I'm not entirely sure where you're getting your information from, but I think you're confused.
I'm an MIT undergrad. All of the machines in the public clusters run Solaris or IRIX. All of the servers run Solaris. There's a very popular student-supported linux distribution. There are more than a few netbsd and freebsd'ers out there as well.
There's really not much of a chance that MIT will go whole-hog to Windows NT any time soon. At a structural level, Windows NT isn't suitable for our type of environment. MIT Information systems has been trying for a few years now to make Windows NT play nice with Athena (MIT's campus network) and, from what I've heard, has been having a miserable time of it.
In general, the _Mac_ is far better supported on Athena than Windows NT.
Yes, a large amount of research work is being done on Windows NT boxes. That's largely due to the fact that those boxes were donated to the Institute. But Microsoft isn't the only one donating software and equipment, either. As I write this, in fact, I'm sitting between a Motorola PowerPC box and a DEC Alpha, both donated by their respective manufacturers.
Yeah, Microsoft does get some mindshare out of the deal, but mindshare doesn't necessarily translate to acceptance. If you spend a lot of time with a crappy product, does that make you want to spend even more time with it?
There's an interesting implied comment here, which goes along the lines of "if no CLI interface exists for a system, then users don't have any power over it". This is due in part by the fact that the Windows and GNU/Linux graphical interfaces do, in fact, provide a limited view of the system and the way it operates. Most win32 and linux hackers root around in the registry and text files (respectively) to get the functionality that they want.
Part of what made the Mac so revolutionary was the fact that the designers decreed that the only way to interact with the system will be through the GUI. The designers knew that there would be no way to fall back on text configuration files if times got tough, so they were forced to implement GUI interfaces to almost every system function.
As a result, they managed to duplicate large portions of the functionality afforded by Linux's configuration files in the GUI./etc/rc has equivalents in the Mac's extensions manager. chmod, in apple's "sharing" window. If the file extensions bug you, try ResEdit. If you feel compelled to build shell scripts, use MPW or AppleScript. Granted, some of these are a bit obscure, but they're certainly available. In fact, I think AppleScript, and it's ability to script lots of different Mac applications, is a place where MacOS has a fairly good lead in terms of flexibilty and power.
As in almost any OS, the amount of flexibility that you get from it depends in large part on the amount of time that you're willing to commit to learning how it works.
Sun (and Oracle) have been saying for a year or two that the "network is the computer", but nobody's yet managed to articulate why thin clients are convincingly better than what we have now. Sure, they're easier to administrate, but are they so much easier to administrate that it's economical for us to yank out the existing infrastructure and replace it?
Sun is a server company, and has been for years. And they make great stuff in that area. But they've made a lot of false starts in the desktop market (such as the Javastation) and folks should be wary.
Slashdot Mirror
I'm not entirely sure where you're getting your information from, but I think you're confused.
I'm an MIT undergrad. All of the machines in the public clusters run Solaris or IRIX. All of the servers run Solaris. There's a very popular student-supported linux distribution. There are more than a few netbsd and freebsd'ers out there as well.
There's really not much of a chance that MIT will go whole-hog to Windows NT any time soon. At a structural level, Windows NT isn't suitable for our type of environment. MIT Information systems has been trying for a few years now to make Windows NT play nice with Athena (MIT's campus network) and, from what I've heard, has been having a miserable time of it.
In general, the _Mac_ is far better supported on Athena than Windows NT.
Yes, a large amount of research work is being done on Windows NT boxes. That's largely due to the fact that those boxes were donated to the Institute. But Microsoft isn't the only one donating software and equipment, either. As I write this, in fact, I'm sitting between a Motorola PowerPC box and a DEC Alpha, both donated by their respective manufacturers.
Yeah, Microsoft does get some mindshare out of the deal, but mindshare doesn't necessarily translate to acceptance. If you spend a lot of time with a crappy product, does that make you want to spend even more time with it?
There's an interesting implied comment here, which goes along the lines of "if no CLI interface exists for a system, then users don't have any power over it". This is due in part by the fact that the Windows and GNU/Linux graphical interfaces do, in fact, provide a limited view of the system and the way it operates. Most win32 and linux hackers root around in the registry and text files (respectively) to get the functionality that they want.
/etc/rc has equivalents in the Mac's extensions manager. chmod, in apple's "sharing" window. If the file extensions bug you, try ResEdit. If you feel compelled to build shell scripts, use MPW or AppleScript. Granted, some of these are a bit obscure, but they're certainly available. In fact, I think AppleScript, and it's ability to script lots of different Mac applications, is a place where MacOS has a fairly good lead in terms of flexibilty and power.
Part of what made the Mac so revolutionary was the fact that the designers decreed that the only way to interact with the system will be through the GUI. The designers knew that there would be no way to fall back on text configuration files if times got tough, so they were forced to implement GUI interfaces to almost every system function.
As a result, they managed to duplicate large portions of the functionality afforded by Linux's configuration files in the GUI.
As in almost any OS, the amount of flexibility that you get from it depends in large part on the amount of time that you're willing to commit to learning how it works.
Sun (and Oracle) have been saying for a year or two that the "network is the computer", but nobody's yet managed to articulate why thin clients are convincingly better than what we have now. Sure, they're easier to administrate, but are they so much easier to administrate that it's economical for us to yank out the existing infrastructure and replace it?
Sun is a server company, and has been for years. And they make great stuff in that area. But they've made a lot of false starts in the desktop market (such as the Javastation) and folks should be wary.