain to run unpriviledged.
No, that's the whole point: If action A saves you money, you've already won. The fact that B gets a free ride does not diminish your own profit, and is therefore irrelevant.
This is of course assuming that the potential profit of denying the free ride outweighs the profit of just being open. In the case of small incremental fixes to an already Free application this is likely to be not the case.
Only those who already have significant investments in their own proprietary tools stand to lose more than they gain. Not surprisingly, this is the proprietary software-only business. And given the monopoly, this means Microsoft and its dependents.
The proprietary folks [...] reduce their users' expectations.
Sometimes even by actively manipulating their users' expectations. Think of it: if Microsoft software is so good that those 95% of computer users run it voluntarily, why do they need such whopping big Marketing department?
Not to single out Microsoft of course, every proprietary vendor does it. I sometimes don't know if I should laugh or cry if I see yet another crummy program being advertised as the greatest thing since sliced bread, when I know that a better and Free alternative already exists. Microsoft is just the most visible exponent of this mode of thinking.
Because you are not a trusted source, and the original seller is?
Unless you massively undercut the original source, and your 'customers' don't need any of the added value the original source could provide, people will prefer to get both binaries and source from the original author(s).
That was the way Cygnus Solutions has always operated. The GPLed software coming out of their labs had more value than any copies anywhere else obtained, simply because Cygnus is a trusted source, and they provided additional services (like porting GCC to new architectures).
Think of it: everytime a release of a major piece of software takes place, what is one of the first things you read on Slashdot? That's right, a cry for mirrors, because everyone starts hitting the project homepage. Now think in terms of a marketeer: how many eyeballs is that hitting a single page? How much is a brand and market goodwill worth?
Obviously, simple rational economic thinking is not the determinant factor in the marketplace. In spite of the possibility of getting the same product for cheaper somewhere else, people will prefer to get it, even at a higher fee, from the original supplier. That's contrary to theory, but that implies your theory is wrong, because it does not conform to the observed facts.
If I may recast your excellent post in different words:
Programmers will be paid for their skill, not for their end product.
Does that cut to the heart of the matter? It is basically my view on this (pointless) discussion, and I thought you articulated it much better than I could.
Wizards releases a 3rd generation dice system free uner a license that isn't too dissimilar in spirit from the GPL, and allows the fan work to flourish. Small wonder they resurrected dungeons and dragons after TSR ran it into the ground.
That so?
Try asking Wizards' legal departement what their stance is on you making a Free (as in GPL) suite of uitilities for Dungeon Masters.
Basically, try mentioning anything trademarked by WotC in your source or documentation, and they'll sic the lawyers on you faster than you can say 'cease and desist'.
Of course you could say that your suite of utilities was for the d20 system. Tough luck. Everything that sets Dungeons&Dragons apart from generic d20 is trademarked.
Oh, and WotC retains the exclusive publishing right to any D&D related software. Your suite of utilities will be pulled faster than you can say 'DMCA' if they send a nice letter to your ISP, and they will. Of course, whatever software they publish will be Windows-only, but what do they care about Linux users? We're all pirates anyway, aren't we?
The only viable project I know of has a sourceforge homepage full of disclaimers to the above effect.
Yeah, really nice company. Please don't compare the d20 license with the GPL. It's not about Freedom, it's about free marketing and free R&D for WotC.
Mart (who gave up in the design stage because of this)
Look at the Netherlands for example, where pot is legal to own (in small quantities) and where the number of first-time drug users (and drug related deaths) is declining.
A few years back we had some numbers on that. About 1 million people had used/still used marijuana. Of these 160.000 were regular/heavy users. These numbers had shown a significant growth over the time the survey was taken.
In the same time period, the number of users of 'hard' drugs (heroin, cocaine, amphetamines) had stayed fairly constant.
In other words, this study showed quite clearly that the 'stepping stone effect' that opponents of legalisation use, does not exist.
I can't be bothered to look up the exact links now, but I believe the study was released by our Central Statistics Agency (CBS, Centraal Bureau voor de Statistiek). If you can read Dutch, you might want to try and search their website for more recent numbers. Or you might try our Health Ministry.
I agree that prohibition is a losing proposition. However, the pro-legalisation side of the debate should never lose sight of the downsides of a permissive soft-drug policy. These downsides have been used by conservatives in the Netherlands to try to make the goverment retract our current policies.
Nicely paranoid, but that doesn't add up. I haven't read the directive per se, but the common way to handle this kind of data in Europe is to only make them available to law enforcement that can present a court order for the info.
So, yes, the fact that this much data gets logged is worrisome, and I'd need some iron-clad guarantees from my government to make sure abuse is curbed before I feel comfortable with it, but it is nowhere near as bad as you make it seem.
I will make sure to watch how my government is going to implement these directives.
If you have esd running, try starting Mozilla with $ esddsp mozilla. That will use the esd mixer daemon to make sure that Flash doesn't get an exclusive lock on/dev/dsp.
I believe Arts has a similar wrapper, but I'm not sure.
Actually, if I remember my Unix history correctly, the BBC is not far off the mark. Remember that Unix started as a research project, and that AT&T was legally forbidden to sell it.
Source availability was what created BSD out of the original AT&T sources. It was the major selling point in the prehistory of Unix, the fact that source was available, and even better, the fact that this source was in a (relatively) high-level language, so the same single OS could easily be ported to multiple architectures.
Wasn't it the CSRG at UC Berkeley distributing source that led to the infamous AT&T vs. BSD lawsuit?
Sure, it wasn't Open Source in the modern buzzword sense of the term, but that's beside the point.
Given that it's Sun we're talking about, I hardly think these are Ye Olde Green Screen Terminals. We're talking X Terminals here.
And since the X Window System is quite able to serve different displays simultaneously, I would think that different workers needing different graphics capabilities is no problem whatsoever.
At least at home my desktop system has no problems serving a 1280x1024x32 display to my monitor and a 1024x768x16 to my laptop simultaneously. I expect Sun to have a little more powerful hardware, but the basic concept is the same.
Sun is acting like they invented the roaming profile. Companies have been using them for years.
Uhh? Earth to Microshoft shill: Roaming profiles is an ugly hack to give a basically single-user system the same functionality Unix workstations had for years before that with NFS-mounted/home directories.
You might also want to check whose name is on all NFS-related RFCs. FYI Sun was making networked workstations when Bill was still flogging BASIC and DOS.
Sorry, but that doesn't solve all problems. I'm having problems with DRI and OpenGL acceleration on my Matrox G450. It's not the Athlon/AGP bug, but it is quite probably a VIA chipset timing issue.
I'm a little hesitant to file this on the LKML, as I have already filed it in the DRI bug reporting system. I may post it to the LKML eventually, as I have heard the DRI guys are rather busy, and if it really is a VIA chipset issue, it is not a DRI bug, but a general kernel bug, right?
No, it's not that bad. It's the deposit on the crate that is the biggest part of the total deposit.
A crate of beer sells for about €10,-. The deposit on the crate is about €2,25 and the deposit on the bottles is about €0,07. There are 24 bottles to a crate, you do the math yourself.
I don't know the exact prices, as I am usually not in charge of the beer purchases among my friends. I will know for sure this afternoon, as I need to buy some for my birthday party.
I've already seen a few dutchmen reply about the KPNQWEST situation, so let me just give you the theoretical background on your great post.
Under dutch law, the equivalent of Chapter 11 is called 'surseance van betaling' (delayed payment), which means that the court will grant a company a delay in paying off the creditors, so that it can get its finances in order. As in the US, this is usually bad news and may lead to the next worse condition.
This next condition, the dutch equivalent of Chapter 7, is called 'failliet'. It occurs when 2 or more creditors ask for immediate payment and the debtor can't pay. The creditors can then ask for the court to declare the company bankrupt. A court-appointed curator will then divide up the companies assets among the creditors. It is possible (wat KPNQ did) to enter banktrupcy voluntarily.
Although I know only about dutch banktrupcy law, AFAIK the rest of Europe is more or less the same. I may not be a lawyer, but I did take two years of law school, and I work in securities, so I ought to have this right.
Because most Canadians live within a hundred miles of the US border and speak English, if you tried to release them later in Canada than the US, it would probably make Canadian distribution unprofitable. So you just factor in the time for dubbing anyway.
And because of the Internet, anyone who wants a copy of a movie that's not released because of slow subtitling can have it and download it, making distribution in those parts of the globe less profitable.
Same reasoning. If dubbing in French during the production process is necessary to have a simultaneous Canadian release in order not to cut into profits, then in this day and age the subtitling should be done in the same stage of the production process.
Instead the MPAA wants to hold on to staggered global releases in order to milk the maximum profit out of the theater run, and when technology makes this strategy worthless, they lobby for restrictive laws instead of giving the customer what he wants. And then they act surprised that said customer doesn't seem to respect their 'Intellectual Property' anymore.
Nice post. Good to see the moderators were awake on this one.
Basically what you are describing is the EU Privacy Directive. The gist of the Directive is that companies may not store information on you without telling what they need it for, and not more information than is necessary for the purposes they state. Additionally, they are not allowed to give out your data to third parties without express prior consent. The national laws that implement this directive are backed up by the governments. Some are a little easy on violators, but others are terrifyingly strict.
That's why I submitted this story in the first place; there have been a lot of stories lately about how companies treat personal information, and this was a nice way to show that somewhere in the world there are laws against this, and governments willing to back them up. I think the EU is a bureaucratic monstrosity sometimes, but this they got right.
Which is why my first link was to the original letter by the Member of European Parliament. I trust you are intelligent enough to check the facts I gave you so conveniently a link to.
I'd say your post betrays a little bit of bias too.
Given that the EU was willing to risk a trade war over the privacy directive, I don't think Microsoft is going to get away with a slap on the wrist on this one
They'll have to work very hard at convincing the Commision that Passport will abide by the EU privacy directive to get out of this mess.
Trust a European on this: the EU countries take privacy very seriously. That's not to say that occasional violators don't slip through, but a giant corporation can't just ignore the directive.
Does it have a feature to hide windows from the tasklist? In Sawfish, I can hide a window from the tasklist based on a regular expression matching the window title.
I really like that feature, as I have set XMMS to 'sticky' mode (it appears on all my desktops), and Sawfish hides it for me. If I need XMMS, I just pick it out of the window menu, or I right click on the titlebar of the current top window (which takes it to the back and brings XMMS to the front).
In which case not even Stallman would ask you to call it GNU/Linux. Thank you for proving the point.
If you don't want to call it GNU/Linux, write your own damn init, your own C library and your own fileutils. These are at the moment the essential parts for an Operating System based on the Linux kernel, making that operating system GNU/Linux.
Or, in other words, try booting any Unix -like Operating System without a C library. What's that? Static linking you say? What C library will you use to link statically into your binaries? Now think on what C library is installed on most 'Linux' systems.
Still think Stallman is wrong?
I do confess that for daily conversational use GNU/Linux is unwieldy. I do try to use where appropriate though.
Well, since there are no moderators watching this thread anymore (cross my fingers), I can continue a bit of off-topic posting...
It's funny that you mention Dell, as that was just the OEM I was thinking of when I hit submit. It's not a secret that Dell shamelessly prostitutes itself to become and stay 'most favoured customer' of Microsoft.
However, to give a counterexample, HP is a Logitech customer, and a big one. Most small OEMs I've seen are also heavily into Logitech, although mostly for the mice. If only their keyboards were half as good...
As for compatibility, that's just the IntelliMouse extensions to the PS/2 protocol, that's not much. Logitech has supported that for quite some time now, together with all other makers of input devices. AFAIK, this is because IMPS/2 is superior to plain PS/2. I don't think this is Microsofts market dominance speaking. Logitech still provides its own drivers though.
I agree that Microsoft now puts out some pretty decent hardware. I just plain refuse to buy it on principle. Aside from that, I just plain prefer Logitech myself. My Cordless Desktop Optical mouse/keyboard combo is worth every cent of the € 160 I paid for it (5 year warranty. Yay!).
Slashdot Mirror
True about FTP. I actually overlooked that one.
On the other hand, a sysadmin that runs a publically accessible FTP server without being aware of the inherent risks like flaky daemons (wu-ftpd anyone?) or having to rely on root-privileged daemons exposed to the outside, not the mention the horrid mess FTP can make of a firewall ruleset, is just plain looking for a quick trip to unemployment.
Thank goodness the major distros no longer ship with services on by default, like they still did about a year ago. Nowadays whoever enables FTP ought to know what they're doing.
Mart
It was put in a rather long message that gets displayed when mkreiserfs is done creating a filesystem.
Sounds like exactly the right place to put it. First time I saw it I got a bit of a start, as I usually suspect an error message when I see such amounts of text scrolling by, but it makes sense to put it there.
Mart
Nice that you mention Apache. From my box:
See any SUID bits there? I don't. BUT! I hear you object: 'it needs root privileges to bind to port 80, so obviously it must run as root'.
To which I can just say that you know fsck all about the modern state of *nix software development. Recent network applications that need root privilege drop it as soon as they have bound to their port. An application doesn't need root to run on a priviledged port, just to bind to it. Apache is the prime example of this practice, and it is by far not the only one. All major daemons do it like this.
Compare this to a well-known webserver that runs as SYSTEM by default and by all accounts is a pain to run unpriviledged.
STFU if all you can do is spread outdated FUD.
Mart
No, that's the whole point: If action A saves you money, you've already won. The fact that B gets a free ride does not diminish your own profit, and is therefore irrelevant.
This is of course assuming that the potential profit of denying the free ride outweighs the profit of just being open. In the case of small incremental fixes to an already Free application this is likely to be not the case.
Only those who already have significant investments in their own proprietary tools stand to lose more than they gain. Not surprisingly, this is the proprietary software-only business. And given the monopoly, this means Microsoft and its dependents.
Mart
Sometimes even by actively manipulating their users' expectations. Think of it: if Microsoft software is so good that those 95% of computer users run it voluntarily, why do they need such whopping big Marketing department?
Not to single out Microsoft of course, every proprietary vendor does it. I sometimes don't know if I should laugh or cry if I see yet another crummy program being advertised as the greatest thing since sliced bread, when I know that a better and Free alternative already exists. Microsoft is just the most visible exponent of this mode of thinking.
Mart
Because you are not a trusted source, and the original seller is?
Unless you massively undercut the original source, and your 'customers' don't need any of the added value the original source could provide, people will prefer to get both binaries and source from the original author(s).
That was the way Cygnus Solutions has always operated. The GPLed software coming out of their labs had more value than any copies anywhere else obtained, simply because Cygnus is a trusted source, and they provided additional services (like porting GCC to new architectures).
Think of it: everytime a release of a major piece of software takes place, what is one of the first things you read on Slashdot? That's right, a cry for mirrors, because everyone starts hitting the project homepage. Now think in terms of a marketeer: how many eyeballs is that hitting a single page? How much is a brand and market goodwill worth?
Obviously, simple rational economic thinking is not the determinant factor in the marketplace. In spite of the possibility of getting the same product for cheaper somewhere else, people will prefer to get it, even at a higher fee, from the original supplier. That's contrary to theory, but that implies your theory is wrong, because it does not conform to the observed facts.
Mart
If I may recast your excellent post in different words:
Does that cut to the heart of the matter? It is basically my view on this (pointless) discussion, and I thought you articulated it much better than I could.
Mart
That so?
Try asking Wizards' legal departement what their stance is on you making a Free (as in GPL) suite of uitilities for Dungeon Masters.
Basically, try mentioning anything trademarked by WotC in your source or documentation, and they'll sic the lawyers on you faster than you can say 'cease and desist'.
Of course you could say that your suite of utilities was for the d20 system. Tough luck. Everything that sets Dungeons&Dragons apart from generic d20 is trademarked.
Oh, and WotC retains the exclusive publishing right to any D&D related software. Your suite of utilities will be pulled faster than you can say 'DMCA' if they send a nice letter to your ISP, and they will. Of course, whatever software they publish will be Windows-only, but what do they care about Linux users? We're all pirates anyway, aren't we?
The only viable project I know of has a sourceforge homepage full of disclaimers to the above effect.
Yeah, really nice company. Please don't compare the d20 license with the GPL. It's not about Freedom, it's about free marketing and free R&D for WotC.
Mart (who gave up in the design stage because of this)
A few years back we had some numbers on that. About 1 million people had used/still used marijuana. Of these 160.000 were regular/heavy users. These numbers had shown a significant growth over the time the survey was taken.
In the same time period, the number of users of 'hard' drugs (heroin, cocaine, amphetamines) had stayed fairly constant.
In other words, this study showed quite clearly that the 'stepping stone effect' that opponents of legalisation use, does not exist.
I can't be bothered to look up the exact links now, but I believe the study was released by our Central Statistics Agency (CBS, Centraal Bureau voor de Statistiek). If you can read Dutch, you might want to try and search their website for more recent numbers. Or you might try our Health Ministry.
I agree that prohibition is a losing proposition. However, the pro-legalisation side of the debate should never lose sight of the downsides of a permissive soft-drug policy. These downsides have been used by conservatives in the Netherlands to try to make the goverment retract our current policies.
MartNicely paranoid, but that doesn't add up. I haven't read the directive per se, but the common way to handle this kind of data in Europe is to only make them available to law enforcement that can present a court order for the info.
So, yes, the fact that this much data gets logged is worrisome, and I'd need some iron-clad guarantees from my government to make sure abuse is curbed before I feel comfortable with it, but it is nowhere near as bad as you make it seem.
I will make sure to watch how my government is going to implement these directives.
MartIf you have esd running, try starting Mozilla with $ esddsp mozilla. That will use the esd mixer daemon to make sure that Flash doesn't get an exclusive lock on /dev/dsp.
I believe Arts has a similar wrapper, but I'm not sure.
Mart (who got bitten by this too)Actually, if I remember my Unix history correctly, the BBC is not far off the mark. Remember that Unix started as a research project, and that AT&T was legally forbidden to sell it.
Source availability was what created BSD out of the original AT&T sources. It was the major selling point in the prehistory of Unix, the fact that source was available, and even better, the fact that this source was in a (relatively) high-level language, so the same single OS could easily be ported to multiple architectures.
Wasn't it the CSRG at UC Berkeley distributing source that led to the infamous AT&T vs. BSD lawsuit?
Sure, it wasn't Open Source in the modern buzzword sense of the term, but that's beside the point.
MartGiven that it's Sun we're talking about, I hardly think these are Ye Olde Green Screen Terminals. We're talking X Terminals here.
And since the X Window System is quite able to serve different displays simultaneously, I would think that different workers needing different graphics capabilities is no problem whatsoever.
At least at home my desktop system has no problems serving a 1280x1024x32 display to my monitor and a 1024x768x16 to my laptop simultaneously. I expect Sun to have a little more powerful hardware, but the basic concept is the same.
MartUhh? Earth to Microshoft shill: Roaming profiles is an ugly hack to give a basically single-user system the same functionality Unix workstations had for years before that with NFS-mounted /home directories.
You might also want to check whose name is on all NFS-related RFCs. FYI Sun was making networked workstations when Bill was still flogging BASIC and DOS.
Mart (Who suspects he just got trolled)
Sorry, but that doesn't solve all problems. I'm having problems with DRI and OpenGL acceleration on my Matrox G450. It's not the Athlon/AGP bug, but it is quite probably a VIA chipset timing issue.
I'm a little hesitant to file this on the LKML, as I have already filed it in the DRI bug reporting system. I may post it to the LKML eventually, as I have heard the DRI guys are rather busy, and if it really is a VIA chipset issue, it is not a DRI bug, but a general kernel bug, right?
MartNo, it's not that bad. It's the deposit on the crate that is the biggest part of the total deposit.
A crate of beer sells for about €10,-. The deposit on the crate is about €2,25 and the deposit on the bottles is about €0,07. There are 24 bottles to a crate, you do the math yourself.
I don't know the exact prices, as I am usually not in charge of the beer purchases among my friends. I will know for sure this afternoon, as I need to buy some for my birthday party.
MartI've already seen a few dutchmen reply about the KPNQWEST situation, so let me just give you the theoretical background on your great post.
Under dutch law, the equivalent of Chapter 11 is called 'surseance van betaling' (delayed payment), which means that the court will grant a company a delay in paying off the creditors, so that it can get its finances in order. As in the US, this is usually bad news and may lead to the next worse condition.
This next condition, the dutch equivalent of Chapter 7, is called 'failliet'. It occurs when 2 or more creditors ask for immediate payment and the debtor can't pay. The creditors can then ask for the court to declare the company bankrupt. A court-appointed curator will then divide up the companies assets among the creditors. It is possible (wat KPNQ did) to enter banktrupcy voluntarily.
Although I know only about dutch banktrupcy law, AFAIK the rest of Europe is more or less the same. I may not be a lawyer, but I did take two years of law school, and I work in securities, so I ought to have this right.
Mart
And because of the Internet, anyone who wants a copy of a movie that's not released because of slow subtitling can have it and download it, making distribution in those parts of the globe less profitable.
Same reasoning. If dubbing in French during the production process is necessary to have a simultaneous Canadian release in order not to cut into profits, then in this day and age the subtitling should be done in the same stage of the production process.
Instead the MPAA wants to hold on to staggered global releases in order to milk the maximum profit out of the theater run, and when technology makes this strategy worthless, they lobby for restrictive laws instead of giving the customer what he wants. And then they act surprised that said customer doesn't seem to respect their 'Intellectual Property' anymore.
MartNice post. Good to see the moderators were awake on this one.
Basically what you are describing is the EU Privacy Directive. The gist of the Directive is that companies may not store information on you without telling what they need it for, and not more information than is necessary for the purposes they state. Additionally, they are not allowed to give out your data to third parties without express prior consent. The national laws that implement this directive are backed up by the governments. Some are a little easy on violators, but others are terrifyingly strict.
That's why I submitted this story in the first place; there have been a lot of stories lately about how companies treat personal information, and this was a nice way to show that somewhere in the world there are laws against this, and governments willing to back them up. I think the EU is a bureaucratic monstrosity sometimes, but this they got right.
MartWhich is why my first link was to the original letter by the Member of European Parliament. I trust you are intelligent enough to check the facts I gave you so conveniently a link to.
I'd say your post betrays a little bit of bias too.
MartGiven that the EU was willing to risk a trade war over the privacy directive, I don't think Microsoft is going to get away with a slap on the wrist on this one
They'll have to work very hard at convincing the Commision that Passport will abide by the EU privacy directive to get out of this mess.
Trust a European on this: the EU countries take privacy very seriously. That's not to say that occasional violators don't slip through, but a giant corporation can't just ignore the directive.
MartNote that I wasn't talking about that. I was talking about essential. Try running all of the above without a C library.
Now who's not getting the picture?
MartDoes it have a feature to hide windows from the tasklist? In Sawfish, I can hide a window from the tasklist based on a regular expression matching the window title.
I really like that feature, as I have set XMMS to 'sticky' mode (it appears on all my desktops), and Sawfish hides it for me. If I need XMMS, I just pick it out of the window menu, or I right click on the titlebar of the current top window (which takes it to the back and brings XMMS to the front).
MartIn which case not even Stallman would ask you to call it GNU/Linux. Thank you for proving the point.
If you don't want to call it GNU/Linux, write your own damn init, your own C library and your own fileutils. These are at the moment the essential parts for an Operating System based on the Linux kernel, making that operating system GNU/Linux.
Or, in other words, try booting any Unix -like Operating System without a C library. What's that? Static linking you say? What C library will you use to link statically into your binaries? Now think on what C library is installed on most 'Linux' systems.
Still think Stallman is wrong?
I do confess that for daily conversational use GNU/Linux is unwieldy. I do try to use where appropriate though.
MartWell, since there are no moderators watching this thread anymore (cross my fingers), I can continue a bit of off-topic posting...
It's funny that you mention Dell, as that was just the OEM I was thinking of when I hit submit. It's not a secret that Dell shamelessly prostitutes itself to become and stay 'most favoured customer' of Microsoft.
However, to give a counterexample, HP is a Logitech customer, and a big one. Most small OEMs I've seen are also heavily into Logitech, although mostly for the mice. If only their keyboards were half as good...
As for compatibility, that's just the IntelliMouse extensions to the PS/2 protocol, that's not much. Logitech has supported that for quite some time now, together with all other makers of input devices. AFAIK, this is because IMPS/2 is superior to plain PS/2. I don't think this is Microsofts market dominance speaking. Logitech still provides its own drivers though.
I agree that Microsoft now puts out some pretty decent hardware. I just plain refuse to buy it on principle. Aside from that, I just plain prefer Logitech myself. My Cordless Desktop Optical mouse/keyboard combo is worth every cent of the € 160 I paid for it (5 year warranty. Yay!).
Thanks for a nice discussion,
Mart