In hindsight, I see that you're right. If a compatable DVD burner was released, you could potentially 'remaster' a unencrypted DVD which was playable on normal DVD decks. However, this depends on the ability/availablity of the remastering software, as well as hardware features that might prevent a full quality dub. As long as the film industry is designing the computer drives, I would expect problems like these to be designed in.
Despite the current troubles, I think the industry is most worried about technology that could make widespread DVD available to the deck-using masses. As long as it's only computer users playing VOB files of a NTFS/ext2-formatted optical media, they won't be too worried. --
Yes, but MS's strategy is to encourage whenever possible to avoid writing raw SQL. They have embraced ANSI SQL, but the "extend" part is OLEDB, ADO, DTS, MS OLAP, and the numerous other services that ride on top of MS-SQL. --
What the question doesn't tell us is what the current backend is... My guess is Paradox, FoxPro or some other XBase system.
So while it's true you can point fingers at MSSQL and say that it isn't as good as Oracle or DB2, in this case it's a probably huge improvement over the existing system.
MSSQL 7 handles small datastores without much trouble, and with virtually no DB administration required. It's often used in 'turnkey' packages like accounting software for this reason.
It sucks that you have to support a new platform (NT) just for one application, but as others have said, the only real alternative is to switch vendors. --
Yup -- and if consumer tape is $2/Gig now, imagine what it will be in a couple years.
Right now the computer industry is sitting on it's hands and waiting collectively for DVD-R/W to stabilize and a compatible version of DVD-R to be released. But while the DVD people are dinking around, they won't wait forever.
I can imagine a situation in a year or two where Iomega, Creative and others in the industry get sick of waiting for writeable DVD and just release their own ~20 Gig disc format. Base it on existing CD-R technology, call it "SuperCDR", and push it out to the consumer at a reasonable cost. It would be designed for computer storage and backup, but of course nothing would stop you from putting VOB files on it. The only thing it couldn't do is play DVDs, but by then an IDE DVD player will probably be about $35 --
At this point, Intel and AMD are competing so hard on speed and price that neither can afford to undermark.
Sure competition between the two is tough, but Intel and AMD aren't stupid either. If they wanted to, they could blow their load and manufacture 1200Mhz chips tomorrow, and still make money.
(Why? Because I purchased a 400Mhz PII in 1998, I probably wouldn't upgrade to any new CPU this year. However next year, 1200Mhz might seem worth it. Businesses think this way, systematically replacing desktops every 18-36 months, and buying whatever middle shelf system is available at the time.)
They get far more revenue by stepping up the speed in a slow and predictable manner. The competition has just forced them to increase their schedules slightly. However, if they increase speeds too much, the buyers of midline chips will suddenly get far more price sensitive, which leads to lower profits. (This has already started happening with folks who decide a Celeron/K6 is 'good enough'.)
The key here from a $ standpoint is the sweet spot at the middle of the market (currently 500-600Mhz), not the dick-sizing going on at the top end. --
The number of people with overclocked Celerons is miniscule compared to the number of non-clocked Celerons humming away on business desktops.
Both Intel and AMD are well aware of overclocking, but they will continue market towards the general case (business desktops) and not the specific (screwdriver guys). The overclockability of a specific chip will continue to depend on the fab output and the overall design of the chip. The loss of sales from overclocking is probably insignificant -- in fact Intel probably sold quite a few Celerons to overclockers that otherwise wouldn't have upgraded their existing systems.
Basically the Celeron was just a nice convergance of good fab capabilities (Intel tries to keep the PII numbers on schedule), and marketing conditions (businesses demand cheap desktops). It could happen again, but then again it could not. --
Mod chips are feasible for the Playstation because there are a billion identical PSX units out there. While you could probably mod a DVD player, selling the kit would be hard because you probably would need a unique kit for every different make and model of player.
I could see DVD mod kits, but they would probably be expensive and marketed towards serious piracy operations.
It's probably a moot point anyway, because in the next couple years we'll have inexpensive computer backup media (removable disk, fast tape, etc) that can easily handle 10 GB of data. This will allow individuals to engage in small time piracy with other computer users, just not the supposedly with more lucrative a/v component DVD player market. --
Anyway, manufacturers are more interested in our marketing, sales numbers, and a product fulfilling a need.
So, basically your arguments about piracy are moot. If the (Linux,OS2,Be,BSD,Solaris,MSDOS) markets were big enough to be profitable, people would sell players.
And while you are correct that the percentage of Linux users saavy enough to use DeCSS is higher than the percentage of Windows users, the saavy-enough Windows users far outnumber the saavy-enough Linux users. If piracy was a serious consideration, there would be no software players at all, or the computer DVD setups would be self-contained sealed hardware boxs. --
Actually, the root problem is that the menuing/control system on DVD-video is just plain braindead and broken. Even obvious things like a pointer/mouse interface were left out.
It's clear that DVD was designed only with playback on an A/V component in mind. They ignored 15 years of research and practice in computer-based multimedia systems and went for a cheesy postcard interface. (The hardware component focus also explains the weak encryption design -- it looks as if they were more worried about people with sodering irons than the computer programming community.)
In hindsight it's obvious that a good portion of their installed base would be DVD-equipped computer systems. The user base is just that much more technically saavy and open to early adoption. Unfortunately, that user base is saddled with a UI designed for people who can't stop their VCR clocks from blinking. What's worse is they ignored the long running predictions of television-computer 'convergence' just when it's finally starting to happen. --
The main reason that binary-only is bad on Linux is that the driver interface is completely unstable. Any particular binary driver works only with a single kernel revision, and is obsolete practically as soon as it's released.
This is a political question -- Linus won't create a stable driver interface precisely to force driver vendors to release the source. If you donate the source, maintenance is freely performed by the Linux community. If you don't, it's a support nightmare.
There's probably quite a few Linux users who don't like this approach, but ultimately, it's Linus' call. --
Hmmm, I see nothing in the GPL which implies that the "You" is a person and not a company or organisation. If the NSA is the licencee of GPL software, it would seem that they as an organization are responsible for following the terms. Letting employees use the software doesn't seem like "distribution" (at least not in the normal sense used in software licences.)
But even if you are correct, most employees have employment agreements dictating what they can do with intellectual property owned by the company. So while they may have the right to redistribute under the GPL, they probaby don't have that right as an employee. With the NSA, such employment agreements have the force of law behind them -- making any internal NSA practice or product public probably carries a very long prison term (even when doing so might be legal in the private sector). --
1) Office Suite integration -- The only thing innovative here was the marketing. Buy a word proc and a spreadsheet and get a slideshow program for free and a desktop database at a reduced price. Plus the cost of MS Office was far below the combined costs of WordPerfect, Lotus 123, and DBase.
2) Linking and Embedding -- An early Mac program, Lotus Jazz had something like this. I think VisiOn did something similar under DOS. What was innovative was making it a universal operating system service -- see #3
3) COM -- Not a new idea, but since MS owns the OS, the desktop apps, the server apps, and the devtools, the integration buys users a lot. Almost no body runs any sort of Corba services on their desktops, but every Windows user uses COM.
4) Internet Ready -- Yawn. If IE didn't exist, they would ship Netscape just like anyone else. Microsoft's 'innovations' in this area, like "Channels" have been roundly rejected.
5) VBA -- The only really interesting thing about VBA is that you can rip the VB part out and use JavaScript or PerlScript. I think one would have quite a bit of trouble arguing that MS stuff is more "scriptable" than virtually anything else. --
Agreed -- Gates has famously said that they are "betting the company" on Windows 2000, and to some extent they are right... If W2000 succeeds as they'd like, Microsoft will become as important as IBM or Sun in enterprise computing; if it fails, Microsoft stays in business but becomes just another Novell hanging onto 'departmental' boxes. Having Gates on the sales team can make a huge difference, especially considering that the upgrade costs are stratopheric enough to require executive level approval in most shops.
(On a side note, if anyone should be fired over Win2000, it should be the people who speced out the product long before engineering had begun. Features like ActiveDirectory have been promised RSN for about 5 years. Microsoft's FUD machine got way way out in front of it's engineering ability. The fact that they've delivered a product that appears to meet marketing's fantasy land specifications is pretty amazing. I know quite a few of you slashdotters have opioned that they wouldn't be able to do it. Now the question is, can they sell it?) --
You are correct that *network browsing* (via the Network Neighborhood or NET VIEW) doesn't work over a routed network without a master browser/WINS setup. However, with LanMan networking, don't confusing browsing with actual accessibility. --
Well for 10 years we've been hearing how the NeXT development environment is one of the best ever invented. If true, why would Apple dump what they've got for a third party tool? (Does anyone know if NeXT IDE comes for free with OS X? I'd be a dumb move if it wasn't.)
I'm sure they consider Metroworks very important for Carbon projects, but the momentum is clear -- new projects should be Cocoa, which means Objective C or Java which means not Metrowerks.
From Apple's point of view, it frees them from yet another Third Party Developer that can crap all over whatever their strategy is. When Symantac started cutting back on Mac development, things got shaky. Having Code Warrior appear was pure luck for Apple, and I don't think they want to go through that again. (As another reference, there's a/. poster that has bitched a few times about MS/IBM/Borland/Novell politics in the early days of OS/2. Things like that can cast a cloud over a new platform that it might never recover from.) --
"Keep all your versions up to date and you can conquer Microsoft."
True -- most WordPerfect and SmartSuite shops I've seen have been at least one or two versions behind. Consequentally the users are bitching about the lack of functionality and import/export filters. (It's hard to import Office 97 files when you are running SmartSuite 96)
Usually this is intentional IT politics. They want to switch, but can't justify the cost. So they let the functionality slide a for a few years, and wait for everyone to start begging for MS Office. --
Look at kernel32.dll, user32.dll, gdi32.dll, krnl386.exe, user.exe and gdi.exe sometime in a hex editor. It's scary how many "exported" function calls are there that aren't documented by ANYONE.
Those file, as a whole, are the Win 9x kernel. Microsoft, for obvious reasons, doesn't want to document internal kernel interfaces.
Look at the Linux situation -- Linus doesn't document internal kernal interfaces, and reserves the right to change them. If you write a closed source driver that uses an internal interface, it will eventually break, and that's your problem. MS doesn't have the luxury to treat closed source developers that way.
If you don't like internal, undocumented interfaces, don't use closed source software. It's as simple as that, and has nothing to do with MS as a monopoly.
Your points about MSSQL leads to Scatter/Gather IO (and so on) hold, but I doubt it's above and beyond what any other OS vendor does. For example, Linux provided interfaces to deal with Apache performance problems specifically after the Mindcraft benchmarks were published. --
AT&T break up was done the way you say. The goverment turned one big monopoly into several little ones, that eventually became big again.
A regional breakup of the phone company made sense, because by-in-large their assets are wires, switches, and franchises -- all fixed in a particular spot. Now with local compeitition plans, some can get different phone service, but the facilities (and most of the operation costs) are still owned by the RBOC. You are essentially just paying someone else for the billing service.
Likewise, with MS -- the 'natural' breakup is along operational lines. However, you are correct that it doesn't have to be this way, because MS's property is primarily intellectual and unlike phone facilities can be transfered for little cost. But, that doesn't mean it's the right way to do it.
To recap, If we break up horizontally, then one of the OS/apps can become the dominant one and we will be in trouble again.
Right -- we would be in the exact same situation as we were with MS-DOS versus IBM DOS versus DR-DOS. The minor incompatibilities weren't worth the hassle of switching vendors (ignoring the Win3.1 beta warning about DR-DOS for the moment.)
Windows NT is not the be-all-end-all and neither is MS Office. (If it were, we wouldn't be yacking on a pro-Linux/OSS board.) Let MS-OS and MS-Apps have their little monopolies -- "Microsoft" loses their single vendor status and any theoretical integration benefits. That alone is huge enough to allow competitors in the door. --
I've heard it said a few times that some of the IBM/MS DOS 1.0 tools had the string "(C) Digital Research" embedded in them.
Considering QDOS was one guy's project to port CP/M to the 8086, we know that it wasn't "clean room" reverse engineering like Linux. (Microsoft, at that time a pretty small company, also had access to CP/M source, as they did the Apple II Softcard port.)
Gary Kildall thought he could have sued IBM & MS, but chose not to. It had nothing to do with look and feel. I don't know why Caldera didn't sue --perhaps the statue of limitations had run out, or they figured that the DOS 1.x market was so small that it wouldn't have been worth it. (DOS 2.x was a major rewrite.) --
OK, the alternative scenario is that everyone buys a new TV in the next 5 years, and we spend the rest of our days watching "Urkel" in full photorealistic resolution. The broadcasters happily give the analog spectrum back to the government. Then Heidi Klum falls in love with me and I win the lotto. It's going to be a glorious future! --
The airwaves are owned by the public, so the theory is that the government is ensuring that they are put to wise use. Five different AM Stereo formats wasn't in anyone's best interest except the people trying to kill AM stereo.
Computer standards on the other hand are controlled by Intel and Microsoft, there is no implied public interest. --
AM Stereo got shot down because the people who own the FM stations also own the larger AM stations. Allowing every religious and foreign language station to broadcast in 'hi-fi' would have killed the market value of the FM stations, so the station owners kept AM stereo in the bottle.
For a few years in the 1980s, American car companies shipped AM Stereo capable recievers. Most people didn't notice, however, because nobody tried to broadcast hifi pop music on AM. --
Slashdot Mirror
In hindsight, I see that you're right. If a compatable DVD burner was released, you could potentially 'remaster' a unencrypted DVD which was playable on normal DVD decks. However, this depends on the ability/availablity of the remastering software, as well as hardware features that might prevent a full quality dub. As long as the film industry is designing the computer drives, I would expect problems like these to be designed in.
Despite the current troubles, I think the industry is most worried about technology that could make widespread DVD available to the deck-using masses. As long as it's only computer users playing VOB files of a NTFS/ext2-formatted optical media, they won't be too worried.
--
Yes, but MS's strategy is to encourage whenever possible to avoid writing raw SQL. They have embraced ANSI SQL, but the "extend" part is OLEDB, ADO, DTS, MS OLAP, and the numerous other services that ride on top of MS-SQL.
--
MSSQL doesn't require NT authentication. That means there is absolutely no need to have PDC/BDC boxes.
An NT domain is really only required to support a bunch of Microsoft infrastructure. One SQL server doesn't really justify an entire domain.
--
MS SQL 7 was a huge rewrite, and any application targetted to version 7 is no longer Sybase-compatible.
--
What the question doesn't tell us is what the current backend is ... My guess is Paradox, FoxPro or some other XBase system.
So while it's true you can point fingers at MSSQL and say that it isn't as good as Oracle or DB2, in this case it's a probably huge improvement over the existing system.
MSSQL 7 handles small datastores without much trouble, and with virtually no DB administration required. It's often used in 'turnkey' packages like accounting software for this reason.
It sucks that you have to support a new platform (NT) just for one application, but as others have said, the only real alternative is to switch vendors.
--
Yup -- and if consumer tape is $2/Gig now, imagine what it will be in a couple years.
Right now the computer industry is sitting on it's hands and waiting collectively for DVD-R/W to stabilize and a compatible version of DVD-R to be released. But while the DVD people are dinking around, they won't wait forever.
I can imagine a situation in a year or two where Iomega, Creative and others in the industry get sick of waiting for writeable DVD and just release their own ~20 Gig disc format. Base it on existing CD-R technology, call it "SuperCDR", and push it out to the consumer at a reasonable cost. It would be designed for computer storage and backup, but of course nothing would stop you from putting VOB files on it. The only thing it couldn't do is play DVDs, but by then an IDE DVD player will probably be about $35
--
At this point, Intel and AMD are competing so hard on speed and price that neither can afford to undermark.
Sure competition between the two is tough, but Intel and AMD aren't stupid either. If they wanted to, they could blow their load and manufacture 1200Mhz chips tomorrow, and still make money.
(Why? Because I purchased a 400Mhz PII in 1998, I probably wouldn't upgrade to any new CPU this year. However next year, 1200Mhz might seem worth it. Businesses think this way, systematically replacing desktops every 18-36 months, and buying whatever middle shelf system is available at the time.)
They get far more revenue by stepping up the speed in a slow and predictable manner. The competition has just forced them to increase their schedules slightly. However, if they increase speeds too much, the buyers of midline chips will suddenly get far more price sensitive, which leads to lower profits. (This has already started happening with folks who decide a Celeron/K6 is 'good enough'.)
The key here from a $ standpoint is the sweet spot at the middle of the market (currently 500-600Mhz), not the dick-sizing going on at the top end.
--
The number of people with overclocked Celerons is miniscule compared to the number of non-clocked Celerons humming away on business desktops.
Both Intel and AMD are well aware of overclocking, but they will continue market towards the general case (business desktops) and not the specific (screwdriver guys). The overclockability of a specific chip will continue to depend on the fab output and the overall design of the chip. The loss of sales from overclocking is probably insignificant -- in fact Intel probably sold quite a few Celerons to overclockers that otherwise wouldn't have upgraded their existing systems.
Basically the Celeron was just a nice convergance of good fab capabilities (Intel tries to keep the PII numbers on schedule), and marketing conditions (businesses demand cheap desktops). It could happen again, but then again it could not.
--
Mod chips are feasible for the Playstation because there are a billion identical PSX units out there. While you could probably mod a DVD player, selling the kit would be hard because you probably would need a unique kit for every different make and model of player.
I could see DVD mod kits, but they would probably be expensive and marketed towards serious piracy operations.
It's probably a moot point anyway, because in the next couple years we'll have inexpensive computer backup media (removable disk, fast tape, etc) that can easily handle 10 GB of data. This will allow individuals to engage in small time piracy with other computer users, just not the supposedly with more lucrative a/v component DVD player market.
--
Anyway, manufacturers are more interested in our marketing, sales numbers, and a product fulfilling a need.
So, basically your arguments about piracy are moot. If the (Linux,OS2,Be,BSD,Solaris,MSDOS) markets were big enough to be profitable, people would sell players.
And while you are correct that the percentage of Linux users saavy enough to use DeCSS is higher than the percentage of Windows users, the saavy-enough Windows users far outnumber the saavy-enough Linux users. If piracy was a serious consideration, there would be no software players at all, or the computer DVD setups would be self-contained sealed hardware boxs.
--
Actually, the root problem is that the menuing/control system on DVD-video is just plain braindead and broken. Even obvious things like a pointer/mouse interface were left out.
It's clear that DVD was designed only with playback on an A/V component in mind. They ignored 15 years of research and practice in computer-based multimedia systems and went for a cheesy postcard interface. (The hardware component focus also explains the weak encryption design -- it looks as if they were more worried about people with sodering irons than the computer programming community.)
In hindsight it's obvious that a good portion of their installed base would be DVD-equipped computer systems. The user base is just that much more technically saavy and open to early adoption. Unfortunately, that user base is saddled with a UI designed for people who can't stop their VCR clocks from blinking. What's worse is they ignored the long running predictions of television-computer 'convergence' just when it's finally starting to happen.
--
The main reason that binary-only is bad on Linux is that the driver interface is completely unstable. Any particular binary driver works only with a single kernel revision, and is obsolete practically as soon as it's released.
This is a political question -- Linus won't create a stable driver interface precisely to force driver vendors to release the source. If you donate the source, maintenance is freely performed by the Linux community. If you don't, it's a support nightmare.
There's probably quite a few Linux users who don't like this approach, but ultimately, it's Linus' call.
--
Hmmm, I see nothing in the GPL which implies that the "You" is a person and not a company or organisation. If the NSA is the licencee of GPL software, it would seem that they as an organization are responsible for following the terms. Letting employees use the software doesn't seem like "distribution" (at least not in the normal sense used in software licences.)
But even if you are correct, most employees have employment agreements dictating what they can do with intellectual property owned by the company. So while they may have the right to redistribute under the GPL, they probaby don't have that right as an employee. With the NSA, such employment agreements have the force of law behind them -- making any internal NSA practice or product public probably carries a very long prison term (even when doing so might be legal in the private sector).
--
My take:
1) Office Suite integration -- The only thing innovative here was the marketing. Buy a word proc and a spreadsheet and get a slideshow program for free and a desktop database at a reduced price. Plus the cost of MS Office was far below the combined costs of WordPerfect, Lotus 123, and DBase.
2) Linking and Embedding -- An early Mac program, Lotus Jazz had something like this. I think VisiOn did something similar under DOS. What was innovative was making it a universal operating system service -- see #3
3) COM -- Not a new idea, but since MS owns the OS, the desktop apps, the server apps, and the devtools, the integration buys users a lot. Almost no body runs any sort of Corba services on their desktops, but every Windows user uses COM.
4) Internet Ready -- Yawn. If IE didn't exist, they would ship Netscape just like anyone else. Microsoft's 'innovations' in this area, like "Channels" have been roundly rejected.
5) VBA -- The only really interesting thing about VBA is that you can rip the VB part out and use JavaScript or PerlScript. I think one would have quite a bit of trouble arguing that MS stuff is more "scriptable" than virtually anything else.
--
Agreed -- Gates has famously said that they are "betting the company" on Windows 2000, and to some extent they are right ... If W2000 succeeds as they'd like, Microsoft will become as important as IBM or Sun in enterprise computing; if it fails, Microsoft stays in business but becomes just another Novell hanging onto 'departmental' boxes. Having Gates on the sales team can make a huge difference, especially considering that the upgrade costs are stratopheric enough to require executive level approval in most shops.
(On a side note, if anyone should be fired over Win2000, it should be the people who speced out the product long before engineering had begun. Features like ActiveDirectory have been promised RSN for about 5 years. Microsoft's FUD machine got way way out in front of it's engineering ability. The fact that they've delivered a product that appears to meet marketing's fantasy land specifications is pretty amazing. I know quite a few of you slashdotters have opioned that they wouldn't be able to do it. Now the question is, can they sell it?)
--
You are correct that *network browsing* (via the Network Neighborhood or NET VIEW) doesn't work over a routed network without a master browser/WINS setup. However, with LanMan networking, don't confusing browsing with actual accessibility.
--
Well for 10 years we've been hearing how the NeXT development environment is one of the best ever invented. If true, why would Apple dump what they've got for a third party tool? (Does anyone know if NeXT IDE comes for free with OS X? I'd be a dumb move if it wasn't.)
/. poster that has bitched a few times about MS/IBM/Borland/Novell politics in the early days of OS/2. Things like that can cast a cloud over a new platform that it might never recover from.)
I'm sure they consider Metroworks very important for Carbon projects, but the momentum is clear -- new projects should be Cocoa, which means Objective C or Java which means not Metrowerks.
From Apple's point of view, it frees them from yet another Third Party Developer that can crap all over whatever their strategy is. When Symantac started cutting back on Mac development, things got shaky. Having Code Warrior appear was pure luck for Apple, and I don't think they want to go through that again. (As another reference, there's a
--
"Keep all your versions up to date and you can conquer Microsoft."
True -- most WordPerfect and SmartSuite shops I've seen have been at least one or two versions behind. Consequentally the users are bitching about the lack of functionality and import/export filters. (It's hard to import Office 97 files when you are running SmartSuite 96)
Usually this is intentional IT politics. They want to switch, but can't justify the cost. So they let the functionality slide a for a few years, and wait for everyone to start begging for MS Office.
--
Look at kernel32.dll, user32.dll, gdi32.dll, krnl386.exe, user.exe and gdi.exe sometime in a hex editor. It's scary how many "exported" function calls are there that aren't documented by ANYONE.
Those file, as a whole, are the Win 9x kernel. Microsoft, for obvious reasons, doesn't want to document internal kernel interfaces.
Look at the Linux situation -- Linus doesn't document internal kernal interfaces, and reserves the right to change them. If you write a closed source driver that uses an internal interface, it will eventually break, and that's your problem. MS doesn't have the luxury to treat closed source developers that way.
If you don't like internal, undocumented interfaces, don't use closed source software. It's as simple as that, and has nothing to do with MS as a monopoly.
Your points about MSSQL leads to Scatter/Gather IO (and so on) hold, but I doubt it's above and beyond what any other OS vendor does. For example, Linux provided interfaces to deal with Apache performance problems specifically after the Mindcraft benchmarks were published.
--
AT&T break up was done the way you say. The goverment turned one big monopoly into several little ones, that eventually became big again.
A regional breakup of the phone company made sense, because by-in-large their assets are wires, switches, and franchises -- all fixed in a particular spot. Now with local compeitition plans, some can get different phone service, but the facilities (and most of the operation costs) are still owned by the RBOC. You are essentially just paying someone else for the billing service.
Likewise, with MS -- the 'natural' breakup is along operational lines. However, you are correct that it doesn't have to be this way, because MS's property is primarily intellectual and unlike phone facilities can be transfered for little cost. But, that doesn't mean it's the right way to do it.
To recap, If we break up horizontally, then one of the OS/apps can become the dominant one and we will be in trouble again.
Right -- we would be in the exact same situation as we were with MS-DOS versus IBM DOS versus DR-DOS. The minor incompatibilities weren't worth the hassle of switching vendors (ignoring the Win3.1 beta warning about DR-DOS for the moment.)
Windows NT is not the be-all-end-all and neither is MS Office. (If it were, we wouldn't be yacking on a pro-Linux/OSS board.) Let MS-OS and MS-Apps have their little monopolies -- "Microsoft" loses their single vendor status and any theoretical integration benefits. That alone is huge enough to allow competitors in the door.
--
I've heard it said a few times that some of the IBM/MS DOS 1.0 tools had the string "(C) Digital Research" embedded in them.
Considering QDOS was one guy's project to port CP/M to the 8086, we know that it wasn't "clean room" reverse engineering like Linux. (Microsoft, at that time a pretty small company, also had access to CP/M source, as they did the Apple II Softcard port.)
Gary Kildall thought he could have sued IBM & MS, but chose not to. It had nothing to do with look and feel. I don't know why Caldera didn't sue --perhaps the statue of limitations had run out, or they figured that the DOS 1.x market was so small that it wouldn't have been worth it. (DOS 2.x was a major rewrite.)
--
You got a few things turned around, pardner --
+ IBM approached Microsoft (primarily for BASIC), not the other way around. Microsoft was a CP/M licencee, and referred IBM to Digital Research.
+ Gary Kildall was not willing to meet with IBM, and basically blew them off.
+ MS zoomed in to cut a deal with QDOS.
+ A IBM PC buyer could choose either CP/M-86 or IBM/MS-DOS (or the UCSD Pascal OS). Only problem was CP/M cost $150, and IBM DOS was only $50.
--
OK, the alternative scenario is that everyone buys a new TV in the next 5 years, and we spend the rest of our days watching "Urkel" in full photorealistic resolution. The broadcasters happily give the analog spectrum back to the government. Then Heidi Klum falls in love with me and I win the lotto. It's going to be a glorious future!
--
The airwaves are owned by the public, so the theory is that the government is ensuring that they are put to wise use. Five different AM Stereo formats wasn't in anyone's best interest except the people trying to kill AM stereo.
Computer standards on the other hand are controlled by Intel and Microsoft, there is no implied public interest.
--
AM Stereo got shot down because the people who own the FM stations also own the larger AM stations. Allowing every religious and foreign language station to broadcast in 'hi-fi' would have killed the market value of the FM stations, so the station owners kept AM stereo in the bottle.
For a few years in the 1980s, American car companies shipped AM Stereo capable recievers. Most people didn't notice, however, because nobody tried to broadcast hifi pop music on AM.
--