The issue is not why IDE is cheap but why SCSI is expensive. Maybe the controller chips and cable cost a little more, but the disks dont. I recently went to a SIG talk by a designer from Adaptec. He was asked why SCSI drives cost so much more when they are so similar to IDE. He answered that they often wonder that themselves, and privately the disk manufacturers told him that a SCSI disk costs about $1 more to make. It is just simply a matter of getting away with charging a higher price for a better product, whatever it costs to make. Sux!!
For anything mission critical to Microsoft they use *NIX or mainframes. Hotmail runs Solaris and *BSD, not W2K. Because W2K doesn't scale anything like *NIX. Bollox.
Hotmail is the exception, and it's because the programmers who wrote it (not ms programmers, mark you) were widely reputed to be shit so the whole thing is about as stable as my butt after a chicken korma. The smart option was to just put it back how it was and stand well back, and that's what they did. Everything else at MS runs on Windows - Win2k exclusively since last fall. That includes microsoft.com which is one of the most heavily trafficked, complex site in the world and has *never* been hacked. Folks, I call that a recommendation.
Which 3 year old products does Microsoft support with patches of any sort at all, leaving out y2k fixes? Don't corporations care about things like this? Gosh, let me think. NT4? That's 4 years old.
I do NOT want to have to wander all over the company campus to configure my servers because I can't do it via telnet. How does Terminal Server (or its freeware competitors) require you to leave your seat? On even 10baseT it is much like sitting at the machine. Why restrict yourself to a telnet session when you can have 5 at once and a GUI? Just wondering.
No it won't. The biggest difference is that Terminal Server comes free in all server flavors, and uses minimal server resources when not in use (couple megs ram, no processor). Telnet is also included. Both work great for me. There is no longer any software reason to be physically at the server or to attach a monitor to it. I have even upgraded my W2k build using terminal server, including all reboots, and logged in successfully an hour later.
Does the fact that the only knowledge you need to operate a Win2K server is where to point and click the mouse not raise some warning flags in peoples heads. The logical implication of your posting is that the OS ought to be harder to use so that it attracts only sophisticated admins. How ridiculous. I would suggest it should be as easy as possible to use and the customer should hire the best admins either way.
In RealityLand, Linux "service packs" ie kernel point releases come out much more often than Windows SPs. (The fact that they contain more new features than SPs does not change the fact that they are essentially SPs.) In fact if SPs came out less often, MS would get shat on right here for being too slow. And Linux releases fix bugs that were known at the time the last version was released just like SPs do. That's because everybody who professionally develops complex software knows that those who claim to get *all* the bugs out of a product of any complexity before shipping either ship obsolete software, are ignorant, or are liars. Instead you should test until you are sufficiently confident that the product is at a shippable level of quality for its application, and you are confident that you know all the major issues that you haven't got time to fix, so you can document them and users can work around them. The space shuttle's control software is reputed to have several severe bugs in it, but since they are well known and documented, they can be worked around. Fixing them would involve too much time, money, and (mostly) risk to be worth doing. And the software does its job just fine. What should be criticised is shipping with inadequate testing or serious known but undocumented issues. If the bugs in a SP or release are serious enough that they ought to have been fixed or documented already, they can be reasonably criticised. But simply shipping with bugs is not anything anyone should be embarrassed about, in almost any circumstances. Flame away.
Pffff! You really think that she would have spouted out something like "6.1.4b" if she couldn't distinguish Microsoft Word from Corel WordPerfect? Noo.
I forgot to add the build numbers onto the internal version numbers, as I should have. NT 5.0.2195; Windows 4.1.1998 etc. The build number does not roll over when the prefixes do. It's a good system.
Everyone wittering here is criticising MS for doing their versioning wrong, because it's confusing people. But everyone here is a geek used to non-marketed freeware and these versions of MS products are aimed at non geeks used to marketed payware. MS knows this and deals with it very well. Their big heavily marketed consumer products tend to have consumer friendly names, "Office 2000". This is a moniker that can be marketed. "Emacs 14.3.4a" cannot be marketed. Of course internally they often have the numerical type version numbers you all love, such as Office 9.0 (Office 2000), NT 5.0 (Windows 2000) and SQL Server 8.0 (SQL 2000). You are criticising MS's version numbers because they do not appeal to slashdot readers. They aren't intended to. They are designed to be marketed to consumers. And whatever you think of MS, you have to agree that marketing software products is apparently something that it is extremely good at.
Web collaboration. Once you've really got this going at work it is extremely cool. And the paperclip actually became more useful than annoying for my mother so she doesn't call with questions anymore
NT4 has the rather lame feature of asking you to reboot whenever you visit the networking ui even when you undo changes that you made, and even when you make changes that do not require rebooting. Hardly any of the networking changes actually require rebooting.
Ooh, ooh! That's not as bad as "you have three choices" rather than "you have one choice with three options"... Can we start saying "these data" again please? It's time to bring back datum.
You would do well to be less cynical about the Jury system if you did not believe as you apparently do that those that do not evade it are 'stupid' - as opposed to principled. Coming from the UK, I am shocked at the widespread belief in the US even among people who seem otherwise ethical and moral that one has no public obligation to serve on a jury. All I can do is wish that if you ever are on trial that you are not tried by an unrepresentative jury composed of the stupid and the jobless.
Slashdot Mirror
The issue is not why IDE is cheap but why SCSI is expensive. Maybe the controller chips and cable cost a little more, but the disks dont. I recently went to a SIG talk by a designer from Adaptec. He was asked why SCSI drives cost so much more when they are so similar to IDE. He answered that they often wonder that themselves, and privately the disk manufacturers told him that a SCSI disk costs about $1 more to make. It is just simply a matter of getting away with charging a higher price for a better product, whatever it costs to make. Sux!!
Hotmail is the exception, and it's because the programmers who wrote it (not ms programmers, mark you) were widely reputed to be shit so the whole thing is about as stable as my butt after a chicken korma. The smart option was to just put it back how it was and stand well back, and that's what they did. Everything else at MS runs on Windows - Win2k exclusively since last fall. That includes microsoft.com which is one of the most heavily trafficked, complex site in the world and has *never* been hacked. Folks, I call that a recommendation.
Which 3 year old products does Microsoft support with patches of any sort at all, leaving out y2k fixes? Don't corporations care about things like this? Gosh, let me think. NT4? That's 4 years old.
I do NOT want to have to wander all over the company campus to configure my servers because I can't do it via telnet. How does Terminal Server (or its freeware competitors) require you to leave your seat? On even 10baseT it is much like sitting at the machine. Why restrict yourself to a telnet session when you can have 5 at once and a GUI? Just wondering.
No it won't. The biggest difference is that Terminal Server comes free in all server flavors, and uses minimal server resources when not in use (couple megs ram, no processor). Telnet is also included. Both work great for me. There is no longer any software reason to be physically at the server or to attach a monitor to it. I have even upgraded my W2k build using terminal server, including all reboots, and logged in successfully an hour later.
Now, brainiac, which do you trust? Do you really think, Einstein, a bunch of Linux fanatics with shaky fingers on the vote buttons mean anything?
What if you die while you are rebuilding your kernel?
Does the fact that the only knowledge you need to operate a Win2K server is where to point and click the mouse not raise some warning flags in peoples heads. The logical implication of your posting is that the OS ought to be harder to use so that it attracts only sophisticated admins. How ridiculous. I would suggest it should be as easy as possible to use and the customer should hire the best admins either way.
In RealityLand, Linux "service packs" ie kernel point releases come out much more often than Windows SPs. (The fact that they contain more new features than SPs does not change the fact that they are essentially SPs.) In fact if SPs came out less often, MS would get shat on right here for being too slow. And Linux releases fix bugs that were known at the time the last version was released just like SPs do. That's because everybody who professionally develops complex software knows that those who claim to get *all* the bugs out of a product of any complexity before shipping either ship obsolete software, are ignorant, or are liars. Instead you should test until you are sufficiently confident that the product is at a shippable level of quality for its application, and you are confident that you know all the major issues that you haven't got time to fix, so you can document them and users can work around them. The space shuttle's control software is reputed to have several severe bugs in it, but since they are well known and documented, they can be worked around. Fixing them would involve too much time, money, and (mostly) risk to be worth doing. And the software does its job just fine. What should be criticised is shipping with inadequate testing or serious known but undocumented issues. If the bugs in a SP or release are serious enough that they ought to have been fixed or documented already, they can be reasonably criticised. But simply shipping with bugs is not anything anyone should be embarrassed about, in almost any circumstances. Flame away.
Anyone know what happened in this case?
Pffff! You really think that she would have spouted out something like "6.1.4b" if she couldn't distinguish Microsoft Word from Corel WordPerfect? Noo.
I forgot to add the build numbers onto the internal version numbers, as I should have. NT 5.0.2195; Windows 4.1.1998 etc. The build number does not roll over when the prefixes do. It's a good system.
Everyone wittering here is criticising MS for doing their versioning wrong, because it's confusing people. But everyone here is a geek used to non-marketed freeware and these versions of MS products are aimed at non geeks used to marketed payware. MS knows this and deals with it very well. Their big heavily marketed consumer products tend to have consumer friendly names, "Office 2000". This is a moniker that can be marketed. "Emacs 14.3.4a" cannot be marketed. Of course internally they often have the numerical type version numbers you all love, such as Office 9.0 (Office 2000), NT 5.0 (Windows 2000) and SQL Server 8.0 (SQL 2000). You are criticising MS's version numbers because they do not appeal to slashdot readers. They aren't intended to. They are designed to be marketed to consumers. And whatever you think of MS, you have to agree that marketing software products is apparently something that it is extremely good at.
Web collaboration. Once you've really got this going at work it is extremely cool. And the paperclip actually became more useful than annoying for my mother so she doesn't call with questions anymore
Because the fool who wrote the daemon uses a Win32 call that returns a DWORD?
NT4 has the rather lame feature of asking you to reboot whenever you visit the networking ui even when you undo changes that you made, and even when you make changes that do not require rebooting. Hardly any of the networking changes actually require rebooting.
Ooh, ooh! That's not as bad as "you have three choices" rather than "you have one choice with three options"... Can we start saying "these data" again please? It's time to bring back datum.
You would do well to be less cynical about the Jury system if you did not believe as you apparently do that those that do not evade it are 'stupid' - as opposed to principled. Coming from the UK, I am shocked at the widespread belief in the US even among people who seem otherwise ethical and moral that one has no public obligation to serve on a jury. All I can do is wish that if you ever are on trial that you are not tried by an unrepresentative jury composed of the stupid and the jobless.