The last time I installed Linux was about 2 months ago, and the distro I installed was Slackware 10.2. The Ethernet adapter that it had trouble with was the built-in Intel Pro/100 on the Intel 845 motherboard that I was using. I was amazed when it wasn't detected and properly configured up-front, but it happened nonetheless.
Beyond that I didn't have much trouble; but I did have to compile drivers at one point and run the standard configure/make/make install process which, IMHO, automatically makes the Linux driver install process more complicated for the average user than Windows XP. Any thoughts?
I'm using Slackware 10.2, the latest as far as I know. After I got it configured correctly I love it and it makes quite the reliable file server; but it did not detect my thumb drive unfortunately.
I do realize that there are GUI apps available to retrieve some of this information. But that doesn't change the fact that on Windows I wouldn't have to go looking for it in the first place:)
I can show you software that installs as easy as an RPM. Its called anything that comes in an MSI format.
MSI installs generally give one the ability to choose where and how one is installing the software, and in an intuitive GUI format with plenty of assistance so it doesn't challenge inexperienced users. If I'm in an enterprise environment, I can use msiexec to install the software anywhere I want, any way I want with no user interaction.
As far as drivers go... I call shinanigans on you! I'll agree with you to the point that Windows might not install the correct drivers for my hardware during the installation. But in my experience neither has Linux. I can't count the number of times I've gone through the following process on Linux:
1. Drop to a command-line to find system information about my hardware. 2. Download the driver from my functional Windows machine onto my thumb drive. 3. Plug in my thumb drive on the Linux machine and find that its not functional. 4. Download drivers for my thumb drive to a floppy and repeat most of this process... 5. Compile drivers. 6. Install drivers. 7. Tweak some files in/etc to make drivers work.
Whereas on Windows XP I put in a CD or download a file, then I double-click it. In the days of Windows 98 I might've been inclined to agree with you, but not today.
"If you're going to go so far as to reimage machines, why do you need to spend money for a tool that performs the equivalent of dd if=image of=/dev/hda?"
1. Symantec Ghost is faster
2. Symantec Ghost is easier
3. Symantec Ghost can reimagine image 100 machines just as fast as 1. You might be able to get the same effect with dd, but its gonna take some tinkering, time your LAN admin probably doesn't have if he's administering a (pretty shoddy) network and teaching classes.
3. Just because they're spending your tax money doesn't mean schools are always looking for ways to save it;)
I was misinterpreting what you said before. I suppose I was writing as much in response to the original article as I was your comment on it. Written the way you just did I'd have to at least partially agree with you. Having a monocultural environment is a bad idea in some ways.
This is the way I see it. If we're not talking about the average user and we're not talking about desktop computers, then we're talking IT workers like you and I dealing with infrastructure, correct? Obviously Linux has earned it's place in server-side technology. It has more than proven itself in that area, I won't deny that.
But we've got to be careful at this point that we're not arguing in favor of having a 100% Linux backend since that defeats the "monoculture is the devil" ideal set forth in this thread. Taking a Socratic approach to it, I propose another question: why is a monocultural computing environment a bad thing? Is it always a bad thing? I don't see this as being a universal truth at all.
Without a doubt, introducing a different platform into a computing environment adds a level of complexity. Making Linux,Windows,Macs,Solaris,whatever play nicely together isn't always easy. So why do it? To challenge yourself? Technology for the sake of technology? These are the kind of ideals that your university should be striving for, at least on the educational front.
But on the backend where you've got student and faculty directories, email servers, files servers, grade databases: the greasy cogs of your university spinning to allow it to run efficiently, shouldn't there be a damn good reason for introducing that added level of complexity? Why introduce a Linux mail system into an otherwise 100% Windows environment? Is your organization too broke to afford Exchange? Are there huge leaps in performance? The benefits have to outway the costs in complexity.
Where I work technology for the sake of technology and doing things different simply to challenge yourself isn't an option. Like it or not, we IT workers are only support staff. We're there to make sure everyone else can do their jobs as quickly and efficiently as possible, not to challenge ourselves or to impress anyone with our ability to rewrite Unreal Tournament in binary or whatever.
My boss is the poster child for the Microsoft drones that you talked about. Every time I talk about Linux he (jokingly) calls me a communist or a "crazy chinese hacker" (paraphrased from Steve Ballmer). The problem is, these people see results from their Microsoft programs day in and day out. "I can do that with open source software too" is never a good enough argument; you've got to show them where it excels, give them a reason to give it a shot. A couple of examples:
My boss commissioned me to write a purchase order program for the school, but it was supposed to be a web application "written in any Microsoft language." Granted I am excused from certain things because I'm the boss's nephew, so I pulled the old bait-and-switch on him. When the program was finished I presented it running on a Windows 2003 Server with IIS 6.0. Everyone fell in love with it. After they treated me to lunch, I told him it was written in PHP, not C# or VB.Net.
I used a Knoppix CD on several occassions to retrieve data off of hard disks that he thought were a lost cause.
Used dd and netcat as a free alternative to Symantec Ghost (thanks to a previous Slashdot post).
People like us love technology, and the fact that we love tinkering with it, even off the clock, is the primary reason we (and thus the IT industry) keeps improving. My point is that this type of creativity has a place. As do single-platform environments as well as multi-platform environments. It certainly is upsetting to we technology fiends that our creativity can and will get stiffled in the working world, but like I said: we're there to keep everyone else productive, not for our own geek need
I love Linux myself, but there are several realities that hardcore linux users just can't seem to fit into their already crowded brains:
Fitting as much canned anti-Microsoft rhetoric into your case as you can doesn't help it. A comment like that might get me lynched here, but can anyone honestly say I'm wrong? ; )
Linux really isn't ready for your average Joe and Jane User. You might say it is, but trust me, it's not. You might even say it's as easy to use as Windows, but that aforementioned rhetoric is clouding your judgement. I work at a school, where most of the faculty has trouble grasping the concept of dragging an icon onto their desktop, or saving a file to a network drive. How can you expect a person to comprehend some of the finer points of Linux when they can't wrap their heads around most of the single keystroke/click actions of Windows? Certainly the latest, very beautiful and simple versions of KDE and Gnome help, but they've got a way to go.
In a work or school environment not specifically involved with computers, people have work to do and the computer is simply a means to an end, not the end in itself. Do students and employees have the time to tinker and learn Linux on top of their workload? Probably not. Which brings me to...
Linux faces an uphill struggle. It's just not as easy as telling everyone at your business / school that they are going to use Linux. As hard as they try, linux file formats still aren't 100% compatible with Windows formats, so you would either have a loss in time / money to convert them, or a loss of your previous investment in those files. The same goes for time and money spent to train people on Windows. Retraining them on Linux is most definately going to take an equal-to-much-greater amount of time and money... would it be worth it?
"Standards get in the way of progress."
That statement most definately bothers me... I keep quite up-to-date on web standards, I closely followed the browser wars of the late 90's as well as the similar struggle that is going on today. And I certainly know web standards are not the only standards that Microsoft set its sights on ruining either. But there are two arguments to be made in their favor:
They're not the only ones. Both Netscape and Microsoft did their fair share of tainting W3C standard HTML with proprietary tags. You can rag on Microsoft all you want, but they're not doing anything unique. Besides, they've done a pretty good job of adhering to standards in their recent browsers. Its just a shame we won't see IE 7 for a good long time....
A nonstandard is still the standard amongst the groups that use it. If Microsoft is going to implement something in a totally nonstandard way or in a manner that is totally incompatible with a competing operating system, its a nonstandard that is going to be implemented on millions of machines worldwide. Would that not be the standard in the Windows Community? Do they give a rat's ass if it breaks compatibility with Linux software like Samba? Probably not, and I'm not entirely sure that they should. Microsoft is here to make money. I'll leave the pros and cons of capitalism for you philosophers to debate... me, I've got work to do.
I'm no Microsoft fan, but if the Linux elitists want their platform to prosper, they most certainly need to pull their heads out of their library of Perl & C++ texts and recognize the things that Microsoft has done right with their platform to get it where it is today.
I'm one half of the two person IT staff at a technical school for junior and senior level high school students. From the IT staff perspective, I completely understand why those Macs were removed from your campus. It all boils down to three simple facts: Macs are expensive (yet very pretty!) paper weights, time, and money.
I am the sole person in charge of 250 windows xp workstations in the building, as well as the 10 Apple G4's in use in our Advertising Art course. 260 computers isn't that tremendous a burdon. I'd say its safe to say, however, that I put an equal amount of time into those 10 macs as I do the other 250 computers in the building. There are a few reasons for this:
1. I don't know Macs, and it's not worth the investment to get trained. The school sure isn't going to pay for it, and I'm not going to invest a cent in learning a skill that won't benefit me anywhere else.
2. The instructor isn't a tech person. Sure, back in the day Apple had a repulation for making computers for those that wanted to get work done instead of fiddle with drivers and INI files and network settings and such. But shipping an operating system build on BSD seems to have changed all that. While school is in, I get more tech support calls from that classroom than the rest combined. And when I go fix it, the instructor goes crosseyed and cries while watching. If gurus like Jeff Zeldman (http://www.zeldman.com/daily/lifeisbeautiful/osxb lues/) have problems with OS X, believe me, our instructor is going to have problems too.
3. They don't integrate well. Maybe this has changed somewhat with Panther, but we don't have Panther yet, nor do I forsee my boss taking money out of his budget to buy an entire operating system that should probably just be a service pack (I've been wrong when I said the same thing about MS products too though). I've got plenty of examples to prove this point:
When I first started I didn't know the difference between Active Directory Schema and my own ass, so we ended up purchasing the steamy pile of crap known as ADmitMac from Thursby software. It solved one problem, created a ton more. Users can now log in, only to reboot just about every hour when their ADmitMac-mapped printers fail. (Updates to the software fixed that somewhat....)
Our firewall requires authentication, which is not an option that Apple's Software Update service has given us. So we either run them unpatched or carry them to the server room for the direct connection out to the Internet.
Obviously Macs aren't going to run our login scripts, so custom AppleScript files need to be written and dropped into the home directories of all the mac users.
I've had no success with ghosting a mac. I've tried dd, CarbonCopy, and a few other methods. I got it working, but the process is nowhere near as quick, simple, or even as well-documented as performing the same simple task on Windows or Linux.
If the school wanted to put the investment into training me or someone else on the finer points of administering this handful of machines, we'd be in good shape. But, if the difference in platform is serving no pupose, its not worth the extra investment in training nor the much steeper price tag on the equipment. Replacing them with more manageable, cheaper hardware and software is the only logical solution in my opinion.
Slashdot Mirror
The last time I installed Linux was about 2 months ago, and the distro I installed was Slackware 10.2. The Ethernet adapter that it had trouble with was the built-in Intel Pro/100 on the Intel 845 motherboard that I was using. I was amazed when it wasn't detected and properly configured up-front, but it happened nonetheless.
Beyond that I didn't have much trouble; but I did have to compile drivers at one point and run the standard configure/make/make install process which, IMHO, automatically makes the Linux driver install process more complicated for the average user than Windows XP. Any thoughts?
I'm using Slackware 10.2, the latest as far as I know. After I got it configured correctly I love it and it makes quite the reliable file server; but it did not detect my thumb drive unfortunately.
:)
I do realize that there are GUI apps available to retrieve some of this information. But that doesn't change the fact that on Windows I wouldn't have to go looking for it in the first place
I can show you software that installs as easy as an RPM. Its called anything that comes in an MSI format.
/etc to make drivers work.
MSI installs generally give one the ability to choose where and how one is installing the software, and in an intuitive GUI format with plenty of assistance so it doesn't challenge inexperienced users. If I'm in an enterprise environment, I can use msiexec to install the software anywhere I want, any way I want with no user interaction.
As far as drivers go... I call shinanigans on you! I'll agree with you to the point that Windows might not install the correct drivers for my hardware during the installation. But in my experience neither has Linux. I can't count the number of times I've gone through the following process on Linux:
1. Drop to a command-line to find system information about my hardware.
2. Download the driver from my functional Windows machine onto my thumb drive.
3. Plug in my thumb drive on the Linux machine and find that its not functional.
4. Download drivers for my thumb drive to a floppy and repeat most of this process...
5. Compile drivers.
6. Install drivers.
7. Tweak some files in
Whereas on Windows XP I put in a CD or download a file, then I double-click it. In the days of Windows 98 I might've been inclined to agree with you, but not today.
I was misinterpreting what you said before. I suppose I was writing as much in response to the original article as I was your comment on it. Written the way you just did I'd have to at least partially agree with you. Having a monocultural environment is a bad idea in some ways.
This is the way I see it. If we're not talking about the average user and we're not talking about desktop computers, then we're talking IT workers like you and I dealing with infrastructure, correct? Obviously Linux has earned it's place in server-side technology. It has more than proven itself in that area, I won't deny that.
But we've got to be careful at this point that we're not arguing in favor of having a 100% Linux backend since that defeats the "monoculture is the devil" ideal set forth in this thread. Taking a Socratic approach to it, I propose another question: why is a monocultural computing environment a bad thing? Is it always a bad thing? I don't see this as being a universal truth at all.
Without a doubt, introducing a different platform into a computing environment adds a level of complexity. Making Linux,Windows,Macs,Solaris ,whatever play nicely together isn't always easy. So why do it? To challenge yourself? Technology for the sake of technology? These are the kind of ideals that your university should be striving for, at least on the educational front.
But on the backend where you've got student and faculty directories, email servers, files servers, grade databases: the greasy cogs of your university spinning to allow it to run efficiently, shouldn't there be a damn good reason for introducing that added level of complexity? Why introduce a Linux mail system into an otherwise 100% Windows environment? Is your organization too broke to afford Exchange? Are there huge leaps in performance? The benefits have to outway the costs in complexity.
Where I work technology for the sake of technology and doing things different simply to challenge yourself isn't an option. Like it or not, we IT workers are only support staff. We're there to make sure everyone else can do their jobs as quickly and efficiently as possible, not to challenge ourselves or to impress anyone with our ability to rewrite Unreal Tournament in binary or whatever.
My boss is the poster child for the Microsoft drones that you talked about. Every time I talk about Linux he (jokingly) calls me a communist or a "crazy chinese hacker" (paraphrased from Steve Ballmer). The problem is, these people see results from their Microsoft programs day in and day out. "I can do that with open source software too" is never a good enough argument; you've got to show them where it excels, give them a reason to give it a shot. A couple of examples:
People like us love technology, and the fact that we love tinkering with it, even off the clock, is the primary reason we (and thus the IT industry) keeps improving. My point is that this type of creativity has a place. As do single-platform environments as well as multi-platform environments. It certainly is upsetting to we technology fiends that our creativity can and will get stiffled in the working world, but like I said: we're there to keep everyone else productive, not for our own geek need
Couldn't have said it better myself! Perhaps someone should call the whaaaambulance to pick him up....
I love Linux myself, but there are several realities that hardcore linux users just can't seem to fit into their already crowded brains:
That statement most definately bothers me... I keep quite up-to-date on web standards, I closely followed the browser wars of the late 90's as well as the similar struggle that is going on today. And I certainly know web standards are not the only standards that Microsoft set its sights on ruining either. But there are two arguments to be made in their favor:
They're not the only ones. Both Netscape and Microsoft did their fair share of tainting W3C standard HTML with proprietary tags. You can rag on Microsoft all you want, but they're not doing anything unique. Besides, they've done a pretty good job of adhering to standards in their recent browsers. Its just a shame we won't see IE 7 for a good long time....
A nonstandard is still the standard amongst the groups that use it. If Microsoft is going to implement something in a totally nonstandard way or in a manner that is totally incompatible with a competing operating system, its a nonstandard that is going to be implemented on millions of machines worldwide. Would that not be the standard in the Windows Community? Do they give a rat's ass if it breaks compatibility with Linux software like Samba? Probably not, and I'm not entirely sure that they should. Microsoft is here to make money. I'll leave the pros and cons of capitalism for you philosophers to debate... me, I've got work to do.
I'm no Microsoft fan, but if the Linux elitists want their platform to prosper, they most certainly need to pull their heads out of their library of Perl & C++ texts and recognize the things that Microsoft has done right with their platform to get it where it is today.
I'm one half of the two person IT staff at a technical school for junior and senior level high school students. From the IT staff perspective, I completely understand why those Macs were removed from your campus. It all boils down to three simple facts: Macs are expensive (yet very pretty!) paper weights, time, and money.
I am the sole person in charge of 250 windows xp workstations in the building, as well as the 10 Apple G4's in use in our Advertising Art course. 260 computers isn't that tremendous a burdon. I'd say its safe to say, however, that I put an equal amount of time into those 10 macs as I do the other 250 computers in the building. There are a few reasons for this:
1. I don't know Macs, and it's not worth the investment to get trained. The school sure isn't going to pay for it, and I'm not going to invest a cent in learning a skill that won't benefit me anywhere else.
2. The instructor isn't a tech person. Sure, back in the day Apple had a repulation for making computers for those that wanted to get work done instead of fiddle with drivers and INI files and network settings and such. But shipping an operating system build on BSD seems to have changed all that. While school is in, I get more tech support calls from that classroom than the rest combined. And when I go fix it, the instructor goes crosseyed and cries while watching. If gurus like Jeff Zeldman (http://www.zeldman.com/daily/lifeisbeautiful/osxb lues/) have problems with OS X, believe me, our instructor is going to have problems too.
3. They don't integrate well. Maybe this has changed somewhat with Panther, but we don't have Panther yet, nor do I forsee my boss taking money out of his budget to buy an entire operating system that should probably just be a service pack (I've been wrong when I said the same thing about MS products too though). I've got plenty of examples to prove this point:
If the school wanted to put the investment into training me or someone else on the finer points of administering this handful of machines, we'd be in good shape. But, if the difference in platform is serving no pupose, its not worth the extra investment in training nor the much steeper price tag on the equipment. Replacing them with more manageable, cheaper hardware and software is the only logical solution in my opinion.