What is needed are metrics for estimating how hard the problem is, not how hard someone worked to solve it. For estimating the cost of writing a program in the first place, there are various measurements of the problem complexity. One is function points. Google found about 10 pages of links. Here is a FAQ (not approved by the function point users group) that discusses the use of function points to measure productivity, among other things.
Why can't everything else go in a specialized room somewhere
Because then you need extremely high bandwidth connections from that room to each workstation. If the processing that generates your video is done remotely, you need more than a 100MHz ethernet connection to get it to the screen without taking a highly noticeable performance hit as compared to a PC that calculates the picture internally. So out in the real world, where high bandwidth costs more than silicon, you need a powerful CPU and quite a lot of RAM in the workstation.
The answer is less clear-cut with the hard drive. Diskless PC's, where the programs and data all must be downloaded through the network before the PC starts crunching on the numbers, clog the network and performs very poorly. If the software was rearranged so the program ran in the server room, and the PC simply expands the results into display screens, then you could get decent performance. But the way this would usually be implemented would be with one dual or quad CPU box serving dozens of people -- and the performance would bottleneck there. Put enough CPU's in the server room, and find software that works in this "thin-client" mode, and you've got a good system for medium and big business. It costs a little more because you've got (nearly) two CPU's and two sets of RAM per user, but it's quite possible to save more on maintenance than it costs in hardware, if you can find the software.
However, you'll probably discover that most people need at least one program that wasn't written for thin clients, and will not perform adequately if either the disk access or the video goes over the network. So you wind up back to a full PC on every desktop, plus all those extra CPU's in the server room, and even more maintenance...
I think that the CD has to be quite flat for the laser to stay focused. So I assume that their adapter sandwiches the flex disk between two sheets of plastic. (This should also settle the questions about weight and balance -- the adapter is going to be the same thickness and very nearly same weight as a normal CD.) But get a little dust on it, and the disk is no longer flat inside the sandwich...
I once had a somewhat sleazy dealer "forget" to put a key clause into the written contract on a mobile home, but I read it carefully and called him on it before signing. It did get written in. I suspect that he'd actually made the price a bit more attractive than he could really afford, expecting to make it back and quite a lot more on that one change. So he scr***d himself.
IIRC, X-windows preceded MS Windows, and in any case, computer journals used "window" to describe a graphical area on the screen back when Xerox and Apple were the only places working on computers with them. That is, MS tried to grab a generic industry term and turn it into a trademark. The trademark "Zipper" was once voided because, even though it was a new word invented by the manufacturer and sole patent licensee of the teethed fasteners, it became a generic word in common use. For MS to claim ownership of "windows" in regards to computer software is to ask the courts to more than reverse the zipper precedent.
And if MS somehow wins on that, to claim infringement they also have to claim that Lindows as marketed is likely to be mistaken for their own product. (Like a "Bolex" watch.) That shouldn't fly either, because Lindows whole marketing pitch is that it _isn't_ Windows.
No, the _customer_ doesn't do configuration tweaking in black boxes, the _developer_ does. And expect to do lots of tweaking, because for cost reduction, you probably aren't using standard hardware.
Sorry, even Win98 had this beat. If you left it set up at defaults, it is:
1. Insert the CD. An install window opens automatically. 2. Click Yes.
Most of the time it works. What MS hates to discuss it that when it doesn't work, you are likely to be really f*d up. And it leaves too many openings for malicious or just badly written software to install things you didn't want. I assume the Mac has similar vulnerabilities, but since Apple maintains much tighter control over software for the Mac, the chances of an installation going bad is lower.
I agree with you, except for one thing: "Special-purpose machines... Microsoft is set to own this market." Microsoft is _trying_ to grab as much of this market as they can, but they do not and will not own it. The most prominent example of special purpose machines today is PDA's -- and the Palm OS is doing quite well in competition with Windows CE.
But you want to build a special purpose word processing machine. So, do you go spend $$$ on Windows CE and the Word component of MS Office, or do you download Linux, KDE or Gnome, and Star Office? Either way, you put lots of work into figuring out how to configure the software to work on your particular hardware. But with Linux, once it works, you can clone that setup indefinitely, for free. With Windows, you pay by the copy, after paying initially to get the development system. With Linux, you'll have to work harder to make it luser-friendly and hide all the system complexity. With Windows, MS has hidden much of the system complexity even from experts -- if it happens to all work right, that's great, but if it doesn't work, you'll have a hard time getting the info to fix it. And you cannot modify the code, or look at it to figure out why things are going wrong -- not that you want to do this in Linux either, but if things really go wrong it might salvage the project.
And finally, with Windows, there is always the risk that Bill Gates will decide he wants to buy Brazil or something, so next year your license costs jump from $50 to $200. It might make it pretty hard to compete with those $300 Linux boxes, but if you've built your whole business around Windows CE, you might not have a choice. Maybe you'd better co-develop an OSS implementation, just in case.
OTOH, when you market the system, you can piggyback onto lots and lots of MS advertising claiming that Windows is the biggest innovation since the wheel. Unless you try to sell it somewhere that lying advertisements are prosecuted as fraud, or to people that are clueful.
Windows will be a big player in this market, but it cannot dominate it. The basic problem is that MS's tactic of changing data file formats to force upgrades is beginning to wear thin even in conventional PC's; if they start telling people that they have to throw out perfectly good hardware because it doesn't support Word 2005, many will shift to something else instead. And if they stick to existing standards, they are going to wind up competing with _free_ without any really superior features. MS marketing is actually good enough to win this -- part of the time.
I interpret that question as "is it possible to cool by convection driven by the temperature differentials." Yes, provided:
1. No extreme hot spots -- like Pentiums. They are too small and make too much heat, so I don't think passive air cooling alone can handle it. Scale back your MHz, or add a liquid-filled heat spreader.
2. The case is designed for air flow, the cables are tied out of the way, and the customer can't get inside and f* it up.
3. The case a tower, and it is always operated right side up with the covers on and the vents clear. Techs have to know how long they can keep the power on while the box is open to work on it. A customer that lays something on the vent, or puts his tower on it's side is going to have trouble.
Yes, it is certainly possible to cool fairly high-powered electronic devices passively, but this device wouldn't be used the way PC's are.
the bearings and oil seals are evenly dispersed over the surface area of the heat sink
Huh??? The site got slashdotted before I could download the pdf or anything with a decent picture, but where do you get this? It sure looks to me like the bearing is still on a shaft up the center. It should help that it's a thin shaft with air moving around it, rather than the bearings being stuck up into a hot electric motor hanging above the hot center of the chip.
Are you thinking that the bearing is all around the rim? That would make for a lot of bearing surface with one surface moving at extremely high speeds. If that's where they put the bearing, no wonder the fans cost $100 -- but I still don't think it would be very long lived, because at the high RPM's times the radius, it racks up the miles too fast.
That "explosive" trace detector must detect traces of nitrate -- that's about the only thing that is common to all common explosives. One trouble is that nitrates also have many non-explosive uses. Walk across a lawn after it's fertilized and your shoes will beep the detector -- that's the likely reason for an incident in San Francisco recently, where they closed the airport after the moron in charge of the machine let the suspect shoes get away. Playing cards will leave nitrate traces on your hands.
And still, it's not going to detect all explosives. NO3 is the usual oxidizer for explosives, but there are other possible oxidizers. The sniffers will catch commercial and military explosives, and probably anything a Tim McVeigh could cook up on his own, but I'm pretty sure my high school chemistry teacher could have made a batch of nitrogen-free explosive if he wanted to. I might even be able to do it myself, if I didn't mind a rather high risk of blowing up the kitchen and myself... A terrorist group with arab oil money could certainly afford a better lab than that high school, and if they could send men to flight school to learn to fly jumbos (but not to land and take off!), they wouldn't have any trouble at all getting a few of their guys into a chemistry major at a US college...
I think that DR just plain didn't have CPM/86 ready when IBM wanted it. Not that Gates had DOS ready either -- he sold DOS to IBM and bought a half-finished program called QDOS = "quick and dirty OS", I'm not sure in which order, then finished it on the fly. That's a Microsoft tradition. Their very first product, Altair BASIC, was also delivered not quite finished and debugged in a marathon session at the Altair plant. Windows is too complicated to do that, way, so apparently they just leave the bugs in.
A business does not have to be smart or efficient to succeed, just less stupid or less inefficient than it's competitors.
It's like the old story about two people being chased by a bear: "I don't have to be faster than the bear, just faster than you."
And actually, I know of companies that have been very comfortable in a niche market for decades in spite of truly terrible management, because by swapping managers back and forth between them, they have both come to be mismanaged exactly the same. That is, their favorite problem solving techniques are shooting the messenger, denying there is a problem, and blaming someone else. They want their decade old equipment redesigned, but no one actually knows how it works, the source code for the control program has been lost, and they can't even tell you how many different configurations have been deployed. Yet, you've got to stay compatible with all of them. And they want the new controller to cost half as much -- sorry, the only thing that will maintain compatibility with machine configurations that we don't even know about is the old controller, and the price of that is going up...
Why haven't they been clobbered by a competitor? It would $100 million to design all new equipment (since compatibility issues will keep you from replacing one piece at a time), then take five years to get your salesmen in to start selling it. American companies don't look that far ahead. Someday some Japanese corporate chief will notice how vulnerable these guys are to anyone that takes the long view and make his grandchildren another billion dollars richer, but it hasn't happened yet.
43... Geraldo with live ammo. Everyone but the Taliban, duck!!! (You're only safe if he's aiming at you...)
OK, for all I know he's an expert marksman. Or maybe he doesn't even know how to release the safety. But on TV in Afghanistan, he looked like the ultimate blowhard. I've known some truly dangerous men, and they act nothing like Geraldo.
So once Paypal has deposited the funds into a bank, they are insured. However, if someone at Paypal takes off to Rio with all the money, leaving Paypal bankrupt, you are scr*wed.
Think that's an impossible scenario? I remember a payroll-processing company where that is almost exactly what happened, almost 30 years ago. The company president disappeared, $5 million was missing, eventually they found his airplane at a remote airstrip in Venezuala, but they never found him.
Yes, searching the raw data would definitely pop up ASCII strings in the slack space -- and it's quite likely the first thing the FBI would do if it was searching your HD for evidence that, say, you were plotting a terrorist attack on the MPAA, would be to search the entire disk for "MPAA" and "bomb".
So, encrypt it before you hide it in the tail. Make sure the encryption format doesn't have a recognizable header. If you don't want to bother with real encryption, exclusive-OR with 0xAA, and it will look like random leftover binary data, just what would be expected to be left in the slack space. Just don't write how and where you hid it on a sticky note...
The market has created clear categories of software that range from the rather unreliable (Windows, piddly silly games, etc) to the extremely reliable
The problem is, MS is selling Windows as being in the "extremely reliable" class, but under the law at present, the fine print in their EULA supposedly means that all their TV ads showing servers allegedly running unattended don't matter. What we need is not laws creating liabilities for all software, but rather laws making it much easier to sue for fraudulent advertising.
Cats just do what they evolved to do, lawyers have gone to school and worked hard to learn how to make so much trouble... 8-)
Actually, the farm cats that I have raised mostly make their kills fairly quickly. They know there are more mice out there, so they don't have to prolong the hunt by playing with the first mouse.
The one problem with extensions is that they do not always indicate a _unique_ file type. That is, there is no central authority to ensure that each extension gets used only once. Maybe this isn't true anymore, but under DOS I knew of at least three word processors that called their files.DOC -- and they were not compatible. I have numerous CAD programs on my computer (got to have anything that any one of my customers uses), and they have conflicting extensions. One of them outputs Excellon numeric-controlled drill files (originally called "drill tapes") with a.TAP extension, which conflicts with some sort of Windows communications setup... There are only 17,576 three-letter combinations, and I am sure that by now there are more distinct file types than that, if you include all the different files used by all the different applications written for Windows. If you want the extension to somehow be meaningful to humans (like.doc), the choices are a whole lot more restricted, and the only way it's going to work is if MS monopolizes _everything_...
Under DOS, this didn't hurt too much because you would type in the name of the.exe first, so if you knew what the file was, you'd pick the right application to handle it. But in Windows, the file is supposed to "know how to open itself", which actually means that the OS goes to the registry to see what application is used for this extension, and if there are conflicts things really go wrong... They badly need another system for identifying file types.
Take for example, Microsoft Outlook (and Outlook Express). The programming teams for these pieces of software were forced to implement a "filesystem within a file" in order to achieve their design goals (I believe the files are called DBX files). Or take for instance, the Windows Registry, or, even better, the Gnome registry, GConf. Why do programmers have to implement dozens of different abstract filesystems in order to achieve their design goals? Simple, the present filesystems are not sufficient.
What design goals are these? Making the e-mail system incompatible with virtually everything else as another way of locking in your customers?
I am forced by company directive to use Outlook. I would much prefer a system where each message would be stored as a text file in the related project folder. (It's possible to export Outlook messages to text, but MS made it as time-consuming as they could.) Full-text searches could be done with tools already in the OS. For other searches (by date, sender, etc.) would be implemented by creating a database with this data plus a pointer to the message file. And I wouldn't have that hundred megabyte outlook.pst file, which is unusable by everything except outlook... Instead, I'd have a bunch of message files, each of manageable size, accessible with any software, and located to be backed up when the project was backed up, plus a database that could be reconstructed at any time just by scanning the message files.
So MS now wants to stuff all files into a framework like the Outlook database? Sounds more like a ploy to sell all new software.
The company claims that it's star wheel that loads 12 rounds per barrel is new. Yes, I have seen gatling style rubber band guns before, but they only fired one shot per "barrel", so you got nowhere near 144 rounds. But what this company really sells, after people get done ooohing at the $395 machine gun, is 12 and 24 round "pistols" and "shotguns" at somewhat affordable prices.
Still, it would be damned nice to smuggle one of these into the courthouse and cut loose at a gaggle of lawyers...
Slashdot Mirror
What is needed are metrics for estimating how hard the problem is, not how hard someone worked to solve it. For estimating the cost of writing a program in the first place, there are various measurements of the problem complexity. One is function points. Google found about 10 pages of links. Here is a FAQ (not approved by the function point users group) that discusses the use of function points to measure productivity, among other things.
Why can't everything else go in a specialized room somewhere
Because then you need extremely high bandwidth connections from that room to each workstation. If the processing that generates your video is done remotely, you need more than a 100MHz ethernet connection to get it to the screen without taking a highly noticeable performance hit as compared to a PC that calculates the picture internally. So out in the real world, where high bandwidth costs more than silicon, you need a powerful CPU and quite a lot of RAM in the workstation.
The answer is less clear-cut with the hard drive. Diskless PC's, where the programs and data all must be downloaded through the network before the PC starts crunching on the numbers, clog the network and performs very poorly. If the software was rearranged so the program ran in the server room, and the PC simply expands the results into display screens, then you could get decent performance. But the way this would usually be implemented would be with one dual or quad CPU box serving dozens of people -- and the performance would bottleneck there. Put enough CPU's in the server room, and find software that works in this "thin-client" mode, and you've got a good system for medium and big business. It costs a little more because you've got (nearly) two CPU's and two sets of RAM per user, but it's quite possible to save more on maintenance than it costs in hardware, if you can find the software.
However, you'll probably discover that most people need at least one program that wasn't written for thin clients, and will not perform adequately if either the disk access or the video goes over the network. So you wind up back to a full PC on every desktop, plus all those extra CPU's in the server room, and even more maintenance...
I think that the CD has to be quite flat for the laser to stay focused. So I assume that their adapter sandwiches the flex disk between two sheets of plastic. (This should also settle the questions about weight and balance -- the adapter is going to be the same thickness and very nearly same weight as a normal CD.) But get a little dust on it, and the disk is no longer flat inside the sandwich...
I once had a somewhat sleazy dealer "forget" to put a key clause into the written contract on a mobile home, but I read it carefully and called him on it before signing. It did get written in. I suspect that he'd actually made the price a bit more attractive than he could really afford, expecting to make it back and quite a lot more on that one change. So he scr***d himself.
IIRC, X-windows preceded MS Windows, and in any case, computer journals used "window" to describe a graphical area on the screen back when Xerox and Apple were the only places working on computers with them. That is, MS tried to grab a generic industry term and turn it into a trademark. The trademark "Zipper" was once voided because, even though it was a new word invented by the manufacturer and sole patent licensee of the teethed fasteners, it became a generic word in common use. For MS to claim ownership of "windows" in regards to computer software is to ask the courts to more than reverse the zipper precedent.
And if MS somehow wins on that, to claim infringement they also have to claim that Lindows as marketed is likely to be mistaken for their own product. (Like a "Bolex" watch.) That shouldn't fly either, because Lindows whole marketing pitch is that it _isn't_ Windows.
No, the _customer_ doesn't do configuration tweaking in black boxes, the _developer_ does. And expect to do lots of tweaking, because for cost reduction, you probably aren't using standard hardware.
Sorry, even Win98 had this beat. If you left it set up at defaults, it is:
1. Insert the CD. An install window opens automatically.
2. Click Yes.
Most of the time it works. What MS hates to discuss it that when it doesn't work, you are likely to be really f*d up. And it leaves too many openings for malicious or just badly written software to install things you didn't want. I assume the Mac has similar vulnerabilities, but since Apple maintains much tighter control over software for the Mac, the chances of an installation going bad is lower.
I agree with you, except for one thing: "Special-purpose machines... Microsoft is set to own this market." Microsoft is _trying_ to grab as much of this market as they can, but they do not and will not own it. The most prominent example of special purpose machines today is PDA's -- and the Palm OS is doing quite well in competition with Windows CE.
But you want to build a special purpose word processing machine. So, do you go spend $$$ on Windows CE and the Word component of MS Office, or do you download Linux, KDE or Gnome, and Star Office? Either way, you put lots of work into figuring out how to configure the software to work on your particular hardware. But with Linux, once it works, you can clone that setup indefinitely, for free. With Windows, you pay by the copy, after paying initially to get the development system. With Linux, you'll have to work harder to make it luser-friendly and hide all the system complexity. With Windows, MS has hidden much of the system complexity even from experts -- if it happens to all work right, that's great, but if it doesn't work, you'll have a hard time getting the info to fix it. And you cannot modify the code, or look at it to figure out why things are going wrong -- not that you want to do this in Linux either, but if things really go wrong it might salvage the project.
And finally, with Windows, there is always the risk that Bill Gates will decide he wants to buy Brazil or something, so next year your license costs jump from $50 to $200. It might make it pretty hard to compete with those $300 Linux boxes, but if you've built your whole business around Windows CE, you might not have a choice. Maybe you'd better co-develop an OSS implementation, just in case.
OTOH, when you market the system, you can piggyback onto lots and lots of MS advertising claiming that Windows is the biggest innovation since the wheel. Unless you try to sell it somewhere that lying advertisements are prosecuted as fraud, or to people that are clueful.
Windows will be a big player in this market, but it cannot dominate it. The basic problem is that MS's tactic of changing data file formats to force upgrades is beginning to wear thin even in conventional PC's; if they start telling people that they have to throw out perfectly good hardware because it doesn't support Word 2005, many will shift to something else instead. And if they stick to existing standards, they are going to wind up competing with _free_ without any really superior features. MS marketing is actually good enough to win this -- part of the time.
Although I do have an old 21" monitor that weighs 70 pounds.
I interpret that question as "is it possible to cool by convection driven by the temperature differentials." Yes, provided:
1. No extreme hot spots -- like Pentiums. They are too small and make too much heat, so I don't think passive air cooling alone can handle it. Scale back your MHz, or add a liquid-filled heat spreader.
2. The case is designed for air flow, the cables are tied out of the way, and the customer can't get inside and f* it up.
3. The case a tower, and it is always operated right side up with the covers on and the vents clear. Techs have to know how long they can keep the power on while the box is open to work on it. A customer that lays something on the vent, or puts his tower on it's side is going to have trouble.
Yes, it is certainly possible to cool fairly high-powered electronic devices passively, but this device wouldn't be used the way PC's are.
the bearings and oil seals are evenly dispersed over the surface area of the heat sink
Huh??? The site got slashdotted before I could download the pdf or anything with a decent picture, but where do you get this? It sure looks to me like the bearing is still on a shaft up the center. It should help that it's a thin shaft with air moving around it, rather than the bearings being stuck up into a hot electric motor hanging above the hot center of the chip.
Are you thinking that the bearing is all around the rim? That would make for a lot of bearing surface with one surface moving at extremely high speeds. If that's where they put the bearing, no wonder the fans cost $100 -- but I still don't think it would be very long lived, because at the high RPM's times the radius, it racks up the miles too fast.
That "explosive" trace detector must detect traces of nitrate -- that's about the only thing that is common to all common explosives. One trouble is that nitrates also have many non-explosive uses. Walk across a lawn after it's fertilized and your shoes will beep the detector -- that's the likely reason for an incident in San Francisco recently, where they closed the airport after the moron in charge of the machine let the suspect shoes get away. Playing cards will leave nitrate traces on your hands.
And still, it's not going to detect all explosives. NO3 is the usual oxidizer for explosives, but there are other possible oxidizers. The sniffers will catch commercial and military explosives, and probably anything a Tim McVeigh could cook up on his own, but I'm pretty sure my high school chemistry teacher could have made a batch of nitrogen-free explosive if he wanted to. I might even be able to do it myself, if I didn't mind a rather high risk of blowing up the kitchen and myself... A terrorist group with arab oil money could certainly afford a better lab than that high school, and if they could send men to flight school to learn to fly jumbos (but not to land and take off!), they wouldn't have any trouble at all getting a few of their guys into a chemistry major at a US college...
I think that DR just plain didn't have CPM/86 ready when IBM wanted it. Not that Gates had DOS ready either -- he sold DOS to IBM and bought a half-finished program called QDOS = "quick and dirty OS", I'm not sure in which order, then finished it on the fly. That's a Microsoft tradition. Their very first product, Altair BASIC, was also delivered not quite finished and debugged in a marathon session at the Altair plant. Windows is too complicated to do that, way, so apparently they just leave the bugs in.
But Digital Research had standards...
A business does not have to be smart or efficient to succeed, just less stupid or less inefficient than it's competitors.
It's like the old story about two people being chased by a bear: "I don't have to be faster than the bear, just faster than you."
And actually, I know of companies that have been very comfortable in a niche market for decades in spite of truly terrible management, because by swapping managers back and forth between them, they have both come to be mismanaged exactly the same. That is, their favorite problem solving techniques are shooting the messenger, denying there is a problem, and blaming someone else. They want their decade old equipment redesigned, but no one actually knows how it works, the source code for the control program has been lost, and they can't even tell you how many different configurations have been deployed. Yet, you've got to stay compatible with all of them. And they want the new controller to cost half as much -- sorry, the only thing that will maintain compatibility with machine configurations that we don't even know about is the old controller, and the price of that is going up...
Why haven't they been clobbered by a competitor? It would $100 million to design all new equipment (since compatibility issues will keep you from replacing one piece at a time), then take five years to get your salesmen in to start selling it. American companies don't look that far ahead. Someday some Japanese corporate chief will notice how vulnerable these guys are to anyone that takes the long view and make his grandchildren another billion dollars richer, but it hasn't happened yet.
43... Geraldo with live ammo. Everyone but the Taliban, duck!!! (You're only safe if he's aiming at you...)
OK, for all I know he's an expert marksman. Or maybe he doesn't even know how to release the safety. But on TV in Afghanistan, he looked like the ultimate blowhard. I've known some truly dangerous men, and they act nothing like Geraldo.
Now, I'll spend $45 on a SoundBug, stick it to my window, and all they'll hear is Britney Spears' songs.
Darn, those CIAnuts like Britney Spears. Put on the yodeling record from Mars Attacks. 8-)
So once Paypal has deposited the funds into a bank, they are insured. However, if someone at Paypal takes off to Rio with all the money, leaving Paypal bankrupt, you are scr*wed.
Think that's an impossible scenario? I remember a payroll-processing company where that is almost exactly what happened, almost 30 years ago. The company president disappeared, $5 million was missing, eventually they found his airplane at a remote airstrip in Venezuala, but they never found him.
Just so long as we don't have to pick up your medical and/or funeral bills. 8-)
Yes, searching the raw data would definitely pop up ASCII strings in the slack space -- and it's quite likely the first thing the FBI would do if it was searching your HD for evidence that, say, you were plotting a terrorist attack on the MPAA, would be to search the entire disk for "MPAA" and "bomb".
So, encrypt it before you hide it in the tail. Make sure the encryption format doesn't have a recognizable header. If you don't want to bother with real encryption, exclusive-OR with 0xAA, and it will look like random leftover binary data, just what would be expected to be left in the slack space. Just don't write how and where you hid it on a sticky note...
The market has created clear categories of software that range from the rather unreliable (Windows, piddly silly games, etc) to the extremely reliable
The problem is, MS is selling Windows as being in the "extremely reliable" class, but under the law at present, the fine print in their EULA supposedly means that all their TV ads showing servers allegedly running unattended don't matter. What we need is not laws creating liabilities for all software, but rather laws making it much easier to sue for fraudulent advertising.
Whos said anything about killing???
Cats just do what they evolved to do, lawyers have gone to school and worked hard to learn how to make so much trouble... 8-)
Actually, the farm cats that I have raised mostly make their kills fairly quickly. They know there are more mice out there, so they don't have to prolong the hunt by playing with the first mouse.
The one problem with extensions is that they do not always indicate a _unique_ file type. That is, there is no central authority to ensure that each extension gets used only once. Maybe this isn't true anymore, but under DOS I knew of at least three word processors that called their files .DOC -- and they were not compatible. I have numerous CAD programs on my computer (got to have anything that any one of my customers uses), and they have conflicting extensions. One of them outputs Excellon numeric-controlled drill files (originally called "drill tapes") with a .TAP extension, which conflicts with some sort of Windows communications setup... There are only 17,576 three-letter combinations, and I am sure that by now there are more distinct file types than that, if you include all the different files used by all the different applications written for Windows. If you want the extension to somehow be meaningful to humans (like .doc), the choices are a whole lot more restricted, and the only way it's going to work is if MS monopolizes _everything_...
.exe first, so if you knew what the file was, you'd pick the right application to handle it. But in Windows, the file is supposed to "know how to open itself", which actually means that the OS goes to the registry to see what application is used for this extension, and if there are conflicts things really go wrong... They badly need another system for identifying file types.
Under DOS, this didn't hurt too much because you would type in the name of the
Take for example, Microsoft Outlook (and Outlook Express). The programming teams for these pieces of software were forced to implement a "filesystem within a file" in order to achieve their design goals (I believe the files are called DBX files). Or take for instance, the Windows Registry, or, even better, the Gnome registry, GConf. Why do programmers have to implement dozens of different abstract filesystems in order to achieve their design goals? Simple, the present filesystems are not sufficient.
What design goals are these? Making the e-mail system incompatible with virtually everything else as another way of locking in your customers?
I am forced by company directive to use Outlook. I would much prefer a system where each message would be stored as a text file in the related project folder. (It's possible to export Outlook messages to text, but MS made it as time-consuming as they could.) Full-text searches could be done with tools already in the OS. For other searches (by date, sender, etc.) would be implemented by creating a database with this data plus a pointer to the message file. And I wouldn't have that hundred megabyte outlook.pst file, which is unusable by everything except outlook... Instead, I'd have a bunch of message files, each of manageable size, accessible with any software, and located to be backed up when the project was backed up, plus a database that could be reconstructed at any time just by scanning the message files.
So MS now wants to stuff all files into a framework like the Outlook database? Sounds more like a ploy to sell all new software.
The company claims that it's star wheel that loads 12 rounds per barrel is new. Yes, I have seen gatling style rubber band guns before, but they only fired one shot per "barrel", so you got nowhere near 144 rounds. But what this company really sells, after people get done ooohing at the $395 machine gun, is 12 and 24 round "pistols" and "shotguns" at somewhat affordable prices.
Still, it would be damned nice to smuggle one of these into the courthouse and cut loose at a gaggle of lawyers...