yeah and then you want Be to not be a general purpose OS
Right, because the "general purpose OS" market is saturated. If they would have specialized (especially after the Apple thing fell through), they might be still chugging along.
Well, the BeBox, as executed, might have been a bad idea, but clearly they could have sold hardware at a much higher margin than they are currently selling software.
If Be could have focused on a single high-performance platform (say 4 PPC chips) and bundled some powerful multimedia applications, they could have sold these things into vertical markets for lots of dough.
(For example, look at the video and sound editing stations that are currently based on standard PC or Mac hardware, but sold at huge markups.)
Copland was another attempt, and probably could've worked out. But the designers wanted to get legacy Mac software running natively on it. I don't think that emulators were ever really considered.
IIRC, the main (marketing) problem with Copeland was that it was it would have broken virtually all applications and drivers, but still wouldn't have delivered a fully 'modern' OS with memory-protection and pre-emptive MT OS (that was planned for the next vapor release, Gershiwn). It was a sort of a Mac version of Windows 95, except without the Win16 mode.
There were also various management problems, such as much of the engineering teaming resigning, which left Apple BSing about how Copeland would ship RSN at the same time they were taking out full page ads in the Mercury News looking for lots of senior MacOS engineers.
I don't think Pink/Taligent got anywhere close to being an OS.
I think you are correct that BeOS was targetted specifically to be Apple's replacement OS. It's certainly closer in spirit to MacOS than NeXTStep was. However, if the stories are true, Be wanted too high of a price and there were some personal issues with Gassee (well, there were with Jobs too, but he's the legend there).
(OT) Using the W3C specification pages as developer documentation sucks. If only because it's not clear exactly what DOM/CSS functionality is implemented and what isn't in Mozilla. But also because it's missing code-examples, references, hyperlinks, compatibililty information and so on. I'd also suspect that there's some superset functionality in Mozilla.
Netscape had pretty good 4.x docs (see http://devedge.netscape.com/tech/javascript/index. html). Is it too much to ask that they actually document Netscape 6/Mozilla in a similar way? They do want people to develop for it, but they aren't making it easy.
There is a group of "People who play games", and basically run a system designed to play games (which probably means Win9x).
There is also another, much larger group of "People who play games on a machine that they do real work on". Usually that machine is Windows, but sometimes it's MacOS or Linux or something else. Because these people are doing real work, they aren't that interested in duelbooting or multiple machines.
Basically for a game to be profitable, it has to convice the first 'hardcore' group. But it also has to sell broadly into the second group. Which is why it wasn't super-critical that Quake III was on the shelves for Linux on Day 1, because the assumption should be that anyone who would run Q3 on Linux is doing so because they want to run Linux, not because they want to run Q3.
I used to be in a similar boat as the Linux folks of group 2 running NT 4.0 on my work/game box. It meant that I was limited to a subset of NT4-compatible games (OGL and DX3 types usually), and it also meant I had to sit back and wait for compatibility reports before running out and purchasing a new game.
Sad thing is that Exchange 6 "Platinum" was originally supposed to ship in 1998. It was rushed to market, and 2 years late, and had a big feature cut after the last beta, all at the same time
Of course that's nothing new for Exchange - the original version was something like 6 years late (was briefly beta'ed for OS/2 in the late 80s.)
I dunno, I worked at a place that had dozens of sales laptops running NT4 Workstation, which is pure non-PnP pain BTW if you've never tried it.
I'm sure that if Linux gets big on the desktop, you'll see manufacturer-supported Linux distros running better than that on laptops.
The bigger problem was the Windows-only sales force automation package that was in use. That would need to be rewritten from scratch at a cost that could pay for a thousand Linux-installing techs (that, or Wine would need to get a lot better).
Furthermore, someone signing a contract for an organization has to be a legally recognized represenative for that organization.
The MCSE installing NT does not have the legal authority to "Press F8 to agree to this licence", and you can't exactly call corporate council down to the server room to push a button.
Which is why, if you are on a site licence program such as Select, Microsoft brings out a paper contract that is signed the old fashioned way and superceeds any click-thrus. You do have some negotiating room on this contract.
These reports are for sale so that IT managers can justify what they already want to do to their own superiors. It's part of the political decision making process, not the strategic process.
For example, if an IT Manager wants to replace his NT webservers with Linux, he can either write a "I think we should do this" memo to get the funding, or he can quote a "Gartner Group Report" that shows that Linux has a.75 probability of replacing NT webservers. Even though it's widely known that Gartner is a bunch of whores that will tell you anything you pay them to say, having a 'higher authority' is still better justification than your personal opinion.
Apple in the old days had a horrible attitude towards corporate deployments. You wanted your Mac clients to talk to any servers? Had to be through AppleTalk. You wanted database integration for HyperCard or FileMaker so you can support Macs on a client-server app? Too bad.
Meanwhile, Microsoft was actually listening to corporate customers and developing things like VB that were actually of some use to them. Apple had a huge corporate sales force, but never produced any tech that was actually of any help to a corporation.
This was all back in the late 80s/early 90s, so today YMMV. Of course, except for the graphics guys, Apple has been wiped out of the corporate market, so it doesn't really matter.
Yeah, you hit the nail on the head there. Piracy has always been Microsoft's best marketing strategy, but there's nobody left to market against for their core products.
You can also see the attitude change between Gates and Ballmer. Gates, since the doomed hobbiest letter, hasn't ever really sweated if someone somewhere was ripping him off, as long as he knew he'd eventually get paid. On the other hand, rampant MS piracy probably keeps Ballmer awake at night.
To clear that up, there's a general rule at play here. If anything you've got running on NT/2K uses the built-in security services, you need to buy either NT client seats (for every client, not just similtanious connections) or expensive per-processor licences.
That means, for example, you can use IIS in normal HTTP authenticaion mode just fine, but if you use NTLM directory authenticiation, you have to purchase NT clients for everyone that might connect.
Another example is MS SQL server. You can get away with only buying SQL server seats and not NT seats if you use the plaintext SQL authentication. Turn on "integrated" authentication and you are on the hook for NT/2K client seats as well.
and a squad of MS consultants who all have completely differing opinions on what requires a license and what does not
That's your problem. Microsoft Consulting is a break-even operation that is effectively part of Microsoft's sales organization. Their opinion on any topic is that your project should use as many MS products as possible, and they'll engineer it that way, and then report the results to sales so that MS can collect after you've deployed. Worse, you are paying them for that service. There's consulting shops out there that are up-to-speed on MS tech, but not in their backpocket. Find them and use them.
(A company I know of made the mistake of letting MS Consulting plan out a commercial product. Ended up based on MS Site Server Commerce, MS Office 2000, and MS-SQL. Turns out the Microsoft licence costs were something like 3 times the company's licence costs, making the product totally non-competitive.)
More typical service organizations (such as that of Lotus/IBM) also want to get their product deployed, but because they are a profit center they at least have some independant 'what's best for the customer' thinking going on.
I should have mentioned that Microsoft obviously favors #1, and will make #2 difficult for you.
It should be a cost/benefit analysis -- if you can't afford the lawyers and the accountants, don't select option #2. Businesses make these decisions all the time, chosing to pay out one large sum of money for low risk in favor of many small sums of money with unknown risk.
One of the worst mistakes is to put the techies in charge of licence compliance (because they usually have a totally lax attitude towards such things, and they are not exactly organizational geniuses).
I lived through a MS audit a few years ago with the kinder, gentler Microsoft. We had our shit in order and had bought certain selective site licences (such as for Office), so it was no problem.
There's been a significant change in Microsoft's attitude. It used to be pretty much "Developers get free comps". And "Deploy Now and we'll worry about the licences later".
This has helped them quite a bit, because (say) a department level project can start today with MS technology without waiting for the endless IT approval process for space in the glassbox Unix farm.
I guess if you are against Microsoft being successful, this is one move you have to cheer. No longer will MIS departments be lax about the Microsoft cheer spreading around if it means they have to be distracted with audits and licence compliance tasks.
Combined with the recent price-hikes of MS stuff, and the inevitable recession rollback of IT budgets, the corporate Microsoft platform free-for-all era is coming to an end. In a few years, it will be just as hard to get MIS to approve your new MS-SQL server as it is today to get an Oracle database approved.
Look, it doesn't take a genius to figure out how to handle Microsoft licencing. You basically have two options:
1) Buy site licences for all of your Microsoft stuff, even if this means paying for certain copies of Windows twice. At least you know you are legit.
2) Keep track of all your individual licences. Cheaper, but higher administrative overhead.
Now, we have these 'legitimate' companies who have chosen a third option: "Don't buy a site licence and don't keep track of licences and don't do any internal auditing". Which means they get to save all sorts of money and hassle in the short term, but are essentially bending over like the goatse guy to pick up the soap.
Obviously it's poor business practice for MS to go around and harrass people that are legit or trying to be legit. The problem here is organizations where the IT department turns a blind eye to what's going on. Look what happened at Virginia Beach -- They were all smug with their 'standard' of WordPerfect, but they knew full well that everyone out there was running MS Office (sometimes paid for, sometimes not).
So, the moral is that stupid companies get screwed. Today it's Microsoft, tomorrow it will be Corel, the next day Oracle.
He's sort of correct actually. Apple developed an internal version of BASIC that had somthing like the hypercard/VB-style of graphical screen building.
Microsoft was already selling a (very buggy) version of their procedural BASIC for the Mac, and wasn't happy about this development so the Apple Basic project was kiboshed and work on HyperCard was started.
Somewhere around I have an ancient MacWorld that talks about using HyperCard as a frontend to a DEC database system. Unfortunately, Apple never picked up on this and never shipped easy-to-integrate database drivers for Hypercard, as MS did for VB.
Speaking of Flash, it's only a matter of time before it starts getting used as a DB front-end itself.
Apparently, XP allows you to "disconnect" from your login session (kinda like Terminal Server does), and login as another user. Presumably your session state is then swapped out to disk. Similar to using virtual terminals on Linux, I imagine.
Agreed that "N programs running" is a bit of information leakage - hopefully you can disable that.
NT has always allowed unprivledged shutdown (of the machine) from the login screen. Administrator change enable/disable this from the registry. I don't know where'd you get the idea that nonprivledge users can kill other user's processes - NT has never worked that way.
You are right on in your description of how this all works (except for the kernel panic part...).
But, I think what you miss is that all of this stuff is designed for is a future class of consumer applicance products that are based on standard PC-type parts. Think TiVo, think digital music deck, all with little IDE drives and proprietary OSes that can be designed around the encryption problems.
What's going on here is that the PC parts companies are seeing their markets stagnate and are making a bid for a piece of the gigantic consumer electronic market. That means changing how they do business, in particular their attitudes about who owns the data. Consumer electronic companies love the idea of building things out of standard parts, so they are more than encouraging the move.
I don't think anyone is seriously suggesting that this stuff be implemented in personal computers and personal computer OSes, at least not yet. There's still too much "personal" in the PC market, and people will just flat not go for any system that removes their control of their data. (And, yes, I know about Windows' secure channel stuff -- different story because it's not hardware lock-in, it's for optional software use only).
But still, I agree that the possibility that one might not have control over the bits bouncing around inside their computer cases is pretty scary. Hollywood has already bucked this tradition mostly successfully with software DVD player (for some reason every company has enlisted for the DVD's forums licencing conditions instead of reverse engineering a clone), and they are going for more pieces with secure music channels and the like. However, the time is not right for a general hardware solution in PC space, nor will it probably ever be.
Actually, I think Matrox would like to release full sources, but they have a contract with the Macrovision folks which prevents them from doing so. (One just needs to flip a bit to disable Macrovision. Windows hacks are all over, so there's probably Linux documentation on this procedure too.)
Slashdot Mirror
yeah and then you want Be to not be a general purpose OS
Right, because the "general purpose OS" market is saturated. If they would have specialized (especially after the Apple thing fell through), they might be still chugging along.
Hopefully Apple will snatch up Gassee and put him personally in charge of Job's iMac-shaped shitbucket.
Well, the BeBox, as executed, might have been a bad idea, but clearly they could have sold hardware at a much higher margin than they are currently selling software.
If Be could have focused on a single high-performance platform (say 4 PPC chips) and bundled some powerful multimedia applications, they could have sold these things into vertical markets for lots of dough.
(For example, look at the video and sound editing stations that are currently based on standard PC or Mac hardware, but sold at huge markups.)
Copland was another attempt, and probably could've worked out. But the designers wanted to get legacy Mac software running natively on it. I don't think that emulators were ever really considered.
IIRC, the main (marketing) problem with Copeland was that it was it would have broken virtually all applications and drivers, but still wouldn't have delivered a fully 'modern' OS with memory-protection and pre-emptive MT OS (that was planned for the next vapor release, Gershiwn). It was a sort of a Mac version of Windows 95, except without the Win16 mode.
There were also various management problems, such as much of the engineering teaming resigning, which left Apple BSing about how Copeland would ship RSN at the same time they were taking out full page ads in the Mercury News looking for lots of senior MacOS engineers.
I don't think Pink/Taligent got anywhere close to being an OS.
I think you are correct that BeOS was targetted specifically to be Apple's replacement OS. It's certainly closer in spirit to MacOS than NeXTStep was. However, if the stories are true, Be wanted too high of a price and there were some personal issues with Gassee (well, there were with Jobs too, but he's the legend there).
(OT) Using the W3C specification pages as developer documentation sucks. If only because it's not clear exactly what DOM/CSS functionality is implemented and what isn't in Mozilla. But also because it's missing code-examples, references, hyperlinks, compatibililty information and so on. I'd also suspect that there's some superset functionality in Mozilla.
. html). Is it too much to ask that they actually document Netscape 6/Mozilla in a similar way? They do want people to develop for it, but they aren't making it easy.
Netscape had pretty good 4.x docs (see http://devedge.netscape.com/tech/javascript/index
Boy do you have it wrong.
There is a group of "People who play games", and basically run a system designed to play games (which probably means Win9x).
There is also another, much larger group of "People who play games on a machine that they do real work on". Usually that machine is Windows, but sometimes it's MacOS or Linux or something else. Because these people are doing real work, they aren't that interested in duelbooting or multiple machines.
Basically for a game to be profitable, it has to convice the first 'hardcore' group. But it also has to sell broadly into the second group. Which is why it wasn't super-critical that Quake III was on the shelves for Linux on Day 1, because the assumption should be that anyone who would run Q3 on Linux is doing so because they want to run Linux, not because they want to run Q3.
I used to be in a similar boat as the Linux folks of group 2 running NT 4.0 on my work/game box. It meant that I was limited to a subset of NT4-compatible games (OGL and DX3 types usually), and it also meant I had to sit back and wait for compatibility reports before running out and purchasing a new game.
Exactly. Which is why the "hackers" quest to get the news media to start using the word "crackers" is ultimately quixotic.
No reputable national publication will use the word "crackers" in a headline, no matter what you think it means.
Sad thing is that Exchange 6 "Platinum" was originally supposed to ship in 1998. It was rushed to market, and 2 years late, and had a big feature cut after the last beta, all at the same time
Of course that's nothing new for Exchange - the original version was something like 6 years late (was briefly beta'ed for OS/2 in the late 80s.)
Copying software from media to memory is explicitly "Fair Use" under US copyright law. Therefore, you can do it without consenting to the GPL.
I dunno, I worked at a place that had dozens of sales laptops running NT4 Workstation, which is pure non-PnP pain BTW if you've never tried it.
I'm sure that if Linux gets big on the desktop, you'll see manufacturer-supported Linux distros running better than that on laptops.
The bigger problem was the Windows-only sales force automation package that was in use. That would need to be rewritten from scratch at a cost that could pay for a thousand Linux-installing techs (that, or Wine would need to get a lot better).
Furthermore, someone signing a contract for an organization has to be a legally recognized represenative for that organization.
The MCSE installing NT does not have the legal authority to "Press F8 to agree to this licence", and you can't exactly call corporate council down to the server room to push a button.
Which is why, if you are on a site licence program such as Select, Microsoft brings out a paper contract that is signed the old fashioned way and superceeds any click-thrus. You do have some negotiating room on this contract.
The company in question was operating out of the Caribbean island of Nevis. You add it up.
For every (say) $200 seat licence for the product, the customer owed $600 in seat licences to Microsoft. Similar story on the server side.
That's a simple way of looking at these services.
.75 probability of replacing NT webservers. Even though it's widely known that Gartner is a bunch of whores that will tell you anything you pay them to say, having a 'higher authority' is still better justification than your personal opinion.
These reports are for sale so that IT managers can justify what they already want to do to their own superiors. It's part of the political decision making process, not the strategic process.
For example, if an IT Manager wants to replace his NT webservers with Linux, he can either write a "I think we should do this" memo to get the funding, or he can quote a "Gartner Group Report" that shows that Linux has a
Apple in the old days had a horrible attitude towards corporate deployments. You wanted your Mac clients to talk to any servers? Had to be through AppleTalk. You wanted database integration for HyperCard or FileMaker so you can support Macs on a client-server app? Too bad.
Meanwhile, Microsoft was actually listening to corporate customers and developing things like VB that were actually of some use to them. Apple had a huge corporate sales force, but never produced any tech that was actually of any help to a corporation.
This was all back in the late 80s/early 90s, so today YMMV. Of course, except for the graphics guys, Apple has been wiped out of the corporate market, so it doesn't really matter.
Yeah, you hit the nail on the head there. Piracy has always been Microsoft's best marketing strategy, but there's nobody left to market against for their core products.
You can also see the attitude change between Gates and Ballmer. Gates, since the doomed hobbiest letter, hasn't ever really sweated if someone somewhere was ripping him off, as long as he knew he'd eventually get paid. On the other hand, rampant MS piracy probably keeps Ballmer awake at night.
To clear that up, there's a general rule at play here. If anything you've got running on NT/2K uses the built-in security services, you need to buy either NT client seats (for every client, not just similtanious connections) or expensive per-processor licences.
That means, for example, you can use IIS in normal HTTP authenticaion mode just fine, but if you use NTLM directory authenticiation, you have to purchase NT clients for everyone that might connect.
Another example is MS SQL server. You can get away with only buying SQL server seats and not NT seats if you use the plaintext SQL authentication. Turn on "integrated" authentication and you are on the hook for NT/2K client seats as well.
and a squad of MS consultants who all have completely differing opinions on what requires a license and what does not
That's your problem. Microsoft Consulting is a break-even operation that is effectively part of Microsoft's sales organization. Their opinion on any topic is that your project should use as many MS products as possible, and they'll engineer it that way, and then report the results to sales so that MS can collect after you've deployed. Worse, you are paying them for that service. There's consulting shops out there that are up-to-speed on MS tech, but not in their backpocket. Find them and use them.
(A company I know of made the mistake of letting MS Consulting plan out a commercial product. Ended up based on MS Site Server Commerce, MS Office 2000, and MS-SQL. Turns out the Microsoft licence costs were something like 3 times the company's licence costs, making the product totally non-competitive.)
More typical service organizations (such as that of Lotus/IBM) also want to get their product deployed, but because they are a profit center they at least have some independant 'what's best for the customer' thinking going on.
I should have mentioned that Microsoft obviously favors #1, and will make #2 difficult for you.
It should be a cost/benefit analysis -- if you can't afford the lawyers and the accountants, don't select option #2. Businesses make these decisions all the time, chosing to pay out one large sum of money for low risk in favor of many small sums of money with unknown risk.
One of the worst mistakes is to put the techies in charge of licence compliance (because they usually have a totally lax attitude towards such things, and they are not exactly organizational geniuses).
I lived through a MS audit a few years ago with the kinder, gentler Microsoft. We had our shit in order and had bought certain selective site licences (such as for Office), so it was no problem.
There's been a significant change in Microsoft's attitude. It used to be pretty much "Developers get free comps". And "Deploy Now and we'll worry about the licences later".
This has helped them quite a bit, because (say) a department level project can start today with MS technology without waiting for the endless IT approval process for space in the glassbox Unix farm.
I guess if you are against Microsoft being successful, this is one move you have to cheer. No longer will MIS departments be lax about the Microsoft cheer spreading around if it means they have to be distracted with audits and licence compliance tasks.
Combined with the recent price-hikes of MS stuff, and the inevitable recession rollback of IT budgets, the corporate Microsoft platform free-for-all era is coming to an end. In a few years, it will be just as hard to get MIS to approve your new MS-SQL server as it is today to get an Oracle database approved.
Look, it doesn't take a genius to figure out how to handle Microsoft licencing. You basically have two options:
1) Buy site licences for all of your Microsoft stuff, even if this means paying for certain copies of Windows twice. At least you know you are legit.
2) Keep track of all your individual licences. Cheaper, but higher administrative overhead.
Now, we have these 'legitimate' companies who have chosen a third option: "Don't buy a site licence and don't keep track of licences and don't do any internal auditing". Which means they get to save all sorts of money and hassle in the short term, but are essentially bending over like the goatse guy to pick up the soap.
Obviously it's poor business practice for MS to go around and harrass people that are legit or trying to be legit. The problem here is organizations where the IT department turns a blind eye to what's going on. Look what happened at Virginia Beach -- They were all smug with their 'standard' of WordPerfect, but they knew full well that everyone out there was running MS Office (sometimes paid for, sometimes not).
So, the moral is that stupid companies get screwed. Today it's Microsoft, tomorrow it will be Corel, the next day Oracle.
He's sort of correct actually. Apple developed an internal version of BASIC that had somthing like the hypercard/VB-style of graphical screen building.
Microsoft was already selling a (very buggy) version of their procedural BASIC for the Mac, and wasn't happy about this development so the Apple Basic project was kiboshed and work on HyperCard was started.
Somewhere around I have an ancient MacWorld that talks about using HyperCard as a frontend to a DEC database system. Unfortunately, Apple never picked up on this and never shipped easy-to-integrate database drivers for Hypercard, as MS did for VB.
Speaking of Flash, it's only a matter of time before it starts getting used as a DB front-end itself.
Apparently, XP allows you to "disconnect" from your login session (kinda like Terminal Server does), and login as another user. Presumably your session state is then swapped out to disk. Similar to using virtual terminals on Linux, I imagine.
Agreed that "N programs running" is a bit of information leakage - hopefully you can disable that.
NT has always allowed unprivledged shutdown (of the machine) from the login screen. Administrator change enable/disable this from the registry. I don't know where'd you get the idea that nonprivledge users can kill other user's processes - NT has never worked that way.
You are right on in your description of how this all works (except for the kernel panic part...).
But, I think what you miss is that all of this stuff is designed for is a future class of consumer applicance products that are based on standard PC-type parts. Think TiVo, think digital music deck, all with little IDE drives and proprietary OSes that can be designed around the encryption problems.
What's going on here is that the PC parts companies are seeing their markets stagnate and are making a bid for a piece of the gigantic consumer electronic market. That means changing how they do business, in particular their attitudes about who owns the data. Consumer electronic companies love the idea of building things out of standard parts, so they are more than encouraging the move.
I don't think anyone is seriously suggesting that this stuff be implemented in personal computers and personal computer OSes, at least not yet. There's still too much "personal" in the PC market, and people will just flat not go for any system that removes their control of their data. (And, yes, I know about Windows' secure channel stuff -- different story because it's not hardware lock-in, it's for optional software use only).
But still, I agree that the possibility that one might not have control over the bits bouncing around inside their computer cases is pretty scary. Hollywood has already bucked this tradition mostly successfully with software DVD player (for some reason every company has enlisted for the DVD's forums licencing conditions instead of reverse engineering a clone), and they are going for more pieces with secure music channels and the like. However, the time is not right for a general hardware solution in PC space, nor will it probably ever be.
Actually, I think Matrox would like to release full sources, but they have a contract with the Macrovision folks which prevents them from doing so. (One just needs to flip a bit to disable Macrovision. Windows hacks are all over, so there's probably Linux documentation on this procedure too.)